Why Do Commercial Offerings Use Linux, But Not Support Linux Users?

Michele Alessandrini writes "Having bought several TomTom One navigation systems at work, I was browsing their web site to find information about maps. There are several pages of documentation about their devices. In one of them, they proudly inform you that their devices use Linux, as a warranty of power and stability. They even prominently display their GPL compatibility. But, when you come to the software (the one used to manage updates, set locations, etc), they only support Windows and Mac OS. Not that surprising, and not a real necessity. Just the same, they probably saved millions of dollars using a free kernel and didn't think to support Linux users. As Linux gains ground in commercial applications like this, how often are we going to see actual users of the OS left out in the cold? Why don't more Linux-using shops reach out to the Linux-using community?"

Comments

Easy Answer

[c0d3h4x0r]

Why don't more Linux-using shops reach out to the Linux-using community? Because the Linux-using community represents such a small percentage of their customer base that it doesn't make financial sense for them to spend the resources to specifically cater to it.

Re:Easy Answer

[gfxguy]

I just found out Canon uses Linux to run some of their products, but they don't directly support Linux printing.

Now, keep in mind there are third party drivers, but you'd think that those Linux developers they have need to print occasionally.

Just venting...

Re:Easy Answer

[tholomyes]

Well, that's a chicken-and-egg problem, then. One of the reasons most often cited for the prevalence of Windows is the availability of software. Your user base is never going to consider moving to Linux if they can't do x, y, or z with it.

Re:Easy Answer

[Anonymous Coward]

Yup... and it costs money to hire developers (and or spend the time) to port to Linux for little/no money in return from (the very few) customers who would use a Linux client... so it doesn't make financial sense to do it.

Re:Easy Answer

[SparkleMotion88]

Yep. They would have to spend an additional X dollars to support linux and that decision would only net them an additional Y dollars in income. They had some (presumably competent) business analyst folks make a prediction that Y is less than X. These things really are quite simple when you look at them objectively.

That was easy. Next question.

Re:Easy Answer

[everphilski]

Now, keep in mind there are third party drivers, but you'd think that those Linux developers they have need to print occasionally.

Internal devs can put up with a beta print driver. Cannon will not support a beta print driver. Make sense now?

Re:Easy Answer

[anagama]

Because the Linux-using community represents such a small percentage of their customer base ...

But wasn't that part of the point of the summary -- they saved a ton by using a premade OS rather than building their own. What's so hard about giving back to the community a tiny little something. After all, it is that very community that made their profits possible in the first place. It's about good citizenship, not an extra two cents profit per device.

Plus, it really is true that linux users probably affect more sales than just the machines we buy for ourselves. I know I have personally influenced the buying habits 5 other users in the last 24 months (all non-linux users). Get the geeks excited about your product, you'll sell to them and everyone they know. So that two cent loss caused by giving back, might turn into an extra dime profit over all.

Re:Easy Answer

[Hijacked Public]

What's so hard about giving back to the community a tiny little something. It isn't that it is hard, it's just that there is no money in it. They call them for-profit corporations for a reason.

Re:Easy Answer

[SparkleMotion88]

But why should a company support linux just because their gadget has linux running inside it? The group that writes the software for the gadget is probably a totally different group than the one that writes the desktop interface software. And an even more different group is responsible for answering the phone and supporting users.

The software that runs in the device specifies an interface. The software that runs on the desktop makes use of the interface to interact with the device. How the device implements the interface is completely irrelevant. So the fact that the device uses linux has absolutely no bearing on whether the desktop software supports linux.

Re:Easy Answer

[metlin]

There is a difference between using Linux at the backend versus targeting Linux end users.

A lot of places use Linux at the back end or at the device level (for which it works well), but front end, desktop applications? Not so much.

This is mostly because of the fact that despite everything, Windows won the desktop war. They literally own it. There are no two ways about it. So, until that changes, you are going to find companies not particularly targeting Linux desktops.

Re:Easy Answer

[Anonymous Coward]

It's about good citizenship, not an extra two cents profit per device

No it's about business, and being responsible to the shareholders of the company. So, yes, it is about the two cents extra per device. That's how business works.

Re:Easy Answer

[c0d3h4x0r]

What's so hard about giving back to the community a tiny little something. Businesses don't care about "giving back to the community". They care about making money. Period.

Re:Easy Answer

[ryanov]

Ah, but they do their homework and buy Linux compatible office printers!

Re:Easy Answer

[glindsey]

they saved a ton by using a premade OS rather than building their own Not to disagree with you, but for an embedded application as sophisticated as TomTom it would be rare (and foolish) to build your own embedded OS when there are options like VxWorks, Nucleus, QNX, etc. out there. Having said that, yes, they probably went with embedded Linux to save money over licensing one of those OSes.

But as I pointed out in my other comment, it is very likely that the folks that developed the firmware have little or nothing to do with those who developed the support drivers and applications, save for a few architecture/API/integration meetings.

I'm not saying the company as a whole shouldn't be trying to give back to the Linux community, just that you may be talking apples and oranges here when it comes to the software developers involved.

Re:Easy Answer

[bmsleight]

I have a Tom-Tom and only have GNU/Linux machines at home. The Tom-Tom via USB will act as a mass-storage device so you can no most things - heck there are just files on the Tom-Tom. I even have my wife's voice giving me directions. The only thing that is not possible is downloading extra maps. But this can be done via a mobile device paired with the Tom-Tom

Re:Easy Answer

[gfxguy]

See, I'm getting all these serious answers, but you seem to get it.

So these developers at Canon go out and buy HPs to use while they're developing their embedded Linux products.

No, not really, but it seems silly to me.

Re:Easy Answer

[kryten_nl]

Who cares if a company can't provide nice and shiny desktop application. Just give us a command line tool, if there is a need for a GUI, someone will build one.

Re:Easy Answer

[wulfbyte]

I don't like easy answers because they all too often gloss over issues and make things seem so simple.

I am thinking that the OP wanted to highlight that a company that goes out of its way to show that is uses Linux on the back end, still doesn't support Linux users on the front end. Once the stage is set he then asks is this usual and when if ever is this likely to change?

I don't like to think that companies can court Linux users on the one hand by touting how much they use and understand Linux and then still refuse to support Linux using customers on the other without some issues arising. I don't think that this is a matter of financial problems either. If they are only concerned with a non-technical, non-Linux using customer base, then why go to the trouble of advertising their use of Linux and GPL compliance?

I, like the OP wonder if lip service to Linux customers will continue to be the normal practice, even while companies benefit enormously from Linux.

Perhaps one day the monopoly will fail and open standards will prevail, but until then, how do I get support for my Linux machines? If all of us Linux users simply roll over and say we aren't worth the trouble, then maybe we aren't. I do like seeing the question voiced though and I will be happy to add mine, ever hopeful that one day Linux users will be seen as commonplace enough to warrant support.

Re:Easy Answer

[im_thatoneguy]

Ahhhh the gift of giving with expectation of return. The philanthropic spirit of open source.

Re:Easy Answer

[harrkev]

What you say is true, but consider the "family geek" effect.

Brand Z starts to ship decent linux drivers, or at least offers up datasheets.

Geek "Y" decides that he loves this company, and recommends them to all of his friends and family, who trust him because he is the family geek. Suddenly, company "B's" sales increase even with non-geeks.

Re:Easy Answer

[DaveWick79]

I think the device OS is a niche that suits Linux perfectly, because it is (or can be) fast, secure, and stable in a controlled environment.
On the desktop side, providing support for multiple GUI's, multiple distros would require almost as much in the way of resources as their existing Windows support structure, and as the parent stated, for a tiny percentage of users.
Consider also that the average user of Linux on the desktop is at a level that transcends most of the support offered. Those who are in the "computer dummies" category and are also using linux, are an even tinier minority.

Re:Easy Answer

[paiute]

It's about good citizenship, not an extra two cents profit per device.

Actually, business are run by MBAs. It is about the extra two cents profit per device.

Re:Easy Answer

[HarvardAce]

But wasn't that part of the point of the summary -- they saved a ton by using a premade OS rather than building their own. What's so hard about giving back to the community a tiny little something. After all, it is that very community that made their profits possible in the first place. It's about good citizenship, not an extra two cents profit per device. In general, a for-profit company is only interested in "giving back" to the community if they can get a tax deduction for it. Unfortunately, geeks are not a recognized non-profit organization so "giving back" to them doesn't constitute a charitable donation for tax purposes. There are definitely exceptions to this rule, but that is how most companies work.

Re:Easy Answer

[sloanster]

> Because the Linux-using community represents such a small percentage of their customer base that it doesn't make financial sense for them to spend the resources to specifically cater to it.

The only problem with that smug little answer is that linux users have arguably a similar percentage to mac users, and they are not going away, but rather growing. Many vendors, the more clueful ones at any rate, know about and support the 3 major OS platforms: pc, mac and linux. The only explanation I can think of for shutting out potential customers is ignorance.

Re:Easy Answer

[Applekid]

But why should a company support linux just because their gadget has linux running inside it? Because they are benefiting from a mature, open source, and well understood pre-established operating system. If there was no Linux they would have to spend much more development costs in building their own OS for their devices.

I liked the prayer on top of SQLite, actually, for this very reason. Here it is:

** May you do good and not evil.
** May you find forgiveness for yourself and forgive others.
** May you share freely, never taking more than you give. Emphasis mine.

Re:Easy Answer

[Stefanwulf]

If they're fully GPL-compliant, then they _are_ giving back to the community by opening up the source that they develop using GPL'd code. The tweaks, improvements, and extensions that they make to the OS and other applications become available, and that helps make the software that we all use better in the long run. Free software isn't about making people write certain programs or support certain platforms in order to offset the benefit they derive from not having to reinvent the wheel. It's about the way that we all inherently benefit from people having time and money to spend on things other than more wheels. No tithing is required, which is the beauty of it all.

Re:Easy Answer

[dirk]

It may be a chicken and egg problem from your perspective, as someone who wants more software and users for Linux, but it's no problem at all from their perspective at all. Their goal is to sell their product and make money. They don't care if more companies having software for Linux will get more people to use Linux, they care that most people use Windows/OSX, so that is where they target their software.

Re:Easy Answer

[Anonymous Coward]

But wasn't that part of the point of the summary -- they saved a ton by using a premade OS rather than building their own. What's so hard about giving back to the community a tiny little something.

Should I have sent a donation to Edsger because I used Dijkstra's algorithm?

If I built a portable device using a 286 to save money, should I feel the need to support users with 286-based PCs?

I thought the benefit of using free things is that they're free. I didn't know that simply because I use a particular technology in one place means I need to support users who use it in another.

Plus, it really is true that linux users probably affect more sales than just the machines we buy for ourselves.

That's nice, but that has nothing to do with the fact that they used Linux. They could have built a system running on Microsoft Windows NT 4, and supported Linux clients, who would be happy and told their friends to use it too.

Re:Easy Answer

[Anonymous Coward]

Because they are benefiting from a mature, open source, and well understood pre-established operating system. If there was no Linux they would have to spend much more development costs in building their own OS for their devices. Enter the GPLv4... if you ever uses GPL licensed software, anything you produce must work out of the box with GPL licensed software.

Re:Easy Answer

[Anonymous Coward]

There's more to keeping your company successful than just profits.

Re:Easy Answer

[Anonymous Coward]

Giving back a tiny bit is fine. Having done a commercial product under linux in a past gig, it's a whole heckuva lot more than a "tiny bit" to provide something for users to install on their own boxen. Adding something into a linux box you can micro-manage is not hard (like the hand held gadget in the original post). Adding something to a linux box you don't have any control over is somewhere between hard and suicidal. I suppose if you had very strict conformance statments (thou shalt use ______DISTRO________, release ___________NUMBER_______) then it wouldn't be too bad. But without that, you'll end up with someone expecting you to support RHEL, someone else all of the FC's, all the Ubuntu's, every Debian ever released and some that weren't, as well as Fred's Retro Kernel 0.1 with pre-fork gcc, cygnus gcc, and gcc4. Windows Sux but there is are some basic assumptions you can make about what sucky things will always be there.

Re:Easy Answer

[Anonymous Coward]

Actually this isn't true. The linux desktop market represents about 3-4 % of all user desktops according to the W3c; while Mac represents 4-6 %. With the ease of porting from OSx to linux, I would bet the problem lies somewhere else. As a programmer, I know that difficulty does factor into project prioritization. If a project is relatively easy it moves up in the priority list (especially if it helps my product reach a previously untapped market). In my world, we call that low hanging fruit.

My first inclination is believe that linux gets less attention from vendors for 3 reasons:
1) Linux users are widely regarded as cheap. The reputation is that linux users don't pay for apps or support, so why build an app for them.
2) DIY Nuts: the perception is that Linux users would rather use software they can tweak infinitely over software that works but is closed source. I have found myself making choices based on source availability over even functionality and performance.
3) Integration: (This is a the biggie) every time an app is released for linux, the first thing the vendor hears is "You released packages for X Linux but you did not release packages form my distro." Many Vendors will avoid releasing linux application packages even if they are relatively simple, because of the huge disparity between distribution configurations. I know that when my company was making a decision how to make our app available to our customers, we decided to build a webapp versus a fat app simply because it allows us to ignore client system configuration.

Until Linux distributions become more compatible as far as which libraries are available and where they are located by default, we will see the whole market ignored.

Re:Easy Answer

[Altus]

yes, but being a good open source citizen isn't one of those things.

Spend the money on a recycling program, or donating devices to schools or other non profits. Thats the kind of thing that gets you positive mainstream press. Or you could spend the same money supporting linux and get little to no press coverage (and probably more than a few people bitching that your linux desktop software isnt open source, or uses the wrong license or whatever).

Re:Easy Answer

[markhb]

If they are only concerned with a non-technical, non-Linux using customer base, then why go to the trouble of advertising their use of Linux and GPL compliance?

Obvious answer: because "Linux on the inside" is properly buzzword compliant, and the "GPL compliance" part boils down to "Moglen don't sue us."

Re:Easy Answer

[Altus]

This is a big issue for open source. No matter what you do with it, somebody is going to be bitching that your not doing enough. These guys are doing what they are supposed to do (I assume, I don't actually keep track of such things) and they get people in the community bitching that they don't support linux as a desktop system. So lets say the decide to go the extra mile and support linux with a nice, closed source application that does what its supposed to do. Well thats not good, they should make that application open source too. So they go and develop the TPL (TomTom public license) and people take that apart and bitch that its no good and evil and why don't they just use the GPL.

And in the end, almost nobody new actually buys their product because of it. Sure, not all open source advocates are like this, but enough are that it causes some serious headaches for many of the companies who do try to work with the community.

Re:Easy Answer

[Random832]

I'll write a GPL-licensed piece of software that doesn't work at all, thereby forbidding everyone from using GPL licensed software (because it doesn't work with my software)

Re:Easy Answer

[Anonymous Coward]

Not if you are publicly traded.

Re:Easy Answer

[zCyl]

It's about good citizenship, not an extra two cents profit per device.

Actually, business are run by MBAs. It is about the extra two cents profit per device.

But not necessarily smart ones. Good citizenship can significantly increase sales, which can result in far more profit than trying to squeeze two cents more out of each device.

In the past couple decades we've been watching a lot of companies run themselves into the ground by trying to maximize profit margins at the expense of customers. This is a foolish approach on par with, "We may not be making a profit, but we'll make it up with volume." Instead they attempt, "We make an excellent profit on everything sold, even though no one really wants to buy it from us."

Re:Easy Answer

[Anpheus]

It's actually much more simple than that and you don't have to pay an analyst to determine why: The Linux kernel is constantly changing, and while the goal is to minimize breaking changes, it can still happen, and devs won't notice it until it affects a large enough group of people. The result is that vendors can't support Linux, it's a moving and vague target to support. When you say "support Linux" are you talking 2.4? 2.6? Which sub-versions of those major versions are you referring to? Are there specific kernel versions which, for whatever reason, have an issue with your software? Not only that, but Linux isn't even an entire operating system, it's the kernel with which a host of other applications are built around or upon to provide a desktop "experience" (to borrow the Microsoft term.) Now, not only do they have to determine which kernels they support, but they have to determine which versions of GNOME offer the appropriate dbus interfaces (admission: I don't know much about dbus, but insert some other interface there if the reference is wrong.) They have to determine which versions of KDE will support X and Y, and whether or not their application is both functional and aesthetic in GNOME and a half dozen other programs. Saying you support Linux is so goddamn blurry that no one could possibly say that they "support Linux" without exerting control over the entire desktop environment and shipping that as or with their software. For example, RHEL can "support Linux" because they release their entire distro, complete with their own kernel patches, the entire system is set up in a way they've tested and assured will work.

Re:Easy Answer

[rnswebx]

I think it's a bigger issue than simply market share of the Linux desktop. In a word, standardization. Developing for the Linux desktop in a way that requires little to no end-user tweaking has to be a nightmare. The libraries for distro XYZ are completely different than those of distro ABC, and both of those are different than my own home-grown distro I installed on my machine. Uninstalling and re-installing different versions of libraries just so you can use a single application which may or may not break other applications (who are dependent on the version you just uninstalled) is a god damn nightmare.

This is the same problem that Linux has had on the desktop for a long time, and I just don't see the light at the end of the tunnel. How can we expect people to write us applications if we don't have a standard platform for them to write it on? Are they (the vendors) expected to write an app that only works with the libraries of a select distribution (Ubuntu? RHEL? Debian? Suse?) and say the hell with everyone else?

Thoughts?

Re:Easy Answer

[arivanov]

This reminds me of the joke that 2+2 is 5 for sufficiently high values of 4.

I had a hilarious conversation with another geek recently (Mac and Linux using one).

He buys wine on the Internet (can't be bothered to go to the shop). The wine shop recently "upgraded" their software and it stopped working for everything but Windows. He wrote to their tech support and asked why. He got the well known answer - that they do not have the resources to support the development and verification for 3% of the Internet user base.

3 months later they called him with a prolonged and sincere apology and asked him to come back and that they have fixed the shop.

Guess what - 97% of the population that buys wine on the Internet by the case at 20+ quid a pop does not run Windows. More likely - windows is under 40% and even that runs firefox or opera. Rest are MacOS and Linux users.

The decision to cut off all non-Windows users was taken by some moron with an MBA who read some "industry press" and did not even bother asking the operations to run browser stats on the logs. As a result their revenue nosedived by 60%+.

So when someone quotes me 97% numbers I usually ask "Which population"?

If the population under discussion is "Buying luxury goods online" - bollocks.
If the population under discussion is "Geeks buying the latest must-have gadget" - bollocks.
Or even if the population is normalised by its buying power - still bollocks.

Re:Easy Answer

[edmicman]

It's about good citizenship, not an extra two cents profit per device.

Ummm, do you understand capitalism and how business works? Actually, it *is* about the extra two cents per device.

Re:Easy Answer

[AndersOSU]

I even have my wife's voice giving me directions

How many times do I have to tell you? Turn right in 500 feet!

I had a TomTom in my rental care ~ a month ago, and the cheery voice started to grate on me. I thought it would be amusing if there had been a bitch setting (not that your wife is a bitch, in fact, I quite enjoy her company ;P). I think having the GPS make a snarky comment about my lack of sense of direction (and lack of ability to follow directions...) might actually help me chuckle and keep my blood pressure down, as opposed to, "Recalculating route. In 19.1 miles make the next legal U-Turn."

Re:Easy Answer

[dch24]

You answer your own question. Companies like TomTom should buy from Red Hat.

Re:Easy Answer

[Anonymous Coward]

Similarly, proprietary vendors are a tiny portion of the Linux GPS-user market. People who care about freedom will be running OpenStreetMap or TIGER maps on OpenMoko or QTopia or MaemoMapper. Companies like TomTom aren't even going to get a look-in at this market...

Re:Easy Answer

[FauxPasIII]

> Well thats not good, they should make that application open source too. So they
> go and develop the TPL (TomTom public license) and people take that apart and
> bitch that its no good and evil and why don't they just use the GPL.

Yeah, it's almost as if we prefer to shop with vendors that offer good support to their customers up front, rather than having to scratch and claw and wail to get even a pittance. How silly of us.

Re:Easy Answer

[FenwayFrank]

The software that runs in the device specifies an interface. The software that runs on the desktop makes use of the interface to interact with the device. How the device implements the interface is completely irrelevant.

I understand the business argument, but their developers are most likely using some kind of Linux desktop tool to download and test the software on actual boxes. Make that software available.

If that's not practical, do the same thing that makes it possible for my PDA to synch up with my Linux system: document enough of gadget's interface so that it's possible for an open-source version can be written. If what the developers are using is too primitive, they benefit by at least providing that amount of information.

Re:Easy Answer

[jcam2]

You've got that right - I work on a commercial product (Virtualmin) that runs on a variety of Linux distros, and the amount of work required to properly support them all is massive. I haven't done much Windows development, but I imagine that in that market you can tell customers "you must have XP SP2 or later", and still manage to cover 80% or more of the market.

Whereas when doing Linux development, targeting just Ubuntu users would miss the other 80% who use CentOS, Fedora, Redhat Enterprise, SuSE, etc. If there were some standards that distro developers actually followed to ensure application compatability then this wouldn't be an issue ..

Re:Easy Answer

[Applekid]

The tweaks, improvements, and extensions that they make to the OS and other applications become available, and that helps make the software that we all use better in the long run. When was the last time you saw a patch submitted by Canon or Tivo or any of the other companies that continually mooch? They do the minimum requirement to satisfy GPL which is make the source available. Their improvements are there, but, how do you find them? How many of them are true improvements or just hacks specially designed for their marketed product?

And the freedom to modify the code running on the device you bought? When was the last time any consumer device went ahead and let you upgrade it's OS (in a way that wasn't a hack or involved JTAG pads).

If these guys were in the warez world, we'd call them leeches and lamers. I'd say using Linux as a foundation for your product and then not building in support for Linux in that product is worse than just using a closed-source foundation and not supporting Linux anyway.

Re:Easy Answer

[bmsleight]

Have a look at :-

http://www.ghostwheel.de/viftool/

It is indeed very fun and does make my chuckle - at last "She" can read a Map. Also it is fun to have Knife and Fork instead of Right and Left.

Re:Easy Answer

[kelnos]

But why should a company support linux just because their gadget has linux running inside it? Because they are benefiting from a mature, open source, and well understood pre-established operating system. If there was no Linux they would have to spend much more development costs in building their own OS for their devices. You're assuming that corporations, in general, exist to "do good" and aren't generally motivated solely by the desire to generate a profit. Using OSS in their product is great for them; they get to avoid a large amount of development costs. Supporting Linux users is completely orthogonal; some companies may decide that supporting Linux users generates them a net profit they wouldn't otherwise have, and some won't.

Re:Easy Answer

[Altus]

Good support is not defined by the availability of code. Maybe it is for you... but not for the rest of the world. Companies know they can support windows (and the mac) without dealing with any of these issues.

Seriously, would you be happy with tom tom if they provided a closed source application for linux like they do on other platforms?

Re: Easy Answer

[westlake]

they saved a ton by using a premade OS rather than building their own

But the parent doesn't really say that - much less prove it - all it says is that: "just the same, they probably saved millions of dollars using a free kernel."

This is what the boss sees when he looks at Windows Automotive:

Based on WinCE 5.0.

Comes with a full set of familiar - though significantly customized - development tools, APIs and so on.

DirectSound. Direct3D Mobile.

"The Automotive User Interface Toolkit (AUITK) - a GUI framework that makes it easy to create advanced user interfaces at a desktop PC and store them as XML markup.

This separates the user interface from the functional part of the application program." The UI designer does not have to be a programmer.

What does all this tell him about development costs, schedules and staffing?

It's about good citizenship, not an extra two cents profit per device.

That two cents can mean the difference between a stillborn project and product on the shelves at Target.

it really is true that linux users probably affect more sales than just the machines we buy for ourselves. I know I have personally influenced the buying habits 5 other users in the last 24 months (all non-linux users)

Twice nothing is still nothing.

Windows is approaching one billion users on the Desktop. Windows Vista has about the same market share as OSX and Linux - and got there in six months. OS Statistics

Re:Easy Answer

[Anonymous Coward]

What's so hard about giving back to the community a tiny little something.

They already gave back to the "community" by using the software in the first place.

Re:Easy Answer

[ClosedSource]

Right! Even if developers license their code under the GPL, they should realize that people are going to take advantage of them to a certain extent while giving nothing in return. If that's a problem, they should keep their software closed.

Re:Easy Answer

[Bert64]

Hmm, pretty sure i can upgrade all the software on my nokia N800, it too runs linux by default but you can rebuild it and install your own apps too.

Re:Easy Answer

[Anonymous Coward]

Actually the joke is 2+2=5 for sufficiently high values of 2 (not 4).

Re:Easy Answer

[FauxPasIII]

> Good support is not defined by the availability of code. Maybe
> it is for you... but not for the rest of the world.

It is for me. I recognize I'm in the minority in that regard. I don't much care.

> Companies know they can support windows (and the mac) without
> dealing with any of these issues.

-nod- And in most cases they lose my business by doing so. Doesn't matter much as
long as it's just me. Maybe some day there'll be enough people who do care that it
will make financial sense to support us. Maybe it won't. -shrug-

> Seriously, would you be happy with tom tom if they provided a closed source
> application for linux like they do on other platforms?

No, I wouldn't be "happy" until they provided all the code and specification data
so that anyone who cared to could develop their own software to interface with the
device in any way they wanted on any platform they wanted.

Take note; I don't mean to suggest they owe me anything, or anyone else. I'm merely
describing the ideal against which I will evaluate their products when considering a
purchasing decision. Very few products or vendors meet that ideal, so most of my
purchases involve finding the vendor offering the closest to that ideal, or doing
without the product when I can and there are no good vendors.

Re:Easy Answer

[Jherek Carnelian]

Well, that's a chicken-and-egg problem, then. No, it is a "Not my problem" then. Most of the guys have no interest in seeing the linux userbase grow. After all, they are already getting what they need from linux as it is.

Re:Easy Answer

[Seumas]

Not only that, but it's simple logistics (I can't even believe this question was submitted).

See, a company may use a single linux distro for their purposes. They can get support from said vendor or from internal sources for this one distro. If they support customers using linux, they might be supporting a dozen distros or more.

Re:Easy Answer

[dupup]

The result is that vendors can't support Linux, it's a moving and vague target to support.

The company I work for (Sun) makes applications that "support Linux". Perhaps it's a different ball game making enterprise software than it is making desktop software for a consumer device, but it's really rather trivial for us. We nominate a set of distros that dominate the datacenter marker (RHEL, SLES) and say, "We support our software running on versions 2.1, 3, and 4, or 8, 9, and 10, respectively. If you choose to run on another distro, might work, might not, but we don't support it." Maybe I'm missing the thrust of your argument, but we have few complaints about this approach. The advantage is the known kernel version. We even track the updates so we can be sure. I don't see why support for any other app on Linux would be different. Granted it may piss off Gentoo users (I am one!), but it would probably appease 80% of the 3% :-)

Re:Easy Answer

[dna_(c)(tm)(r)]

Cannon will not support a beta print driver. Make sense now?

No. They could publish the incomplete driver as open source. Or make the specs public. Would not cost them much and gain them some extra business, goodwill...

Re:Easy Answer

[SavedLinuXgeeK]

That doesn't mean that there is no profit in giving back, I mean Google et. al have found the benefit of giving back to the community, and the increased profits and word of mouth advertisement, etc. I mean, if you just weighed the time, effort it takes to give your changes back to the community in cost of advertisement, I would **assume** that you would at least make even if not gain on the entire endeavor (and after the initial setup of giving back, each additional change is minimal to give back to the community and the continued reputation improvement and word of mouth advertisement would inevitable outweigh any costs of giving back to the community)

Re:Easy Answer

[darkpixel2k]

Because they are benefiting from a mature, open source, and well understood pre-established operating system. If there was no Linux they would have to spend much more development costs in building their own OS for their devices.

That's retarded. It's like bringing in cookies to work and telling everyone to chow down and enjoy. Then a few days later, you start bitching and moaning that no one else brought in cookies, or made your recipe any better, or whatever.

Linux is given to them FREE. There is no requirement that (because they use it) they suddenly have to improve it and give it back. The knife cuts both ways.

Re:Easy Answer

[everphilski]

Presuming that is an option, releasing something they can't/won't support is generally considered a Bad Thing by most people. The overwhelming public perception would probably swing negative...

Re:Easy Answer

[DrgnDancer]

And, as a general 99.9% of the time rule, if you software works on those specific distros, it can be make to work on almost anything else. Sure that Ubuntu user may have to go out on the Internet and find the website of the clever guy who figured out which libs need to linked where, but chances are that someone figured out how to make your product work on every distro with half a following and has a website with the needed info. Honestly, if you can more or less keep your software up with one of the major vendors' releases, someone will figure out how to make it work with everything else, even if you don't release source code.

Re:Easy Answer

[Jimmy King]

Why do the linux devs necessarily need to be printing from linux? I develop software that runs on linux, but I print from windows. All of my development is done on a remote server via ssh while my workstation (unfortunately) runs XP.

Re:Easy Answer

[RobertLTux]

in the Closed > GPL device range you have a number of options
1 fully closed deliberately locked
2 closed but the protocol is guessable
3 closed using a known protocol
4 spec sheet available
5 closed but workable ap given
6 open but not hackable ap given
7 GPL ap given
I would like to run into somebody that has an Ximeta NDAS box that can code a GUI ap for the driver (i would also like to get a current and working driver but...)
btw somebody needs to convince Ximeta that they can get what they want and GPL the linux driver

Re:Easy Answer

[PenguinBoyDave]

Simply put, it's mind over matter. They don't mind, and you don't matter.

Re:Easy Answer

[sgartner]

... The group that writes the software for the gadget is probably a totally different group than the one that writes the desktop interface software...

It could even be worse than this. Often a company that is good at embedded systems (like TomTom) is no good at writing application software, so they completely outsource the development of the application and just provide them with API documentation for the device. I don't know that this is the case with TomTom, but I've known of companies that did this (they even stupidly closed the vendor contract after the application software was written, thereby making updates both expensive and difficult).

The world is not homogeneous...

Re:Easy Answer

[ednopantz]

Isn't the linux user's first question: why isn't this free?

Who would want customers like that?

Re:Easy Answer

[JimDaGeek]

Bad analogy. A more correct "cookie analogy" would be:
I work in IT and bring in cookies every Friday. I give my cookies to Accounting, IT and HR. Someone from Accounting, who eats my cookies, brings in cookies every Wednesday. However he/she only shares his/her cookies with Accounting and HR. Is the person from Accounting required to share cookies with IT? No, but it is a pretty crappy thing to not share their cookies.

Yeah, cookie analogies are pretty dumb ;-)

The way I see it is that TomTom is saving a nice chunk of change by using OSS/GNU/Linux to build the base of their systems. It would be nice if they took a small part of those savings and just... maybe... wrote some software for OSS/GNU/Linux users. Hell, I am sure they saved enough by using Linux in their devices to hire just one Linux GUI developer to build an equivalent GUI software that is available for MS Windows and Mac. It is not like they are making tons of money from Mac users. The majority of their users will being using the devices under MS Windows. At least WRT a Linux GUI, they can say the cost was offset by the savings generated by using Linux.

Re:Easy Answer

[Anonymous Coward]

If the decision is to always round down

2.5 + 2.5 = 5

would be expressed in the integer domain as

round(2.5) + round(2.5) = round(5)

which is

2 + 2 = 5

so the joke is technically correct, just that it omits to state its assumptions.

Re:Easy Answer

[greenbird]

Actually, business are run by MBAs. It is about the extra two cents profit per device.

And it's a sad fact but those MBAs for the most part are clueless as to how to most profitable run a business. They're to busy counting beans to understand that supporting the Linux community actually helps their business since their whole business relies on a strong Linux community.

Re:Easy Answer

[mrchaotica]

The "overwhelming public" would fail to notice that the unsupported Linux driver even existed, much less have a negative reaction to it! The only people who would notice would be the Linux community, which is almost entirely composed of geeks that aren't scared by unsupported code (especially if it was also Free Software itself).

Re:Easy Answer

[JimDaGeek]

But they did write software for Mac. Mac has only about 4% or so of the desktop market, that is certainly not a huge target group. At least if they made a Linux GUI/control software they could say they used the money they saved by using Linux/OSS.

Re:Easy Answer

[morcego]

I'm pretty sure judges in many countries would also have problems with a company "releasing something and not supporting it".

That is sure to be another show stopper.

Re:Easy Answer

[xenocide2]

Even for-profit companies have a line item that says "Goodwill" on the budget.

Re:Easy Answer

[Bellum Aeternus]

That's becoming know as the Ubuntu Effect. My mother actually asked be about "On Blue Two", took me a moment but I figured it out (they really do need a better name). She'd heard of it through a coworker, who'd heard of it from her 17 year old geek of a son. You're absolutely right, people will trust technology when they believe that someone will be around to help them when it breaks. The problem with Linux is that it's not easy to find somebody to help fix it - with Windows you can go to a number of local shops, and there's always somebody's son/daughter who can fix it in your office. It's virtually the same for Mac (plus their advertising makes people believe they never break - don't believe it).

Re:Easy Answer

[mrchaotica]

What's so hard about giving back to the community a tiny little something.

Nothing, which is why they do give back, in the form of complying with the license and releasing the code for the device itself! They've already done everything we expected them to do; conversely, if you want them to do more then write that into the license requirements next time.

Look, I'm about as pro-Free-Sofware as they come. I'm even an EFF and FSF member (are you?). But do I have a problem with a company not going above and beyond the license requirements? No, because "above and beyond" is called that exactly because it's not required.

Sure, it would be nice if the TomTom people gave me a free GPS unit, a million dollars, and a pony, but I'm not about to start bitching at them for choosing not to do so! In exactly the same way, it would be nice if TomTom supported Linux, but I'm not about to bitch at them for choosing not to do that, either.

Re:Easy Answer

[kaizokuace]

Yea but as we all know (i hope) word of mouth advertising is the most powerful form of advertising possible, and the cheapest. On top of that geeks are the ones that will rabidly praise products they like to anyone willing to listen to them talk. Those people that find the new and cool things are the ones that control the word of mouth networks.

Re:Easy Answer

[Anonymous Coward]

Well according to their webpage they support only WinXP, 2000 and Vista, so I don't see why it would be a problem to require linux kernels after 2.6.7 and Qt versions greater 4.1.3 (just examples) for their linux support.

Re:Easy Answer

[Anonymous Coward]

Then you support RHEL and SLES, you don't support linux. Not to mention that datacenters are a lot more of a niche market than GPS users.

Re:Easy Answer

[Chris Mattern]

Even for-profit companies have a line item that says "Goodwill" on the budget.

Which has a very specific meaning relating to accounting for corporate mergers. "Goodwill" has nothing to do with going out and making people feel good about your company.

Chris Mattern

Re:Easy Answer

[cstepan]

I wonder how many people will post dsimissive comments of the variety, "Because it's not profitable to do so, you moron." I would like to believe that all the very smart people out there who read/post to /. would be able to figure out that the original poster probably already knew that answer. I think a more pertinent question would be:

"How can we convince companies that use Linux in their products that it's in their best interest to support the very geeks who produced Linux in the first place, when any mongoloid (with or without an MBA) can spout out extremely obvious and simple-minded answer that there just aren't very enough Linux users to waste our precious resources on?"

I think a combination of shame and threats to do something like plant malicious code in the kernel. But I grew up Catholic. Shame and threats are all I know.

Re:Easy Answer

[vosester]

"They have to determine which versions of KDE will support X and Y" well supporting X is a bitch but i awfully sure Y is dead http://en.wikipedia.org/wiki/Y_Window_System

Re:Easy Answer

[mpcooke3]

Yeah also if you waste resources "giving back" and your competitor doesn't, then you are at a disadvantage.
That is unless "giving back" gives you some direct benefit rather than a general feeling of wellbeing.

Realistically if you want companies to give back, then you have to explain whats in it for them. Eg, building a community, etc,

Re:Easy Answer

[c0d3h4x0r]

Your revised question also has an easy answer: contribute in any way you can to make Linux more widely acceptable. Once it's finally good enough that most people prefer it over Windows, then it will have enough marketshare for it to be worth companies' efforts.

Re:Easy Answer

[DarkVader]

Your numbers are old.

The Mac has over 6% of the US desktop market, and nearly 15% of the US portable market - and GPS is more likely to be used with a portable.

Which is not to say I don't want to see a Linux desktop client here - I certainly do.

Re:Easy Answer

[Hijacked Public]

Yeah, but it isn't thankless giving back to any communities to even things out, it is photo ops and newspaper articles. An executive ladling out soup in a soup kitchen. Charitable Donations.

This kind of thing comes up a lot at Slashdot, this expectation that corporations are obligated to act equitably. As annoying as Jon Katz was most of the time, he wrote one of the better editorials ever posted here. (this was back when we actually had editorials rather than the occasional rant against toolbars that rate Slashdot poorly).

At the very least he explained the way corporations view things in what I thought were geek-understandable terms. Maybe everyone missed it, I don't know, but he pretty well conveyed the total absence of any moral or ethical deliberations prior to doing any given thing. The guy went into an interview expecting to argue his case from an "is this the right thing to do?" standpoint and got "if we don't do it, somebody else will.".

An absolute lack of even the desire to consider, at all, in any way, his question.

Anyway, same thing here. An accountant looked at a cell in an (Excel) speadsheet labeled "Code our own OS". It had %1,000,000 in it. The next cell "Use Linux" has $0 in it. They could legally use it without giving anything back so they legally used it without giving anything back.

No one is going to motivate them with stories about how Linux was created through the selfless work of volunteers. Their spreadsheets have assigned $0 to that as well. Show them that the market is worth something and maybe they'll pay attention.

Re:Easy Answer

[fishbowl]

>What's so hard about giving back to the community a tiny little something.

The demand curve doesn't justify it, and that makes it *very* hard to, say, go to the board of directors and convince them that your "giving back" idea is necessary or even beneficial.

What's so hard about getting into a position of authority where you get to make these decisions? You're smart and know what makes a successful business decision better than them, right? Make a business plan and get investors!

Re:Easy Answer

[Anonymous Coward]

As a developer, I don't think the kernel as a 'moving target' has as much to do with companies not supporting linux as the fact that each distribution of linux is so different. For example, on Windows and Mac OS X, we know where the libraries are. They're always in the same place so we can just link to them and everyone's happy. But who knows where a certain library might be installed on any given Linux system. I'd need to release a differently-linked binary package for each differing distribution... for every version of my all my programs! Obviously that's not cost-effective, and therein lies the real problem. Someone needs to come up with some userland standards.

But I agree with you in that most of us don't support Linux, not because of market share (we really do want to have our stuff run on linux), but it's just too hard/expensive to support at the moment, unless we just jump behind ONE distribution, which may be happening with Ubuntu.

Re:Easy Answer

[the_bard17]

Unfortunately, geeks are not a recognized non-profit organization so "giving back" to them doesn't constitute a charitable donation for tax purposes.

So maybe we need to get our arses in gear and start lobbying for tax breaks to those corporations that contribute to the open source community.

Yeah, I know... lobbying & tax breaks for corp's are bad/evil, but giving back to the OS community is good, right? Ends justify the means, and all that...

Re:Easy Answer

[Watson Ladd]

Why do you think POSIX and the LSB exist?

Re:Easy Answer

[solid_liq]

Enter DKMS, the Dell project, which abstracts the kernel version from the device driver. This project renders your point about device drivers moot.

See more here: Dell Linux Projects.

Re:Easy Answer

[dupup]

Then you support RHEL and SLES, you don't support linux.

Agreed. I submit, though, that if Tom-Tom supported RHEL and SLES, that would make plenty if not most Linux users happy. As a sibling to your post points out, once this hurdle is crossed, support for other Linux distros becomes a lot easier.

Not to mention that datacenters are a lot more of a niche market than GPS users.

Agreed again. However, this is the extent of my professional and relevant experience with supporting Linux. If I were in the business of writing Linux apps for more general consumption, I'd post about my experience with that.

Re:Easy Answer

[JimDaGeek]

I guess that depends on whose numbers you look at. At MacNN it says that Apple had 5.6% of shipments for Q2 2007. Then at this site it shows Mac had 3.33% of the market share. So who knows the real numbers. I would bet it is between 3% - 5%.

Re:Easy Answer

[Anonymous Coward]

+5 cool claymore reference in sig!

A real Claymore reference would have to include something about Raki being such a worthless emo retard, and the entire last eight or so episodes sucking horrendously. Evem the zomg-2nd-season ending was totally pathetic.

And it started off so cool... >_>

Re:Easy Answer

[Bluesman]

Wife's voice, eh?

Can you also have it second-guess the way you're driving and change its mind about which way you should turn at the last minute? Or how about having it shout "Oh my GOD!!!!" at random while you're driving in traffic, and then telling you that they're putting in a new Banana Republic at the shopping center you just passed.

Until I can buy a GPS that does that, I'll stick with the real thing.

Re:Easy Answer

[Anonymous Coward]

companies pay $$$s for tax consultants specifically so that they don't have to give back to the community.

they save money twice here by not even having to pay consultants.

Re:Easy Answer

[Anonymous Coward]

Why don't more Linux-using shops reach out to the Linux-using community?
Because the Linux-using community represents such a small percentage of their customer base that it doesn't make financial sense for them to spend the resources to specifically cater to it.

If you believe that, then I have a bridge to sell you.

McAffee Webshield e500 is a stand alone network virus scanner. Its web interface has a java applet that only works under Internet Explorer. Not, Netscape, Mozilla, IE for Mac, or HotSpot. Yes, this appliance runs Linux.

APC IPKVM for Sun servers works via an Active-X applet. Are you going to tell me that a shop full of Sun servers will have Unix admins with XP or Vista on their desktop? Does it even matter that this appliance runs Linux? Talk about not researching your user base!

Re:Easy Answer

[Anonymous Coward]

That's why the release of specifications is preferred.

It's easier to support if you have to, and avoids $500M lawsuits (EU vs MS).

Re:Easy Answer

[curlynoodle]

Okay, I can understand that argument. Linux desktop support is troublesome for an application vendor. Why not then release an unsupported yet documented SDK helping Linux desktop users to build a community supported interface?

Re:Easy Answer

[NoMaster]

Yeah, it's almost as if we prefer to shop with vendors that offer good support to their customers up front ...

Yeah, but face it - the product is shiny and cool; you're going to buy it anyway.

Or are you going to give up your TomToms and Linksys routers and Canon cameras and Tivos and ... ?

It's great to have ideals; it hurts when others don't abide by them; and it's hard to maintain them when they come up against something you really really want...

Re:Easy Answer

[NoMaster]

But, when comparing Mac market share to Linux market share, you've got to remember this:

1. Mac users won't touch your product unless it supports the Mac out of the box
2. As soon as the first blog post containing the line "And, guess what? It runs Linux!" appears, Linux users will start buying your product - and somebody will be working on an app / driver to make it work.

Why do the work of writing an app, when you know a bunch of naïve idealists will gladly do it for you?

The only real downside is you end up with a bunch of petulant whingers when you change your hardware / protocol. And, even then, they're dumb enough to quickly get right back to the hard work of supporting your product for you!

Re:Easy Answer

[Mista2]

I remember a slashdot admin saying that on their stats, 60% of internet users would be running a browser other than IE, mainly becuase the audience tend to run all sorts of systems and even with Windows often don't settle for the default. Trade-me here in NZ were doing a site refresh. They wanted to change the layout to that it looked better at 1024x768. But when they ran their numbers, they found a large population of users browsing with screen resolutions of 640x480 or 800x600, or many people just keeping trademe running in a smaller windows while they monitored an auction, not running at full screen size. So they had to make sure all sorts of systems would still work well, and support multiple browsers (this was still when IE had poor css support.) Small gadgets are going to be primarily bought by geeks in the first instance, especially when they are new. If there are several devices or applications being released that do the same thing, you need to do something to stand out. Support all platforms I say! I don't use skype because they cripped the client in linux (no file transfer and no video, and wont talk to other IMs) so I use Gizmo, full featured on Windows and Linux, regular updates on both platforms and can talk to some other jabber based IM's) Both do the same thing, but I chose one for better platform support. However, most will be happy with Skype as they can talk to their grandma overseas just fine. Go figure

Re:Easy Answer

[dbcad7]

I was thinking along the lines of the "Nvidia" effect, where good support for linux has made it the card recommende by geeks regardless of OS

Luckily, I don't think there is much support needed for a Linux user. If it can't be installed through Synaptic, then it's probably not a good idea for a non tech user.. luckily pretty much everything asked for is in there... although the multimedia codec thing is a bit of a problem.. I've been looking at Linux Mint though (I want to call it Mint Linux.. but I don't think that's correct) which may be an alternative as a give away distro, as then you don't have to explain how to get codecs installed. (although the "light" version you would). I haven't tried it out yet, so can't say yet, one way or the other, if this would be better for newbies than Ubuntu.

Re:Easy Answer

[SEWilco]

If you intend to write a piece of software that doesn't work at all, then when it doesn't work it will be working. So in order for it to not work at all it will have to work.

Re:Easy Answer

[Jimithing DMB]

The RHEL support seems to be a big one. Most of the time, software designed to run on RHEL will also run on properly-configured Fedora and of course also on the generic builds of RHEL built from the same source packages Red Hat used. Then again, I am primarily familiar with Red Hat's offspring so I have no idea how well Debian-based distros fare. I imagine if a product works on both RHEL and SLES then it probably has a pretty good chance of working on a Debian-based rig as well.

Then again, I am really only familiar with server products. The server space seems to work well for Linux because most people are happy with the built-in functionality. Where that functionality is lacking, they are (at least now) able to purchase products which will run on their server. Most importantly, they are willing to do at least a little bit of work to get products running if they don't happen to be using one of the officially-supported distros.

The desktop space is much harder to get into. Desktop Linux can be good for people with little to no knowledge of computers if someone else sets up and installs the system. But as a third-party software vendor you run into a problem. It's very hard to write software that works cross-distro and there really isn't a clear winner in the desktop distro market. On top of that, typical desktop users aren't willing to futz with various files just to get a new program installed. Of course, there are desktop Linux users with knowledge who are willing to futz with the system to get a third-party program installed, but that's a niche of an already niche market.

I used Linux as my main desktop system for a number of years and found it reasonable. But then again, I'm a developer. But for about 5 years now I've been using Mac OS X on my desktop. The reason is simple. I can do nearly everything I could do with Linux (MacPorts is great) plus I can "cheat" once in a while and (gasp) actually buy a program that runs on my machine.

In fact, I think one of the biggest attractions to Mac OS X is the availability of third-party programs. Sure, there's less than Windows but what is available is generally really damn good. For instance, GPSNavX and its newer brother MacENC are top-notch marine GPS navigation programs. Likewise, EyeTV is a damn cool piece of software for watching and recording TV and, perhaps most importantly, they have a list of supported capture hardware and their supported hardware is actually AVAILABLE for purchase, even in a big box store. Oh, and did I mention that Microsoft makes Office for the Mac and that it actually works better and generally more intuitively than the Windows version? And if you don't like Microsoft and/or don't need perfect Word/Excel compatibility you can always buy iWork for $80 now. There is basically no alternative to Office on Linux. OpenOffice.org? Please. They sure succeeded in creating a near clone of the Windows version of Office. I wonder who the hell really wants that though? I want something better.

Then there's the open source stuff. VLC on the Mac is great. You drop the app bundle somewhere on your drive. Double-click. Done. You don't go hunting through your distro's available software list. You don't go find the .rpm or .dpkg for your distro and double-click to install it and then get to use it. You drag the damn thing into your Applications folder and you're ready to roll. Likewise for Mozilla stuff like Firefox, Seamonkey, Thunderbird, and Camino.

Furthermore, there's even a lot of stuff out there that is cross-platform between Windows and Mac. Take something like TurboTax or TaxCut. You don't see those on Linux but by god there's enough people doing their taxes on Macs to justify porting them. The quality is a little shitty, looking like a crappy port from Windows, but hey, if I can do my taxes without rebooting into Windows, great.

I could go on a lot more but the bottom line is that desktop Linux isn't really headed in a direct

Re:Easy Answer

[anagama]

And if you read all the way through, I suggested that penny pinching like that might cost them money overall. People I know ask me for my opinion about electronics -- as in they tell me what they want to do, I tell them what to buy, and they do it. Make me happy and I bring my own money plus a bunch more. Make me unhappy, and the manufacturer loses my money plus a bunch more. At some point, it is penny wise and pound foolish to be a poor corp. citizen. Are we there yet? Doubt it. Will we get there? Probably.

Re:Easy Answer

[jsebrech]

There's more to keeping your company successful than just profits.

Nonsense. There are tons of companies that care only about profits and keep doing exceedingly well. Walmart, microsoft, most banks, most oil companies, most drug companies. Corporations have figured out that if you behave like an ass but are cheaper, people will still buy from you.

Re:Easy Answer

[Anonymous Coward]

So let's see what really any portable device needs in terms of "resource spending":

you need to transfer data: easy, if either the device runs some kind of ssh or ftp daemon (as it already runs Linux). Alternatively, make the device support USB mass storage (which all good devices do, anyway). This also doesn't require the user to use any specific OS.

Secondly, you need to manage the device, maybe sync with it. This is easy. Just write something using cross-platform libraries, such as Qt, Wxwindows, Java, Python... All of this is so ridiculously easy, and the costs to produce something like it are so much not-more-expensive than "regular" software development, that I can't believe no business ever plans for the future and states portability in their software requirements when the engineer the damn shit.

Re:Easy Answer

[Blkdeath]

Internal devs can put up with a beta print driver. Cannon will not support a beta print driver. Make sense now?

Actually based on my experience, most notably with their multi function devices, their drivers are beta. When I asked why all users of a Windows XP workstation had to be administrator (not even Power Users would suffice) in order to even PRINT to the device I was given two pieces of pertinent information;

1. the device "does more than most" so it was necessary. The fact that I could have plugged in 4 separate devices and accomplished the same thing but with less available desk space I was brushed off.
2. Their drivers are programed in Japan which, somehow, means the North Americans have no control. So I guess it's kinda a roll of the dice when those Japs are done with them. Or something. I'm still not sure. My car was made in Japan, but it still works on Canadian roads. {shrug}

Re:Easy Answer

[joto]

I'm pretty sure judges in many countries would also have problems with a company "releasing something and not supporting it".

Uhm, no. Maybe if you sell it, and if you sold it under such circumstances that it is reasonable for the buyer to assume that he gets a working product, and you refuse to return the money to the buyer when he tries to get a refund for his useless new acquisition. Otherwise, forget it, you're not getting rich by suing companies releasing unfinished or incomplete open source software just yet.

Re:Easy Answer

[Torvaun]

Japanese car running on Canadian roads is a bad analogy for software drivers. Try getting a Honda and a Ford, then switching the engines. I'm not a car guy myself, but I have it on good authority that you can create one hell of a mess by trying it.

Re:Easy Answer

[MadJo]

I have a TomTom device, and a Linux pc at home.
I only use Linux, no Windows, and have no inclination to do that.
So from me they won't get money to get new maps.
THAT's a lost sale!

Re:Easy Answer

[tuxicle]

How about a program like Google Earth? Fairly complex, uses OpenGL for 3D, etc. Runs without trouble on several distros.

Re:Easy Answer

[Johnny Mnemonic]

It's about good citizenship, not an extra two cents profit per device.

Boy, are you wrong. I'm sure you'd like to think that, but it has simply never been the case that a company has acted out of "citizenship" instead of "2c profit per device". That's 2c they get to keep and spend on something they want to spend it on, instead of doing you a favor.

If you don't like the terms of the license, don't use it. It's as simple as that. If you can't bear to see your code used for free, with nothing in return, well, you have standard copyrights to use. "Freely use" includes acting in ways you don't like.

Re:Easy Answer

[isorox]

I even have my wife's voice giving me directions

Thought about that, but I instinctivly go the other way when SWMBO gives me directions, she's that bad!

Re:Easy Answer

[Goth Biker Babe]

Because they are benefiting from a mature, open source, and well understood pre-established operating system. If there was no Linux they would have to spend much more development costs in building their own OS for their devices.

Me laughs loudly. I'm currently working on an embedded product using this mature, well understood, well established operating system. We're currently investigating a core operating system issue which should work but just doesn't in this case and results in a crash which is recreatable with a small test application. We've found the cause and it will be released to the open source community but its such an obvious fault I'm surprised that no one has found it so far. It would have definitely been found if there were unit tests and quality control procedures which there aren't. We found it because we have unit tests and quality control procedures for our software which runs on the top.

Because I know linux quite well and will stick my arms, up to the pits, in to the code, I'll not use it as a day to day operating system. I have a Mac. Considering the relative user base size I'm not surprised that commercial companies avoid the issue of developing for Linux.

Re:Easy Answer

[ReceptiveIT]

I am the director of an IT company looking after 100+ sites with a mixture of Linux and Windows OS. Some sites are Windows only. I still make my purchasing decision based on how well certain hardware works on Linux. I may be small, but Linux friendly suppliers get the majority of my approx $500,000 per year in IT spending. I'm also sure I'm not alone.

Re:Easy Answer

[cool_arrow]

maybe you could just ask the server to set a cookie on your computer ;)

Re:Easy Answer

[jddunlap]

Why don't more Linux-using shops reach out to the Linux-using community? Because the Linux-using community represents such a small percentage of their customer base that it doesn't make financial sense for them to spend the resources to specifically cater to it. That's not entirely correct. Speaking as a Linux user(desktop and server) I am sorry to say that the thing holding Linux back is the Linux community itself. The Linux community has, by and large, taken the position that any company which releases something for Linux, which is not open source, is an evil and should be boycotted. I work in the computer game industry and I actively try to support the idea of cross-platform games. However, I've been told by many people in the Linux community that if the game is not shipped with the source code that they will not buy it. This is an extremely stupid mentality. Games for Linux are one of the last things binding people to Microsoft. If you don't reward companies that make an effort to support Linux games there won't be any support for Linux games at all. For crying out loud, just look at how many people say they have Vista just because there game doesn't run anywhere else? If the Linux desktop is to succeed the Linux community must get used to the idea that there will be closed source applications written for it. Get real, the Linux Desktop is a very small market share right now. No major vendor is going to GPL their source code JUST to support Linux(especially not for a game).

Why don't company's support Linux? When the Linux community is dominated by "My way or the highway" attitudes corporations invariably choose the highway because the Linux desktop is a small market share and can be plainly ignored. Good God people, we don't need Microsoft to sink the Linux desktop. The Linux community is doing a good enough job already.

Re: Easy Answer

[mpe]

This is what the boss sees when he looks at Windows Automotive: Based on WinCE 5.0. Comes with a full set of familiar - though significantly customized - development tools, APIs and so on.
DirectSound. Direct3D Mobile.
"The Automotive User Interface Toolkit (AUITK) - a GUI framework that makes it easy to create advanced user interfaces at a desktop PC and store them as XML markup.
This separates the user interface from the functional part of the application program." The UI designer does not have to be a programmer.
What does all this tell him about development costs, schedules and staffing?

In the real world probably very little. Given that it's mostly "buzzwords".
Missing is such important information as how does this safely interoperate with the driving controls? These "advanced user interfaces" are of little use if they are mutually exclusive with several decades experience of automotive ergonomics.

Windows is approaching one billion users on the Desktop. Windows Vista has about the same market share as OSX and Linux - and got there in six months.

Vista only got that way through "Hobson's Choice" OEM deals.

Re:Easy Answer

[Anonymous Coward]

Developing for the Linux desktop in a way that requires little to no end-user tweaking has to be a nightmare. The libraries for distro XYZ are completely different than those of distro ABC, and both of those are different than my own home-grown distro I installed on my machine.

And DirectX on my XP machine is different from DirectX on my colleaues Vista machine, just like every other DLL.

The problem has been solved for years. Just include the required libraries with the program. Has been done that way on Windows since the invention of DLL hell, and on Linux at least since Loki Games. Just don't do like on Windows and drop everything into the system directories, that's the cause of DLL hell. On Linux (and AFAIK newer Windows versions), they can be conveniently placed in the application directory.

Re:Easy Answer

[rnws]

With all due respect I think you miss Stallman's point of "Free as in speech..." and what the GPL license (I assume) you are working under allows you to do.

There is no "Linux Corp.", there is the Linux Community (of which you are part). YOU (and the rest of the community) are the unit test and quality control procedure.

How does using a Mac or any other closed code GUARANTEE that they have done those things too? Because they say so? Excuse my bitter (20 year veteran) cynicism at not taking large IT companies at their word (I work for one of the biggest).

If, for example, Apple had a similar "obvious fault" in a library/driver/whatever, that you were using (under license) in your project and they didn't think it was necessary to fix - even if you contributed the code back to them (probably losing your rights to it in the process), then you're stuck at their mercy.

Yes those cost your company money to perform - but how much would it have cost if Freed Software didn't exist in the first place? Was it cheaper to find and fix that bug instead of writing the entire code-base from scratch?

Thanks to you and your team for performing the tests and procedures while taking advantage of the GPL and your community spirit in returning your fixes back - that's what it's all about!

Re:Easy Answer

[starnix]

Then use Java

Re:Easy Answer

[halber_mensch]

But why should a company support linux just because their gadget has linux running inside it? Because they are benefiting from a mature, open source, and well understood pre-established operating system. If there was no Linux they would have to spend much more development costs in building their own OS for their devices.

Or invest more in hardware and support costs and embed Windows...

And why should the blue screen of death and the business director's kludgey VB apps be relegated only to the PC world, when there are so many devices out there performing vital functions in our society that should be able to crash hard through the floor too?

Re:Easy Answer

[mosch]

Seriously.

There's a strong business case for using Linux (or other free unixes). Reasonable stability, flexibility, no per-unit licensing fees.

There's pretty much no business case for supporting Linux end users. Not only would it involve creating software for a small contingent of users, but it would do so using software that has to deal with all of Linux's weakest points (X11, GUI, etc.)

Re:Easy Answer

[illumin8]

Because they are benefiting from a mature, open source, and well understood pre-established operating system. If there was no Linux they would have to spend much more development costs in building their own OS for their devices.

What a terrible argument. Because they are saving money by benefitting from open source on their embedded platform, you expect them to waste money on developing support for open source on the desktop platform?

** May you do good and not evil.
** May you find forgiveness for yourself and forgive others.
** May you share freely, never taking more than you give.

That sounds all nice and fluffy and bunnies and hearts and flowers and stuff, but this is business. Sorry, but I do not OWE you anything. This trite communist propaganda is the reason why it has taken business so long to adopt open source. The sooner you realize that people are selfish individuals, and that capitalism works because of each individual's self interest, the better off we'll all be. You want to write open source, go ahead, knock yourself out. You'll benefit, and I'm sure many others will as well. Once you start to force people to write open source for you, against their will, you've crossed the line from freedom of choice to totalitarianism.

Re:Easy Answer

[morcego]

I have seen companies getting sued (and loosing) based on a free product they offered.
Maybe the GPL or some other OSS license will stop that from happening, but I really don't know.
This is, anyway, a valid concern for their legal department.

Re:Easy Answer

[Heddahenrik]

According to Google Analytics, Elftown has the following kind of visitors last month:

Windows: 93.75%
Mac: 4.93%
Linux: 0.71%
Unknown: 0.26%
Playstation: 0.26%
The rest: less than 0.10%

As Elftown is an artist community, there is a little bias towards Mac, but it's not big and these numbers are very close to the internationally average user. Elftown runs on Linux and is of course 100% Linux compatible. To be honest, I don't care much about (Safari or Firefox) on Mac though, but they seem to work fine except a few problems I've forgotten about now.

Re:Easy Answer

[Anonymous Coward]

>> such a small percentage of their customer base

So, which would you rather have, 1% of a $1 billion market, or 100% of a $1,000 market?

Lame reasoning. The answer is easy: the people running the place don't know about Linux and don't care about it, except when it allows them to freeload. They don't know and you don't know what the difference would be, or if it would be profitable. Especially when the Linux community actually offers to do much of the development for you.

Your thinking demonstrates that you could be an executive. You have the perfect combination of absolutely confident, condescending ignorance for the job.

Re:Easy Answer

[dupup]

Then again, I am really only familiar with server products.

Excellent point. I, too, am focused on the server side. Come to think of it, why not avoid the fuss and bother and write your desktop app in Java? I have a Rio Car: an mp3 car stereo. It has a Linux kernel and originally came with Windows-only compatible desktop software, too. Some clever wag wrote a Java desktop version, so it now runs on my Mac, my Linux box, DW's Windows machine, or my Solaris box.

There is basically no alternative to Office on Linux. OpenOffice.org? Please.

I can't agree with you here. I use OO exclusively for intercommunication with my coworkers. I find it perfectly adequate. The only shortcomings I have seen that affect me are problems with the presentation software (Impress, I think it's called). I have frequently seen presentations made with PowerPoint or StarOffice that come off differently when I look at them. The point comes across, though. I think that's the main concern; will the software allow me to communicate with my coworkers effectively, regardless of what Office software they're using? Yes. That's all I need or want from the package.

I have yet to meet a person who had any real reason for running Linux on their desktop

I have an OSX Mac as well, and really like it. I still use Linux for my primary development platform because I have access (through work) to powerful PC hardware, but none that's approved for OSX. I suspect that the answer is the same for most Linux folks. Obtain a Windows machine, replace Windows with a happier-making OS. If I could install OSX on my current box, I probably would. I just can't score the work-$$ to buy an equivalent piece of OSX hardware (2xAMD Opteron 250, 3GB RAM).

Re:Easy Answer

[Jaqenn]

If you intend to write a piece of software that doesn't work at all, then when it doesn't work it will be working. So in order for it to not work at all it will have to work. So, put the software in a box, with a 50% chance of it aborting, but you never know if it will work until you open the box.

Then, run it on a quantum computer. That should get what you're looking for.

Re:Easy Answer

[Goth Biker Babe]

Don't compare consumer/embedded devices with desktop or server OSes. The two are not the same.

In the latter the customer is the end user. Generally an individual customer has no sway since they are just one in millions. An individual user is generally powerless against Apple or Microsoft say.

In the embedded market the customer is not the end user but the developer of the device who usually pays a licence on each device sold which can amount to an awful lot of money. In that case developers can have a significant influence over the supplier of the operating system. If there's a bug you get on the phone and report the issue and pressure for a fix.

Its where you want the responsibility to lie. If you are a middleman who, for example, develops a device by subcontracting for the hardware and developing the software, and you have the hardware built by more than one supplier, it is actually more scalable and manageable to have each supplier provide a set of drivers to a defined API and you never see the source. If there's a problem you get the manufacturer to fix it. It's part of the contract for the hardware. You don't want to have engineers debugging and maintaining multiple code bases. You validate the thing and if doesn't pass you get them to fix it.

Re:Easy Answer

[Altus]

Don't get me wrong. I have no issue with people who choose not to purchase a product that doesn't meet their needs. Thats your call. My issue is with some people in the open source community (and you might not be one of these people, but they do exist) who behave as if the company owes them something. There are plenty of examples just in this thread. Because Tom Tom is using linux, there are those that think they owe the community something. Well they do, they owe the community any changes to the source code that they distribute and any bug fixes that they make. Presumably they are providing that.

If they went the extra mile and provided linux support they would not be required to do so in an open source way, but you can be sure there would be those who would complain (we share our source why don't they).

There is a strong sense of entitlement among some in this community and I don't think it is doing a service to the community as a whole. If anything it is making it less likely that companies will provide linux support. I respect your choice to avoid products that don't meet your standards, but I don't respect those (not you) who feel they are simply entitled to something from a private organization.

Re:Easy Answer

[Mr. Jaggers]

Such as? Only F/OSS examples, and only situations of full license compliance, please.

Re:Easy Answer

[thegnu]

Presuming that is an option, releasing something they can't/won't support is generally considered a Bad Thing by most people. The overwhelming public perception would probably swing negative...

whatever happened to calling something a BETA release, and noting that bug reports are appreciated? Voila!

Re:Easy Answer

[russotto]

Because I know linux quite well and will stick my arms, up to the pits, in to the code, I'll not use it as a day to day operating system. I have a Mac. Considering the relative user base size I'm not surprised that commercial companies avoid the issue of developing for Linux.

Delve into the internals of any operating system and you probably won't get a warm and fuzzy feeling. This certainly applies to Windows, but also to MacOS (classic and X) and likely everything else out there. Including the commercial RTOSs.

Re:Easy Answer

[ChronosWS]

You are expecting the company to do more than it is contractually obligated to do. And expecting that is disingenuous (or perhaps deluded) on your part. If you expected them to support you, then the GPL should also state that the software must be released with support for users of Linux comparable to that of other systems for which the device was primarily indented. Or something to that effect. You can't just write a contract and then later claim that there were additional 'moral' clauses in it to which the recipient must now adhere.

You got exactly what you should have from the GPL - the company has to release modifications. If you expected more, you should have stated as much.

Re:Easy Answer

[Corwn of Amber]

"Yeah, right."

I've been thinking those very words... Scratch that. Make it "This is the nth post that makes me think "yeah,right" and it feels like those two words are playing in a loop inside my head."

How fuckin' EASY it is to write a driver in Linux... Do you have any idea how simple it really is? I've at least read sched.c (gotta begin SOMEwhere), the driver for my SCSI card (trm395) and emu10k1.c, and written a gadget module (could read to and from some kernel space, cool). Let's see... mmap() the memory region, declare structures that hardware understands, fill interface to kernel calls, voilà - driver done.
And that's not printing drivers, those are generated by a PHP script from the specs. Anecdote : I use a Lexmark E210 and the LinuxPrinting website fucks up the PDD, so I have to select the Samsung ML-1210 instead - that must be because the spec is better written. (The printer is the exact same.)

The companies are not releasing specs because they're afraid that someone would reverse-engineer the device from the spec. It's stupid because if another company did that, they'd be spending much on that effort and they would be later to market than the first. The other thing that keeps companies from releasing specs is that they're afraid they haven't invented everything in their devices, so they might be infringing patents, and there might be some intellectual property they have licensed and may not publicly release.

Re:Easy Answer

[cthulhu11]

Cannon will not support a beta print driver? Will Crossbow or Rifle? Is it REALLY that hard to spell "Canon"???

Re:Easy Answer

[Jimithing DMB]

Then again, I am really only familiar with server products.

Excellent point. I, too, am focused on the server side. Come to think of it, why not avoid the fuss and bother and write your desktop app in Java? I have a Rio Car: an mp3 car stereo. It has a Linux kernel and originally came with Windows-only compatible desktop software, too. Some clever wag wrote a Java desktop version, so it now runs on my Mac, my Linux box, DW's Windows machine, or my Solaris box.

Perhaps because at least on Macs, Java GUIs are horrid. Actually, just about every cross-platform GUI toolkit fails to provide a respectable experience on Mac because the UI paradigms are different.

There is basically no alternative to Office on Linux. OpenOffice.org? Please.

I can't agree with you here. I use OO exclusively for intercommunication with my coworkers. I find it perfectly adequate. The only shortcomings I have seen that affect me are problems with the presentation software (Impress, I think it's called). I have frequently seen presentations made with PowerPoint or StarOffice that come off differently when I look at them. The point comes across, though. I think that's the main concern; will the software allow me to communicate with my coworkers effectively, regardless of what Office software they're using? Yes. That's all I need or want from the package.

Again, from a Mac users perspective, it's really clunky. You're right, it's perfectly adequate. Not great, just adequate. Microsoft has done a really nice version of Office for Mac and Apple has done really good Office-like programs for Mac. The screenshots I'm seeing from Office 2008 are looking really good. I still don't like the fact that Office uses Windows-style keybindings (e.g. Home goes to the beginning of line instead of Cmd+left, etc.) but it's otherwise a pretty good product. I fortunately can get by quite nicely with the much cheaper iWork so my usage of Office Mac has been limited to other people's Macs.

I have yet to meet a person who had any real reason for running Linux on their desktop

I have an OSX Mac as well, and really like it. I still use Linux for my primary development platform because I have access (through work) to powerful PC hardware, but none that's approved for OSX. I suspect that the answer is the same for most Linux folks. Obtain a Windows machine, replace Windows with a happier-making OS. If I could install OSX on my current box, I probably would. I just can't score the work-$$ to buy an equivalent piece of OSX hardware (2xAMD Opteron 250, 3GB RAM).

Yeah, money is usually the only reason. Then again, money isn't all that hard to come by. Befriend your local bean counter and it gets easier. Of course, if you're working for Sun it might be hard to justify purchasing a competitor's product. You could always, of course, hack OS X to run on a non-Apple system. It's not particularly difficult. I've gone as far as getting a base Darwin system going. To get OS X though you'd have to break the license agreement as well as the binary encryption. The latter is trivial but the former is questionable. I've always found it funny though that the license agreement says you may run it on an "Apple-labeled" computer. And the boxed OS includes two white Apple stickers. I guess it depends on how you parse "Apple-labeled."

Re:Easy Answer

[kelnos]

How fuckin' EASY it is to write a driver in Linux... You're ignoring support costs. They're not zero. You also need release management, testing/QA, etc., and people with experience in these things (which may or may not all overlap with current staff). And regardless of how easy it may be to write a driver, it requires non-zero work (money).

There's also the somewhat grey area of binary-only kernel modules. Some companies may consider it a legal minefield not worth navigating, and they may not want to GPL their driver. Most companies (esp. large companies) are very conservative where legal issues are concerned.

I'm not saying they're "right" to have this attitude: I'd love to see more vendors acknowledge Linux as a desktop OS. But these are valid concerns and issues to corporate types, for better or worse. The question boils down to: "what does supporting Linux users give us?" If the answer is only "a warm fuzzy feeling plus support headaches," what do you think they'll do?

Re:Easy Answer

[joto]

Please provide at least one verifiable example. I haven't even heard of any software company that has lost such a case---with a product they sell. Microsoft Windows offers exactly the same warranty as GNU Hello, none!

Re:Easy Answer

[morcego]

Can you show me where in my previous posts I ever saw "software companies" ?
Can you guys please look outside your little "software/OSS" windows for a moment ?

I'm just pointing to you guys some red flags that might be triggered on the legal department of those companies. Companies that work with mostly with non-software stuff, and get sued mostly by non-software stuff. Do you really think their lawyers are going to stand and say "Oh, lets forget everything we know, and all our past experiences, because this doesn't apply to software. I read on Slashdot about it!".

Better understanding a problem is the best way to try and solve it. I don't expect anyone to agree with this point of view (potential litigation). I don't. But we must understand this is the kind of thing that might be going on, and unless we do, we can't do anything about it.

Note do moderators: I like "Troll" better than "Flamebait", please.

Re:Easy Answer

[joto]

Can you show me where in my previous posts I ever saw "software companies" ?

The article is about companies releasing linux-powered devices, but the software for controlling them that can only be used from windows (or mac) but not linux. The post you were responding to, was talking about the possibility for companies to release unsupported linux drivers. Whether the companies are "software companies" or not, is irrelevant. But it should be pretty clear by now that we are talking about "software products"

I'm just pointing to you guys some red flags that might be triggered on the legal department of those companies. Companies that work with mostly with non-software stuff, and get sued mostly by non-software stuff.

The companies in question already release drivers for other operating systems. Their legal departments are therefore already familiar with the legalities of software.

Re:Easy Answer

[darkpixel2k]

The way I see it is that TomTom is saving a nice chunk of change by using OSS/GNU/Linux to build the base of their systems. It would be nice if they took a small part of those savings and just... maybe... wrote some software for OSS/GNU/Linux users.


I agree--it would be nice. But this attitude from the open source community where they get on their soap box and say "I gave it to you for free--but you didn't give me anything back." is retarded. So what if they don't give anything back. They don't have to.

I think Microsoft software sucks, so I run only linux at home. I got it for free. Sure--it'd be nice if I was able to give something back to the open source community, but it's not required.

Re:Easy Answer

[Corwn of Amber]

I was trying to say, it's so easy to write a driver or have the community write it (go on the insanelymac forum about driver development sometime, one guy pretty much alone does the intel wireless 3945 driver for OSX86) that it's stupid NOT to support "alternative OS's", at least by releasing specs.

Re:Easy Answer

[iNetRunner]

You don't have to round down. Just standard rounding works. (Though, just dropping the fraction to form the integers doesn't obviously work.)

2.49 + 2.49 = 4.98

Well, obviously anything in the range 2.25 > x < 2.5 (non-inclusive) works..

The point of the joke is of course about the internal data type vs. representation, dressed in 1 + 1 = 2 form. But you know that..

Re:Easy Answer

[iNetRunner]

Wups. Could have written that as 2.25 >= x < 2.5.

Re:Easy Answer

[JimDaGeek]

No where in my post did I say TomTom is required to give anything back. I think it would be seen as a fair exchange if the did though.

Re:Easy Answer

[similar_name]

This is really a public relations issue with the developers(OSS community) that created the OS(Linux) TomTom is using. The real downside I see for TomTom is they missed an opportunity to have an enormous enthusiastic technological group of people contribute to their product for free. Although, lucky for them, I'm sure some in the community will still hack the TomTom and create their own interface anyway. To further your cookie analogy: How will IT respond when Accounting puts in a Help Desk Ticket?

Re:Easy Answer

[renoX]

>>What's so hard about giving back to the community a tiny little something.
>It isn't that it is hard, it's just that there is no money in it. They call them for-profit corporations for a reason.

Actually, there are some benefits to supporting Linux: they save money by using Linux, so they need Linux to be successful to keep being able to use it.
As supporting Linux increase the number of the users, it ensure that Linux will still be available in the future.

But of course, Linux being already quite successful they think 'I can reduce my costs not supporting Linux, and Linux will still be available' except that if everyone do this, Linux will stall..

Re:Easy Answer

[Anonymous Coward]

getting sued (and loosing)

"losing".

Re:Easy Answer

[Anonymous Coward]

Generally, people from accounting are female, and it's probably against company policy to allow females to "share their cookies", at least on company property.

Re:Easy Answer

[Otter]

It isn't that it is hard...

Anyway, who says making and supporting a Linux client isn't hard? Just because it's easy to demand it (as if the people making the demand actually wrote the kernel themselves) doesn't mean it's "a tiny little something" when some Slackware user wants his problems fixed.

Re:Easy Answer

[kelnos]

You're still missing the point. The act of "writing the driver" is often the *smallest* cost to the company. Supporting users of the driver, release management, providing updates, security fixes, etc. -- that's where the cost is. If a company doesn't think supporting an alternative OS will generate enough extra revenue (e.g., from increased sales of their hardware) to offset these costs, they probably won't support alternative OSes.

Releasing specs is another matter. Some companies may have patent concerns, or may have licensed technology from other companies under terms that don't allow them to release specs. And sometimes it's just simple "turf protection." They're afraid -- reasonably or unreasonably -- that releasing specs might give their competitors some kind of advantage. Maybe they're right, maybe they're wrong, but that's not really up to us. They spent the time and money to develop the hardware; it's really their choice whether or not to release specs. Fortunately, for most classes of hardware, Linux users can go elsewhere when we encounter this sort of company. Even now we're making inroads into video hardware, and there's an increasing number of well-supported wireless chipsets. Most other hardware has been "open" in one form or another for some time.

Re:Easy Answer

[Corwn of Amber]

You're still missing the point. The act of "writing the driver" is often the *smallest* cost to the company. Supporting users of the driver, release management, providing updates, security fixes, etc. -- that's where the cost is. If a company doesn't think supporting an alternative OS will generate enough extra revenue (e.g., from increased sales of their hardware) to offset these costs, they probably won't support alternative OSes.

So they can release specs and the Community will take care of support and such.

Releasing specs is another matter. Some companies may have patent concerns, or may have licensed technology from other companies under terms that don't allow them to release specs.

I already adressed that. The System is broken. Abolish patents now. Not gonna happen? I'll ignore them and never sell anything save from countries that don't imagine that ideas can be monopolized.

And sometimes it's just simple "turf protection." They're afraid -- reasonably or unreasonably -- that releasing specs might give their competitors some kind of advantage. Maybe they're right, maybe they're wrong, but that's not really up to us. They spent the time and money to develop the hardware; it's really their choice whether or not to release specs.

Then why are the winmodems and winprinters and "windows-will-do-it-because-we're-cheap-assholes and will certainly never tell how we plugged the wires from PCI to phoneline! Someone else might figure it out and sell it for even cheaper!"

Those companies deserve to die an horrible economic death.

I've adressed that already, too : Reverse-engineering HW from specs is not profitable for several very good reasons. ONE : companies who do that are late to market. TWO : the label of "Follower" is Not Good in tech.

Fortunately, for most classes of hardware, Linux users can go elsewhere when we encounter this sort of company. Even now we're making inroads into video hardware, and there's an increasing number of well-supported wireless chipsets. Most other hardware has been "open" in one form or another for some time.

I need an OS that installs itself with drivers. Out of the box, no question asked. Same for codecs. Install all I'll ever need, NOW. (Automatic updates will take care of those that don't exist yet).

NO ONE needs a widget that's just a couple of wires and a Windows-only "driver" where all the functionality is. It WILL break, as soon as Microsoft introduces one more incompatibility layer between their old bad wrong broken design decisions and their new bad wrong broken design decisions, because the crappy driver will want to talk to the hardware by a channel that's FINALLY beeen abstracted... This applies to so many things I don't even know where to begin picking up examples. SB16 compatible cards that are no more SB16 compatible than mice and grape DNA is. And they need to have magic registers tweaked, or they won't be recognized as SB16. Tweaks only possible on DOS and Linux, it seems, because WinNT will NOT let you. That's ONE. (Or maybe motherboards makers only hire terminally incompetent morons for driver development...)

Re:Easy Answer

[kelnos]

I already adressed that. The System is broken. Abolish patents now. Not gonna happen? I'll ignore them and never sell anything save from countries that don't imagine that ideas can be monopolized. I don't disagree that the system is broken, but while you're doing that, the rest of us will be living here in the real world. Patents, copyrights, corporate secrecy, etc. won't go away (or be reformed) overnight, and ignoring them won't make them go away.

Re:Easy Answer

[dedalus2000]

yeah, but we don't have to like companies that take but don't give back. nor are we required to do business with them or to even speak well of them. Quite honestly the open source community owes nothing to companies like tomtom. It seems perfectly in line with community interests to trash companies who take but don't give back and to hold companies that do give back up as shining examples.

Re:Easy Answer

[darkpixel2k]

Yes.
I think everyone missed my point though.
You may not like tomtom, they may not give back, you can refuse to do business with them, or help them, or promote them, etc...

I just hate the people who whine and bitch about it like they are entitled to get something back from tomtom.
They aren't.

But even though I'm not entitled to get anything from tomtom, and they aren't giving anything back--I'm still not going to buy one.

Re:Easy Answer

[geminidomino]

If you intend to write a piece of software that doesn't work at all, then when it doesn't work it will be working. So in order for it to not work at all it will have to work. So, put the software in a box, with a 50% chance of it aborting, but you never know if it will work until you open the box.

Then, run it on a quantum computer. That should get what you're looking for. Why bother with a quantum computer, when you can run it on windows with the same result?

Re:Easy Answer

[geminidomino]

I thought the same thing... Does she really think the OSX code looks like roses and smells like baking cookies and unicorn farts?

It's about the programmers.

[PlusFiveTroll]

Because their web interface programmers are using Windows or Macs.

obviously

[Zashi]

It's the same reason they use linux in the first place that they don't support linux-desktop users.

To save money.

For most companies, linux is too small of market to be worth devoting development time to. As companies follow in IBM's and AMD's footsteps, though, I think linux support will continue to increase, but I doubt it will ever match Windows and OS X levels.

Re:obviously

[HartDev]

You would think though that after saving oh so very much buy not paying through the nose to Microsoft that you would pay some kid to type up Linux documentation! Heck I'll do it........time providing or if they paid me a decent amount.

Re:obviously

[AlexBirch]

For most companies, linux is too small of market to be worth devoting development time to. As companies follow in IBM's and AMD's footsteps, though, I think linux support will continue to increase, but I doubt it will ever match Windows and OS X levels.

My prediction is that the desktop will start to die off. It will be more hand held devices that plug into docking stations. If Apple had a great docking station for my iPhone, I wouldn't need a laptop for personal use; just for programming.

For this reason, I think Windows is going to die. It's too big and bloated. If Nokia's iPhone knockoff runs on linux, you are going to increase the number of linux users. Imagine if there is a GPhone as well.

Re:obviously

[johnsie]

Why Mac and no Linux though? Surely Macs and /Linux have a similar number of users?

Because....

[llamalad]

They have enough trouble supporting Windows users.

Imagine trying to deal with some bumbling idiot with an Ubuntu box?

And then... Which distro(s) should they support?

Re:Because....

[evanbd]

Some of us would be quite happy with "Here's the linux binary; we won't help you with it, but we'll maintain a user support forum and pay attention to bug reports."

Or, "Here's the Windows binary and source code; that should get you started. We won't help you with the Linux port, but we promise not to actively hinder it with malicious firmware updates." After all, for a company making a hardware device, the profit center is the device, not the computer-side software. Why not make it open?

Re:Because....

[Anonymous Coward]

The other problem is Linux users just plain love to bitch.

You give it to them for free, and then they demand it be free as in speech

You give them the source, and then they cry and moan that they need to be able to compile the firmware, for what reason who the fuck knows

You give them an RPM and they get in a schoolgirl huff because they want a TCL installer

You give them an X installer and they break into a full on cry because they only use KDE and they don't want to install the compatibility libraries

You give them a nice shiny graphical/TCL/everything goes installer and they bitch like an old man at a a deli because they don't want to install it as root

There's just no pleasing these people

Re:Because....

[ILongForDarkness]

I think your bang on here. Linux is just too hard, because in windows land you only have to worry about 2k, XP, and Vista, in linux land you have people with custom hacks to the kernel, not to mention the flavor of the month for installers, development environments, compilers etc. There is just too many choices to be made, and they'd rather come across as offering great support to 99% of their users, than fight to try and help the 1%, and then have the occasional screw up.

Also, Linux users IMHO tend to be more tech savy, they'd be the ones that would post on blogs everywhere, 'this company sucks they couldn't figure out how to make their code work on my custom MySql engine, I hacked directly into this random distro, what kind of losers are they?' Answer: the kind that aren't coding for a hobby but a paycheck.

Re:Because....

[Anonymous Coward]

They tend to not complain if you give them source code they can compile, and then redistribute binaries as they please.

Re:Because....

[slashdotmsiriv]

"And then... Which distro(s) should they support?"

That's easy: (K)Ubuntu

Re:Because....

[Lodragandraoidh]

This can be mitigated in several ways without having to break the business model:

1. Expose the APIs used to access the device. This way the FOSS community can build an interface that will get the job done.

2. Make the interface non-OS specific using standards. An http interface can be programmed once on the backend, and support multiple OSs via web browser (similar to how commodity IP router/switches are configured today).

These are ways of providing value add for the user, while at the same time saving your company money by only having to maintain one code base. WIN-WIN!

Re:Because....

[gardyloo]

That's easy: (K)Ubuntu I think you mean (X)Ubuntu. :)

Re:Because....

[Isaac-1]

A lot of the time it is because they have intentionaly cripled the device to not compete with a higher end product. See Cisco and their history of selling wireless routers and acess points with artificail limits on number of connections, etc.

Re:Because....

[Anonymous Coward]

Some of us would be quite happy with "Here's the linux binary; we won't help you with it, but we'll maintain a user support forum and pay attention to bug reports." Bullshit.
One of you will be happy with that.
Two will call our tech support and complain that there's no version for PPC or Sparc.
One hundred will call and complain that there's no AMD64 version and you don't want to install IA32 libs.
Two hundred will call and complain that the package system doesn't work with your distribution or that you don't know how to run install.sh.
Five hundred will call and complain that it doesn't run in VMWare (because their USB subsystem is broken).
One thousand will call and complain that it depends on a non-free library that's not included in Debian.
Two thousand will call and complain in depends on a library too new to be included in the stable Redhat.
Ten thousand will call and complain that it depends on something no longer present in the newest Ubuntu release.

Linux support will turn into a corporate fiasco until management has had enough of it and shuts it down. And then the message board bitching will begin. So that's a losing proposition.

Or, "Here's the Windows binary and source code; that should get you started. We won't help you with the Linux port, but we promise not to actively hinder it with malicious firmware updates." I don't know if you've ever seen Windows source code, but it is essentially worthless for, oh, just about anything. A much better solution is to release the interface specs. Now, the problem with that is that you'll be writing some program that does a close but imperfect job of whatever is required, and you'll expose some bug that is invariably in the system. So whatever device you had that never ever crashed when used with the well-tested Windows software will start having trouble, you'll return the product to the store because it's buggy, and then it'll be on my desk a week later and I'll have to say "someone used a third party app to do this and that and this field shouldn't be this value and that's the problem with it". Returns are the worst thing you can have in retail. They cost manufacturers money. Can you guess what the management will do when they have a thousand returned boxes sitting in the office? If you're thinking it's "let's go through another release and test cycle and put out another firmware version instead of working on our current project", you may want to think again.

Re:Because....

[JackieBrown]

Just support the distro that id. google, postal, unreal, adobe, nero, nvidia, intel, ....

Oh wait...

It looks like some software and hardware companies have figured out how to do it.

Re:Because....

[Harik]

I think your bang on here. Linux is just too hard, because in windows land you only have to worry about 2k, XP, and Vista, in linux land you have people with custom hacks to the kernel, not to mention the flavor of the month for installers, development environments, compilers etc. There is just too many choices to be made, and they'd rather come across as offering great support to 99% of their users, than fight to try and help the 1%, and then have the occasional screw up. I hear this a lot, but it's bullshit. There's no "Windows" out there, except perhaps for exactly windows XP SP2. (what, 80%+ of the market still?) 95/98/ME/2000/XP/2003/Vista, plus each service pack, what version of .net, bla bla bla. Are they on a home computer or using a managed profile? What parts of the registry can we touch, and if we can't touch locked parts on newer OSs, are the appropriate replacements usable on older ones? Where do we install? Where do we keep user data? What APIs require Administrator privledges. Do we require Administrator for everything? Do we laugh at the customers then who upgrade to vista?

Or, you know, linux. If you're LSB compliant you're probably set. Target Ubuntu for desktop apps, Redhat for enterprise. Everybody using something else is smart enough to figure it out (or read the support forums). You can assume they're not running 10 year old software because it's free to upgrade. You can require new (not bleeding edge, but new) packages because again, you're not asking someone to shell out hundreds of dollars just to run your crap.

Realistically, it's not that hard to write decent unix software. You know you can't run as root because nobody does. You know you can't update system files except via the installer, and you know they'll have libX11 installed and the API is pretty fixed. After that, you've got a subset of options (GTK or another widget set) and possibly some kernel glue (2.6+ only) but mostly you just use libUSB for consumer-grade hardware.

Your second paragraph sounds like someone who hasn't actually required technical support for a windows based product, ever. When your target market can't tell the difference between their display and their computer, (and all the assorted other horror stories, cupholders and blackouts), it doesn't take very skilled support to explain to them how to plug in the cable and click Start/My Computer/Widget Installer CD. This means that anyone who actually has a PROBLEM with it is told the three Rs of tech support: restart, reboot, reinstall. Yay, helpful.

Re:Because....

[ricegf]

That's easy: (K)Ubuntu

I think you mean (X)Ubuntu. :)

No, he means (K)(X)(Edu)Ubuntu. And Mobile Edition, too. :) :)

Re:Because....

[ILongForDarkness]

How does this benifit a device manufacturer though? They've already made the sale, now like it or not they are in the "support as cheaply as possible" part of the transaction. These aren't devices that people have paid 10k+ for and have SLA's on, just warantees. Plus, the vendor doesn't want to get into a situation where they have to mandate what you install. Eg. I work in the health industry, our vendors tell us the OS to run, right down to which patches to install. So they can know what they are getting into, and test it in their lab. Letting users use a *NIX box of their choosing is asking for headaches.

Unfortunately, they don't have to support Linux for two reasons:

1) It is not a large market

2) Most people dual boot, or have an windows box anyways, so if they don't like it they can suck it up and by the windows version. I'm supprised that their support for linux isn't just, 'if you don't like your current firmware, here is how you switch it to running windows'.

Re:Because....

[LDoggg_]

Agreed 100%

The only reason for them not release source code or specs is their own incompetence. It could be that the firmware on the device has no controls or validation for the parameters that are passed to it. It's possible that once the protocol is figured out a user may wish to send a value to the device that it can't handle.

For example, let's say a device has a little speaker that can have its default volume level in a range of 1-10 set by the windows only client application. If the application was suddenly opened, and the client side validation was removed someone might want to set the volume to 50 and blow out the speaker.
The hardware manufacturer could easily put an acceptable range validation check in the firmware, but may have been too stupid or lazy to do so. They then might say they only support their target windows because other operating systems are too small of a market which is probably bullshit. They may also say that giving out specs would give their competitors an advantage which may also be bullshit.

Re:Because....

[ILongForDarkness]

Your post highlights the difficulty in making sure that the software will work on Windows well. Why would vendors add to that Linux?

You know you can't update system files except via the installer, and you know they'll have libX11 installed and the API is pretty fixed. After that, you've got a subset of options (GTK or another widget set) and possibly some kernel glue (2.6+ only) but mostly you just use libUSB for consumer-grade hardware. Installed where, what is the installer, what shells are installed. As an aside, X11 sucks big time, yet the *NIX community continues to rely on it.

There is no way that a company selling a device for a few hundred is going to hire support staff for both windows and linux. Examples of Google, IBM, etc are lame, because they are already heavy users of LAMP so have other reasons to justify having the in house expertise.

If you're LSB compliant you're probably set. Target Ubuntu for desktop apps, Redhat for enterprise. Everybody using something else is smart enough to figure it out (or read the support forums). This thinking is what keeps Linux from going mainstream. First off, Ubuntu is based off of Debian which isn't LSB compliant.

Secondly, how many people check the compliance of their distro before they start using it? My experience is people go to a friend, the friend says something to the effect "wow Debian is so cool, I'll help you install it". Most likely the system is a dual boot, so if something doesn't work for them in Linux, no problem they just use Windows for it. Or if they really like the distro, they figure out how to make their stuff work, they aren't bothering to see if their distro, is 100% compliant with X11, installer etc.

Lastly, why does the user have to spend hours of their time to figure out how to make things work on their distro? They paid for the product, were told it would work with Linux, and now have to figure out what mods they have to make to port it to their flavor on their own?

Business Sense

[Necreia]

A business wants to make money.

Using a secure and reliable O/S that's free to run your unit/server/whatever is a great business move.

However: Most Linux users are used to 'free' software, in both cost and open sourced. Ones that are willing to pay for products will usually run dual boot with Windows or own a Mac. This being the case, it doesn't justify the resources (as a company) to create a client that must work on all or select distros and/or make the source code public.

I would love Linux to get more desktop applications. I can also see companies perspective as-to why it isn't going to happen strongly anytime soon.

Re:Business Sense

[James Anglin]

Exactly: not paying is one thing; seeking customers who don't pay is another.

Re:Business Sense

[Dunbal]

Exactly: not paying is one thing; seeking customers who don't pay is another.

      Excuse me, but if your business consists of selling hardware (and not the app that lets you hook it up to a PC), how exactly do you expect linux users to "not pay"?

      Release the technical specs for your widget and the community will do the rest. However if you feel you can live without that extra 5% of the market, well, fair enough. Some companies would kill for 5%.

Re:Business Sense

[JackieBrown]

I meant to mark insightful and my mouse slid on redundant.

Posting to undo moderation.

Re:Business Sense

[the_arrow]

A business wants to make money.

Ah yes, but the business in question don't sell PC software, they sell hardware. They don't make any money on the software that talks to the device, they make money on the actual device.
Besides, if they make a Linux version of the PC software, wouldn't lots of geeks buy the device and recomend it to their (non-linux) friends?

TFA answers its own question

[ACS Solver]

Like the TFA says, they save millions by using free software. Showing that your hardware is stable also brings you extra cash. Recruiting extra specialists and devoting extra resources to help what's a tiny part of your user base is not financially profitable, so they don't.

Sometimes things are that simple.

Re:TFA answers its own question

[Tackhead]

> Like the TFA says, they save millions by using free software. Showing that your hardware is stable also brings you extra cash. Recruiting extra specialists and devoting extra resources to help what's a tiny part of your user base is not financially profitable, so they don't.

Speaking of extra resources and budgetary allocation...

Embedded system: Free.
Linux port of client software: Expensive.
Sufficient web server CPU and bandwidth allocation: Priceless.

Re:TFA answers its own question

[Neo_piper]

Wait they support MAC?!? When did that sort of thing become unremarkable in it's self.

Re:TFA answers its own question

[Anonymous Coward]

Right around the time that spelling and grammar checkers were expected to be used.

LOWRANCE

[Anonymous Coward]

use the Lowrance GPS unit for your car. They support Linux users.

Re:LOWRANCE

[russ1337]

That looks quite good. My biggest complaint with in car GPS's (having owned an Navman iCN-510 for three years), is the inability to update the maps easily or cheaply.

My next purchase is the Nokia n800 with a bluetooth GPS. While it has some limitations, it can cache the maps from Google, for an area you specify, and as detailed as your memory card will allow.

I'm also intersted in the openmoko, and understand the retail kit will include a GPS. I believe it runs Linux (or other open source), and as long as they come up with a similar cache method, then great.

it's all about the money

[Adult film producer]

how many companies listed on the exchange even give dividends anymore? Not many.. stick that junk in your mutual fund and you've become a speculator, not an investor. The money goes to the top, ceos deserve big 8 figure bonuses.. why bother supporting a bunch of greasy haired hippies? That makes no sense from a ceo's perspective.

Re:it's all about the money

[Anonymous Coward]

Look at "canroy" -- Canadian Royalty Trusts, eg, HTE, PGH, CNE, AAV. Oil & natural gas, dividends paid monthly in Canadian dollars (woo hoo). There are downsides -- the Canadian gov't raped these companies last year by breaking its promise to not change the tax structure on trusts. In 2011 the tax rules will change (taxes will go up), but many companies have built up a lot of tax credits to set that back a number of years. What the Canadian gov't really did though, is set a firesale price on these trusts. So Dubai just bought up PWI. Essentially, Canadian gov't greed and duplicity will result in the sale of their natural resources to foreign concerns. Because all the trusts pay high distributions (double digits on an anual percentage basis), and are likely to get bought out, you get the potential to speculate and take payments along the way. Of course, the Canadians could really rape these trusts by forbidding sale to foreigners AND raising taxes, so the distributions come at high risk. But if they did that, development would stop, people would lose their jobs, and then there might be pressure to change. Anyway, plenty of risks but the distributions really do tempt.

Because..

[Chineseyes]

Because their job is to make money not support linux users. If you want to see a business that supports linux users start one.

Re:Because..

[Anonymous Coward]

"If you want to see a business that supports linux users start one and watch it fail."

there fixed

Re:Because..

[Anonymous Coward]

I did. Guess what? I'm making money. Wanna know why? Because no one else in my area supports Linux users. Who gives a fuck if it's only 3% of the computer-using public or (pull out some arbitrary number here) if every single one of them comes to me?

Let's play with numbers. According to these guys (I don't know who the hell they are either, it's just what Google spat out), there were 223 million PCs in the United States at the end of 2004. Let's use the commonly accepted 3% number. That adds up to roughly 6.6 million Linux PCs in America. Let's assume every one of those cost $500 (another completely made-up number). That gives us a figure of 3.3 billion-with-a-B dollars in Linux PCs in America. Three years ago, given these admittedly arbitrary numbers that I pulled out of my ass, but I don't see why they'd be too far off the mark. Three billion dollars. That's completely discounting the "family geek" phenomenon described elsewhere in this discussion. Now, I'm just a local geek for hire, and my share of that pie has been pretty good. As a national/international concern, you ignore that number at your peril.

for the same reason

[darth_linux]

Novell can buy a Linux, partner, with M$, and be a "Linux company". It's good to use Linux to our advantage and bad to support Linux when it doesn't. (Mod -15 karma. you know you want to)

It's GPL, right?

[PhxBlue]

So can you piggyback on their software to develop a Linux solution? I'm asking because I don't know enough about the GPL to know how much of the source gets distributed under a GPL (i.e., all of it or just what parts of it were GPL'ed before they wrote the software).

Re:It's GPL, right?

[Anonymous Coward]

So can you piggyback on their software to develop a Linux solution?

No.

All the "good parts" - the mapping, navigation, desktop-updater-interface, even the UI - are all in this one, gigantic, binary-only application that you promise not to disassemble in the shrink-wrap EULA.

Tom-Tom leeches the Linux kernel and gives back exactly sweet fsck all. Not code, not cash, just SFA.

Re:It's GPL, right?

[Dan Ost]

And that's just fine. Using Linux without giving back, as long as you're not doing something that directly damages the community, strengthens the community. It raises awareness which is what we need the most of right now.

Re:It's GPL, right?

[Hatta]

Nope. The GPL only covers derivative works. Works that include the actual GPL code in them. Just because a program was built with GCC and runs on linux doesn't make it a derivative work of either. It has to incorporate the actual GPL code.

If they are distributing a device that runs the Linux kernel, they are obligated to proved the linux source. If they had to write any drivers, those have to link with the kernel to run so you can probably get the code for them too. Though sometimes vendors (like nVidia) use silly tricks like an open source adaptor to a binary blob, so even that's not certain.

Cost

[Reality Master 201]

Linux desktop users are a small segment of the market. Developing tools costs money, and there needs to be a large enough payoff for the development costs to make it worthwhile.

And some things about development of commercial apps for Linux are bit of a pain. What widget set do you use? How do you determine if the appropriate libraries are installled, where does the OS mount devices, what device numbers do you get, etc. Nothing insurmountable, just more complexity than with Windows or OSX.

Re:Cost

[Anonymous Coward]

This is assuming that not everyone who thinks about developing for linux is as stupid as you are.
Lets take the top 3 Distributions and see if we can answer these asinine questions.

What widget set do you use? Most @POPULAR_DISTROS ship with gnome so I guess gtk is a safe bet.
A lot also ship with KDE so qt is also another safe bet. Pick one.
If users don't have these installed their distribution will handle
installing it for them as the popular distributions handle that.

How do you determine if the appropriate libraries are installled, See my point above. If Version x of distro y uses library version z then I guess you're going to be using
library = z.

I mean it looks like Nero for Linux figured out how to release a sensible product for the masses in Linux.
They chose the following.

Package: Red Hat & Debian based. These will ensure that it will run on a popular distribution
Kernel: 2.4 but 2.6 recommended
Widget: Gtk 2.x, not too hard

Wow look at all those hard choices they had to make. Man I can see their development house going up in flames because of the 10 million different packages they release.
Also somehow they seem to be able to detect the CD burners on the system. ZOMGPONIESMAGIC.

Stop spewing Fud.

Why? Here's why.

[fyngyrz]

Why don't more Linux-using shops reach out to the Linux-using community?"

Because linux users, as a general rule, have a strong aversion to paying for a commercial product. They're used to free software, and free software, service models excepted, is a very poor model for a company to earn with.

Service models won't do for consumer products, either. They have to work, they have to be intuitive, etc. The optimum consumer product (like the GPS in my car, now that I think about it) has to "just work."

Depending on the product, there may be some issues with licenses as well, if the company wishes to keep the product a closed-source project (quite likely.) The GPL in particular can be a problem in this regard, and even the LGPL can be a problem (see section 4d, which specifies that either source code sufficient to recompile and relink, or a shared library already present on the user's computer must be used.)

It really is too bad that the barrier for entry for commercial entitles is so rough; I got a good look at Ubuntu the other day, and I was downright astonished. That's an OS ready for my grandmother. Finally.

Re:Why? Here's why.

[Cal Paterson]

Because linux users, as a general rule, have a strong aversion to paying for a commercial product. They're used to free software, and free software, service models excepted, is a very poor model for a company to earn with.

This is nonsensical crap. Everyone pays for hardware. Tom Tom is a hardware company.

Re:Why? Here's why.

[tepples]

free software, service models excepted, is a very poor model for a company to earn with. This is nonsensical crap. Everyone pays for hardware. Tom Tom is a hardware company. Unless the hardware is a common embedded FPGA/microcontroller reference design that happens to have proprietary firmware loaded onto it.

Re:Why? Here's why.

[fyngyrz]

This is nonsensical crap.

No, actually it isn't. Most GPS units have considerable software support at the PC level. Mine, for instance (a Garmin), has a complete out-of-unit mapping system (for Windows) with the ability to tag places, make custom icons, route out-of-car and plan, plus optional maps and some other features.

Creating these tools for each OS involves considerable additional investment, and each tool or feature that is a cost+ add-on brings the issue of who will or will not pay more and more to the forefront. As GPS units subsume more and more features - MP3 playing, DVD playing, NWS, XM, traffic, and other radio services, folding your images into the maps and so on, the investment in PC level software can increase dramatically.

Linux simply isn't fertile ground for this for all the reasons I mentioned, plus a few others here and there (such as the royal pain in the neck of trying to pick a widget set with the least annoying license, best feature set, lowest extra cost, and broad distribution - as compared to the standard GUI tools available on OS X and Windows.)

So while I decline to be as rude as you were with your "nonsensical crap" kind of remark, the fact remains, you're simply wrong.

Re:Why? Here's why.

[npsimons]

Because linux users, as a general rule, have a strong aversion to paying for a commercial product.

I keep hearing this, and more and more, I suspect it's not true, especially as corps pick up Linux and want enterprise products on their servers. I know that anecdotes don't count for crap, but I have paid plenty of money for software (even closed source!) that runs on Linux. Granted, most of that has been for games, but it's still a valid point. Besides, we're talking about a physical product here that is sold (the in car GPS device) with no relation to being able to interface it to a computer; then the company tells us you can connect it to your computer to do neat things with it. Great! Where's the free Linux client that should be available beside the free Mac and Windows client? Where does not wanting to pay money factor into the free-software-to-support-a-product-I-already-paid-for business model?

On a related note, I only use Linux at home; it covers everything I need (accounting, gaming, programming, photo management, etc). The only reasons I have Mac and Windows boxes is for my users (as both a sysadmin with users who won't use anything else, and a programmer who develops for such users). If your company doesn't offer a Linux version or doesn't have hardware that works with Linux, great, that's one less choice I have to consider. And for those who think I am limiting my choices, saying that I'm ignoring "higher quality" choices that otherwise work on other platforms, I have generally found that it's the exact opposite: if they don't support Linux, their product is usually inferior to those who do.

Re:Why? Here's why.

[jedidiah]

Bullocks.

They could use QT and immediately have all of their bases covered.

Since they are using this to "make money" they don't have to act
like they are some shareware developer just scraping by. They can
buy tools to avoid the sort of labor costs some people seem to be
whining about.

Re:Why? Here's why.

[everphilski]

Qt is expensive, and is licensed per developer, not per project. .NET is (essentially) free, and in many ways a lot better ... (I've used both)

Re:Why? Here's why.

[fyngyrz]

QT is commercial, extra cost. Which case I specifically addressed. So the only "bollocks" in sight are yours. There are days when it seems like the lower the slashdot ID, the more likely the poster will be rude, uninformed, and just plain wrongheaded. Thanks for making this one of those days.

Re:Why? Here's why.

[PitaBred]

Then they better not use open-source based software in the firmware program. You can't have your cake and eat it too. You get Linux for free, you need to abide by the copyright agreement.

Re:Why? Here's why.

[Cal Paterson]

QT won't cost anything, because Tom Tom is (apparently) Free Software anyway.

Re:Why? Here's why.

[jedidiah]

That criteria alone is hardly compelling. Commercial software that businesses buy in the course of doing what they do tend to cost a considerable amount of money. So do a lot of other things that work out as a per employee cost.

So the fact that you personally might get sticker shock from something (like Photoshop, Cubase or AutoCADD) is not a compelling argument.

Much is made about the "extra cost" of supporting "marginal platforms". Since MacOS is in the mix there's an obvious potential value to having a cross platform solution. If supporting Linux poses a problem then so does supporting the Macs.

Even the current Troll prices don't work out to very many billed developer hours.

Either the cost savings of avoiding the 2nd platform dev costs are worth the Troll licenses, or the cost of the 2nd/3rd platform being developed aren't all that bad afterall.

If the Mac version is outsourced, the Troll licenses might even be cheaper/better than dealing with the management
of the external dev team.

Re:Why? Here's why.

[Risen888]

You are mistaken. QT is dual licensed. You as the developer may choose to use it subject to the GPL or pay Trolltech money. Since Tom Tom is already releasing their software under the GPL, it seems like a pretty easy step.

Re:Why? Here's why.

[fyngyrz]

So the fact that you personally might get sticker shock from something (like Photoshop, Cubase or AutoCADD) is not a compelling argument.

First of all, Photoshop is a high level application. That has no bearing on what a *developer* might consider unreasonable as a development cost going into their own application. Secondly, I am a developer, I am responsible for an application of Photoshop's approximate class, we're completely debt-free and cashy, and I still wouldn't consider licensing GUI widgets. As far as I am concerned, the day linux gets GUI widgets that are always there and available on the same terms as those in OS X and Windows is the day I'm willing to release a port to the platform. Other people may have other opinions, and I'm not saying they aren't valid, but that's mine. Either the OS provides the GUI, or as far as I'm concerned, there is no GUI. And incorporating anything using the GPL... not a chance on this earth. But we do have a working linux port ready for the eventuality that the OS changes to provide a standard GUI. There are projects running to get that done, thank goodness. All we'll have to do is move the widgetry over and we should be good to go.

Since MacOS is in the mix there's an obvious potential value to having a cross platform solution. If supporting Linux poses a problem then so does supporting the Macs

Not so. Mac users pay, and pay well, and in large numbers, for good applications. If your app isn't a support problem, every sale is a profitable sale, and the Mac OS, being extremely stable and reliable (just like linux, I might add), is a wonderful platform for selling software into. The linux market isn't even remotely comparable. The GPL is the perfect example of the linux attitude towards commercial software - and it is not commercial friendly. Selling support doesn't work either unless your app is so unfriendly people require help to use it, or else if it is buggy, or has compatibility problems. Applications that "just work", which is our actual goal, have to be sold on initial perceived value, actual value in use, and perceived value of upgrades. Selling someone a "service contract" you know they'll never have to use isn't a very ethical thing to do either.

Even the current Troll prices don't work out to very many billed developer hours.

Oh. I get it. You think the cost of going with a third party widget set is the initial monetary outlay. Well, that's certainly part of it, but what happens when trolltech goes out of business, and linux just keeps evolving? Or the opposite - when Trolltech decides that they're not going to support an older linux, but we want to support our customers? Why should we risk tying our application to a third party? With a better OS design - meaning, one that actually has its own GUI - you can be pretty certain that your stuff is going to continue working. Windows 95 software still works and its been 12 years. Trolltech would never do this, I hear you say? Whoops, wrong. They already have. I can't compile or run the current Gimp on a stock RH9 system, not all that old, frankly. If it isn't complaining about the font libraries or the version of the C compiler, it's having a meltdown over some obscure library I've never even heard of. When I spoke up about this, I was told, "update the linux system"; but that's precisely the wrong answer. A commercial app needs to work on the widest possible number of systems, not only the latest and greatest. At least, as far as I'm concerned. I admit I've run into developers who grab at new OS features like chimps after bright yellow bananas, but we're not one of those. Our objective is to get the app working, and keep the app working. If something shows up we want from a later OS, and we can't special case it in and out based on OS level detection, we just won't use it. Because to lock out our users with old OS's is unaccept

Re:Why? Here's why.

[Air-conditioned cowh]

They could use the GPL version of QT. That's what I'm doing at work (with pyqt and eric IDE) to make a cross platform app to work with an embedded device. If they use GPL software in the camera then they have to make the source available. Ditto with the desktop app. No big deal for a hardware manufacturer.

Re:Why? Here's why.

[everphilski]

Right but then you have to open source your project (gotta love that viral GPL) and now your interfaces are wide open ... again, Tom Tom is more than a piece of hardware and I gotta believe they have a vested interest in keeping their software closed (read some of the other threads, other people have said points I won't bother to repeat)

Re:Why? Here's why.

[Anonymous Coward]

Garmin's newest Nuvi, which should be out before Christmas, is Debian Linux based. Don't give me the extra stuff, just give me the ability to upgrade the device on linux (through a binary blob driver), or possibly some specifications I can use to write a driver and application, or better yet, just give me what the developers have:

The firmware on this Nuvi is upgraded using a USB-Ethernet interface through apt-get and a .deb repository maintained by the developers while the software is being developed.

None of that will see the light of day outside Garmin's development offices. There will be some development pack for 3rd party games, but that and firmware upgrading will be windows-only. The thing uses gtkmm, as well as mobile nvidia-accelerated graphics. They are using the Debian/GPL/Linux codebase, which I'm sure is creating quite a bit of value for them, yet they somehow see more value in locking the thing up and losing customers who might find better uses for it? Let there be binary drivers and proprietary applications, they're not losing anything the way they hand out maps (each map can only work with the one device it was activated for, and without which the GPS loses quite a bit of it's functionality). Just let me put other things I want on it. If you must, make the hardware "TPM" application specific, and not OS specific. It's quite possible with the newest virtualization techniques... and you won't lose any security. Granted, somebody would have to license vPro from Intel and adapt it for an ARM processor, if somebody at ARM isn't working on something similar already.

Re:Why? Here's why.

[geminidomino]

Out of curiosity, what do you use for accounting and photo management?

They saved 'Thousands' not millions

[warren_spencer_1977]

Compared to developing their own kernel for their product, sure, perhaps they saved millions. Maybe tens of millions.

But products aren't designed like that. We buy commercial kernels like pSOS or QNX or (lately), use Linux for free. We pay royalties based on units shipped to the kernel provider, which amounts to cents or dollars per unit. This is much cheaper than the Care And Feeding of kernel developers.

Re:They saved 'Thousands' not millions

[Anonymous Coward]

True; the summary is grossly exaggerated.

It's also possible to find royalty-free kernels commercially. Binary licenses for such are typically available for ten or twenty thousand dollars, no per-instance fee at all. Source code licenses can break into six figures, but not seven.

There are also a number of free kernels that happen not to be Linux.

Even developing your own kernel won't cost millions. Linux is, frankly, a bloated pig compared to what most embedded products really need. You don't need seven forms of IPC with three different messages interfaces, some of which work in signal handlers and some which don't, some of which work in device interrupt handlers, some of which don't, on top of three threading models and four schedulers. The question isn't how much it would cost to re-implement Linux and its Unix history line-by-line, but rather how much to implement what you really need. The answer there is more on the order of six person-months; call it $100K.

(And yes, I've implemented such a kernel before. Three times, actually.)

Re:They saved 'Thousands' not millions

[warren_spencer_1977]

Good points.

Three times huh? No wonder your estimate is 6 months. Mine is more like 18 lol!

Market considerations

[l33t-gu3lph1t3]

Submitter's logic is fuzzy. Tomtom runs on linux because Linux is a good candidate for an embedded operating system. From a technical and business standpoint, it makes sense to use linux here: no license fees to a proprietary vendor, greater control over the OS, etc. From a business standpoint, supporting Windows clients makes sense as well. It's a question of numbers: There are more Windows desktop users than Linux desktop users. The right tool for the right job. Making your own standardized device run on Linux is a lot easier than making software that supports an entire ecosystem of OSes.

Re:Market considerations

[Anonymous Coward]

No, your comprehension is fuzzy. All the submitter is asking for is the devices that run Linux kernels offer the noddy tools to update their firmware and perform the basic crap when setting the device up. The companies in question can get around this my merely releasing these trivial applications as open source and let the Linux community create their own versions. Even less effort is to just publish the specs on how to talk to the device.

Considering the vast sums of money they save by using the Linux community's efforts, it's not much to ask in return.

Re:Market considerations

[l33t-gu3lph1t3]

Where's the ROI in this case? "It's not much to ask" sounds like you're asking for a favor. What possible incentive do they have to spend resources to do you a favor?

Re:Market considerations

[Hatta]

Releasing the specs to the protocol that the windows client uses to speak to the linux app is free. We'll do the rest.

Re:Market considerations

[anagama]

The ROI is getting geek mindshare. As I mentioned before here, I've been the key factor in the computer purchases several people I know made -- five computers in 24 months. None of these people use linux, but they still ask me what I think. That amplifies the value of my opinions five fold.

Re:Market considerations

[Grail]

Never forgetting that if someone is running Windows XP, you have two options to support - XP Home or XP Professional. If someone's running Mac OS X, there is one option to support. If someone is running Linux... was it a home brew? Ubuntu (which release)? Debian (sid? sarge? etch?) are they using X windows or just text mode? Gnome, KDE, Athena?

So suddenly that 3% of the market turns out to be highly fragmented. How do you write a support script for your level 1 tech support people?

The best you can do is document the API and electrical interface, and let the community write their own UI.

answer

[Lord Ender]

They think supporting linux desktops is too expensive to be profitable.

End of discussion.

Next question!

Not surprising

[Goofy73]

While I too find that a bit disturbing, it doesn't surprise me.

1) There isn't enough people using linux to really hurt them dollar wise by not supporting it.

2) They probably saved a lot of money by not licensing an os or trying to develop one on their own.

3) There is nothing really preventing them from doing so as long as they abide by the GPL etc...

As I said, I'm not saying it's right but it is what I would expect at this point.

Re:Not surprising

[PitaBred]

1) Not directly, no, but my opinion as a Linux user affects the tech purchases of at least 10 people very strongly, and the purchasing decisions of my company at least moderately

2) Definitely. Good business sense.

3) Nope. But it's not good business to alienate the people who support your operations. Goodwill may not go on a line in your ledger, but it's a quite real value (or cost, should you misuse it).

You expect it at this point because execs were educated in the 80's, where the bottom line is the ONLY line that matters. But that's not really the best way to approach things, because you lose loyalty when you show none, which will screw you over whenever you make a single misstep, rather than people being forgiving of a faux pas (such as with companies like Google) simply because the company shows so much goodwill towards their customers.

Support costs

[ardor]

Percentage is one reason, but a big one (if not the main one) is that supporting Linux is expensive. Technically, the platform isn't that hard to support, but there are *many* distros out there.
Many companies are waiting for one or a couple of distros to become dominant enough. Ubuntu, Fedora, Suse, Debian, Redhat are potential candidates. It is also very important to rule out setups with modified kernel, since they can have unpredicatable side-effects. In fact, this rule extends to all software packages which are about to be supported. THEN commercial support is possible. There are too many variables otherwise, and those poor helpdesk guys have to KNOW all known issues.

Windows XP has several versions (Home, Professional...) but the differences are often subtle enough to make it easy to treat them as one. Vista, I don't know. OSX is a no-brainer, its a totally locked platform, support is quite cheap.

Maybe an Ubuntu with support for vendor repositories (for patches) would be an option. But right now, Linux support is just too expensive. Quite ironic, actually: OSX is the most locked down of all those platforms, and the cheapest to support. With Linux its the other way round.

What about server/client discrimination

[7-Vodka]

Don't forget the numerous companies who release linux server versions of their application and completely ignore linux when it comes to releasing a client. It irks. I want to use the software or play the game myself, not host lame windows clients so they can play on my server.

Also, companies which promise a linux client is "coming soon!" and then years later still haven't delivered a damn thing. (I'm looking at you ventrilo on both counts).

Re:What about server/client discrimination

[L0rdJedi]

Why bother with a client when it seems to work under WINE.

http://gentoo-wiki.com/HOWTO_Ventrilo_Via_Wine

Re:What about server/client discrimination

[Hatta]

Also, companies which promise a linux client is "coming soon!" and then years later still haven't delivered a damn thing. (I'm looking at you ventrilo on both counts).

Have you tried Mumble/Murmur?

Re:What about server/client discrimination

[Indian]

The biggest offender that fits the bill on this: Google.
Where is a voice+video capable native google talk client for linux?
Ditto for google earth, picasa, ........

Re:What about server/client discrimination

[gronofer]

Ditto for google earth, picasa, ........

These exist for Linux, but are somewhat hidden. I'm not sure about voice+video google talk, never used it.

Picasa for Linux. Google Earth for Linux (the instructions for the latter seem to be out of date: the .bin file that downloads when you visit the download page with Linux seems to be the one you want.)

You think it's bad to be a user?

[Anonymous Coward]

How about being a developer at a shop like that, to have a corporate requirement to use the services, and to have to respond "I'll use our services as soon as our client runs on the operating system I use" during all-company meetings.

Not that I would have ever done anything like that. No, sir!

In any case, the usual justification is that the linux user community is too small to be worth the effort required to support it. Fortunately, for the company I work for, there is a serviceable, if somewhat clunky, Web-based interface that is browser-agnostic, so I can use our services no matter what OS I choose to run. There are also movements afoot to expose the services we provide through open protocols like WebDAV and LDAP, which would, in the fullness of time, allow Linux users the ability to develop their own clients.

One guess...

[glindsey]

One guess is that the TomTom firmware was developed by their embedded engineering team (or outsourced), while their drivers and applications are developed by their (non-embedded) programming team. This is not uncommon; at the place I work we often design and/or develop the firmware for a company, while the company develops supporting applications in-house.

Them's fightin' words - Sis, send the kids out

[Anonymous Coward]

Them's fightin' words - Sis, send the kids out 'cause we commencin' to duel

Why not leverage Open Source?

[zonex]

There are several comments about the cost of developing and supporting a Linux version of a company's software for a (relatively) small user base.

If that is indeed the case then why don't these companies use the innate strength of the Open Source community and have them contribute in creating this software? All they would need is to create an API that can be used by the OS community to create the appropriate front-ends.

Incidentally I am sure it would be a trivial task to port the same API to Mac OS X thereby reducing the cost of supporting that platform as an added benefit.

Re:Why not leverage Open Source?

[czmax]

My guess is 'differentiation'.

If there was a standard (opensourced) API for communicating with WidgetX -- and some form of opensource WidgetControler application was available -- then a competitor can step in with WidgetY and offer much the same user experience (it's even running the same software)!

The status quo lets the WidgetXControler application be part of the offering. The box can read: "Includes the award wining WidgetXControler software that only we provide!".

Duh, I wonder why...

[cjonslashdot]

All of the comments here are true and right on the mark. I would add that the root cause is that Linux is too hard to get onto your computer, if you are a mainstream user. Linux will not be widely adopted until either: 1. Retailers order it pre-installed; or 2. It can be installed with the same ease as installing an application. Forget ISO files - most people don't even know what those are. And if there is not built-in driver support for the range of widescreen displays, bluetooth, and wireless hardware, then forget it. In that case, it will stay an early adopter niche, and will not be supported by most vendors.

Re:Duh, I wonder why...

[Anonymous Coward]

With the EU poised to ban shipping Microsoft Windows preinstalled and bundled with PCs (suck on that, Microsoft "Naked PC" FUD campaign!), that situation might change rapidly in the world's largest economy (=> europe, not america)

Maybe their software is written by different group

[whois]

I just recently got a tomtom, and I'm very impressed with the actual unit. It's highly functional and useful and the menu system (while sometimes cumbersome) is generally really good. The only thing I find its missing is trip time and average speed (something I have on another portable GPS, so I don't miss it much)

The Windows software though, blows.

It's super super super awful. Some of the things they could've done (contact list synchronization, or just a csv based import of addresses for favorites aren't possible. About the only thing you can do is download new maps to it and update the software. Granted, I only used the software for 5 minutes to make sure I was running the latest version, but thats my impressions of it. If you truly want to, you can control your tomtom through your computer (via an onscreen representation of the tomtom screen, not an API).. it might be possible to use the tomtom as a GPS for one of the windows mapping programs but I dunno.

Anyway, the software sucks.

There are a lot of factors involved here...

[PFI_Optix]

I'd have to say the biggest barrier (aside from the relatively tiny potential market) is the lack of standardization in Linux. Dozens of distros with multiple shells and several desktop environments and a lack of a unified standard on libraries and...well, you get the point. It all adds up to a support nightmare with Linux User #32,469 calls because his customized DSLinux USB key won't properly sync with their device.

With Windows, you can specify "requires Windows XP with SP2 and .Net Framework 3.0". But if you specified a handful of Linux distros and library sets and everything else necessary to ensure it can be supported, you'd only be getting a fraction of the Linux market, which is but a fraction of the PC user market.

The most I could ask of any company in the way of Linux support is a solid driver with good documentation, a wiki to allow the Linux community to fill in the blanks when unexpected problems crop up, and a web forum to facilitate the community and allow developer to monitor/communicate with the users.

Re:There are a lot of factors involved here...

[legirons]

"I'd have to say the biggest barrier (aside from the relatively tiny potential market) is the lack of standardization in Linux. Dozens of distros with multiple shells and several desktop environments"

Weird how all the programs I use "just happened to work" in about 10 different versions of about 5 different Linux distributions.

Actually no, it's not weird at all, since they all run the same software with interfaces defined by UNIX and RFC standards, and with separation between software and GUIs and interfaces and file formats, and each one only changing where required.

"With Windows, you can specify "requires Windows XP with SP2 and .Net Framework 3.0"."

I think you answered your own question about which OS is the least stable between versions...

Re:There are a lot of factors involved here...

[walt-sjc]

You know, it's amazing just how wrong that is. Look at nvidia, where their driver works on just about every Intel based linux box, or Acrobat Reader, firefox, open office, Java, etc. Even pretty invasive software like VMWare works just about anywhere.

Figure out how to link statically and a single binary works fine.

It's fine to limit "official" support to a subset like Suse, RedHat, and Ubuntu, but it's fairly trivial to be able to run on damn near any flavor of Linux.

also not just numbers but money

[hesaigo999ca]

Aside from the fact that linux community is smaller in comparison to windows....it is cheaper to not support the linux community, where as someone can tell you....you have a dell or winodws, then call them for support. If you have linux, everyone knows there is no one to call except for those of us that bought a package to allow for support...which is extra...

As far as I know, I can only see this being a real money thing, everything is easily developed for windows(.net framework etc...) I can throw an app together in less then 1 hour, and publish it, whereas, not sure but i think for anything developed linux, there seems to be alot of hoops to jump through (ie- setting up distros, and patches, needing to recompile the kernel etc...)

"The best way to predict the future is to invent it"

I bet they do, They just don't test it.

[pUNX.h]

I bet they do have linux versions. They just don't go through the QA process. As stated before, Linux users are a small percentage of the user base. Cutting QA costs increases the profit.

Just my two cents.

Another easy answer - support staff

[shawnmchorse]

At my workplace, we have client software used by our players that is Java based. All of our developers use Linux, so it is well tested on that platform. However our two supported platforms are Windows and OS X. Why? Those are the platforms that our support staff actually know something about. And even then, their OS X knowledge is pretty marginal. We do however provide a "use at your own risk" type installer for Linux, though the link to it is buried on our forums. It's kind of a "if it works for you then great, if not then tough luck" type of deal.

Which linux?

[Anonymous Coward]

Supporting "Linux" is a difficult proposition when you aren't a huge company. There are so many variations of linux and variations of configurations that it is very difficult to provide support. In order to support a distribution you have to test on it, have a machine available with that version to perform support, etc.
If you are trying to support hardware it gets even worse.

Finally, if you are dealing with an open source distribution that does not provide commercial support you don't have anywhere concrete to go to if you have a problem. If the community doesn't want' to support the change or fix you need or doesn't consider it a priority you have to locally branch that distribution (and so does your customer) and then support that branch.

So it's a pain in the ass. That's why.

Re:Which linux?

[Dunbal]

There are so many variations of linux and variations of configurations that it is very difficult to provide support.

      Not really, see, because if you build your app for a very popular linux distro and release the source code, the community will do the rest of the porting for you.

      But once again we see how wanting to keep things secret and hush hush this is proprietary stuff just slows down progress.

Re:Which linux?

[Lurks]

But once again we see how wanting to keep things secret and hush hush this is proprietary stuff just slows down progress.

No. We don't see that at all.

The application in question, which clearly almost none of the Slashdot posters here have even the slightest clue about given the repeated mantras of Linux evangelism, isn't a required aspect to support the device at all. You plug it into USB. It appears as a mass storage device. The large buttons on screen guide extremely inexpert users as to how to obtain various bits of third-party add ons and copy them to the right places on the device.

If you're a Linux user, you almost certainly would prefer NOT to run this kind of application at all. You'd simply read the readme, figure out where to copy the files and be happy. So it's not even that they didn't bother to make a Linux version, it would be absurd to even bother. And if you really DID that, you could make a duplicate application in a heart beat because all the damn thing does is copy files to a USB mass storage device. Or do you consider that proprietary?

it's about their target audience

[Locutus]

Having Linux on the device saves them tons on support by using a reliable system, saves them tons in licensing fees by using GPL'ed software, and saves them tons on development time by leveraging many API's available and again, due to the GPL.

When it comes to providing software for users to load to interface a computer with that device, most are still using Microsoft Windows and far far fewer using Mac. IMO, the Mac gets support because it has a long history in the industry and not supporting it pisses of some vocal users( media, etc ).

With this in mind, do you now understand why Microsoft went all out to destroy the C++ frameworks businesses in the 90s? Why they have done the same when any cross platform development tool gains acceptance in the community? If they were using Qt for their desktop app development then it would be one thing but IIRC, Qt 3.0(2001) was the first time it supported Mac and so many companies were/are still tied to other development platforms. Ones which don't easily port to Linux.

BTW, this was the same thing happening when Sharp release the Linux based Zaurus but it was worst there. Sharp wanted developers to help with application and the dev env was Linux but the QtopiaDesktop PIM/syncing application was only for Windows. How stupid is that? Trolltech did release some version of the QtopiaDesktop for Linux but there wasn't a whole lot of activity and eventually, it became outdated and unable to sync with the newer Sharp ROMs.

Hopefully, as OEMs around the world start providing Linux pre-loaded, vendors like those behind the TomTom will start porting their desktop apps to cross platform frameworks and tools so they can support Linux desktop users. Too bad they don't learn from the router companies and put a web server in the device so any browser can work with it using standard protocols.

LoB

Re:it's about their target audience

[ealar dlanvuli]

Mac users get targeted because they have an extremely large amount of disposable income compared to their market share.

Sean

Re:it's about their target audience

[kocsonya]

Or take a look at Motorola mobiles. A handful of them runs Linux. The "developer network" is a bunch of users/companies who write applications that run on the phone. I repeat, not end-users but SW developers who develop SW to run on a Linux machine. You get the MOTODEV newsletter with boldface lines shouting "Motorola unveils $FLAVOR_OF_THE_MONTH Mobile Linux Platform!" and alike.

Now that means the platform in the phone. The SDK is *only* available for Windows. So they are catering only for developers who are happy to program *for* Linux but are otherwise living in Windowsland. For some strange reason I believe that Linux people would be quite inclined to write Linux SW, possibly more so than Windows people. Motorola apparently doesn't, they told me that they do not even plan to release dev. software on Linux (or on anything, actually, apart from Windows).

They probably figured that supporting the SDK on Linux would cost them an other engineer, which is a direct expenditure while the benefit would be a very indirect one, by having slightly more apps available for the phone making a tiny increase in phone sales - no deal.

Re:it's about their target audience

[Locutus]

funny you mention Motorola because they purchased Metrowerks and their CodeWarrior tools. Metrowerks was the company Sharp used to provide an IDE/dev env for the Sharp Zaurus on Windows via CodeWarrior. As I mentioned earlier, Sharp was looking for developers to write apps for the Zaurus and initially only had the Linux toolchain/SDK for 2-3 years but didn't provide desktop sync tool for those Linux developers. Late in the game, they later provided a Windows based SDK via Metrowerks/CodeWarrior but by then, many Linux developers were leaving the product. Sharp went and changed the PIM storage mechanisms and was really twisted in how much they'd allow open to developers. They also would not open source their PIM apps and people were wasting their time on building completely new replacements in an attempt to increase PIM satisfaction. BTW, one had to copy the new ARM based PIM database access library from the handheld and into the toolchain path to write apps for it on Linux. No source and the only documentation was a few header files and one API doc.

Motorola, now already has a Windows SDK(CodeWarrior) for their Linux phone and they do have both Windows and Linux versions of that. IIRC, quite pricey but they do have both versions.Still, there are a boat load of Linux developers who'd love to develop for the Motorola platform but will bypass that if it won't work with their Linux desktop and requires Windows for syncing to calendar and address books.

BTW, CodeWarrior is only $3,000 - $4,000 so most Linux or Windows developers are would be happy to develop for the Moto phone platforms. Not!

LoB

Re:it's about their target audience

[Locutus]

good point. It's gotta help that the Apple product pricing is typically in the mid to upper range. Their user base is mostly as you mention due to this. Maybe they also tend to purchase more devices like the TomTom because of this. They, IMO, are probably more likely to let vendors know they are disappointed by any lack of Mac support along with holding higher professional positions.

It was mentioned elsewhere but mostly, Mac is a fixed and controlled platform and it still makes it into the corporate desktop market better than Linux does. Mostly because of perception/acceptance but still, being the 3rd or 4th ranked OEM gives them recognition over Linux too.

LoB

LGPL

[tepples]

and even the LGPL can be a problem (see section 4d, which specifies that either source code sufficient to recompile and relink, or a shared library already present on the user's computer must be used.) As I understand the LGPL, you have to give the source code of the LGPL covered parts and the object code of any proprietary parts, which are called "Corresponding Application Code". So the other libraries don't have to be already present on the user's computer if they are designated as "Corresponding Application Code".

Re:LGPL

[fyngyrz]

So the other libraries don't have to be already present on the user's computer if they are designated as "Corresponding Application Code".

That's not how our IP lawyers see it; may I ask who did the analysis of the LGPL for you?

Re:LGPL

[tepples]

That's not how our IP lawyers see it In that case, they may have found a bug in the LGPL. Have they discussed the issue with the FSF?

Not just a Linux thing

[swordgeek]

At the ISP I work for, our customer-facing servers are all Sun Solaris systems, and there are a lot (many hundreds) of them. However, some of our applications are only supported on Windows, and in fact our previous webmail package only worked on IE.

Not much else to say. It sucks, but that's how things are at present.

Access to filesystem from Linux works.

[whoever57]

In fact, accessing the files on the TomTom itself is trivial under Linux. Just plug it in to a USB port and mount the filesystem (vfat).

I was recently moving all my TomTom data to a new (larger) flash card and the Windows application kept hanging, so I just plugged it into my Linux box, mounted it and used "cp -a" -- problem solved.

Easy Solution

[giminy]

Work it into GPLv4!

"Any use of GPL'd code for profit requires commercial support interfacing the device to GNU operating systems."

Okay, that's a horrible idea. Microsoft would just write a thousand GNU operating systems, and then demand support and put the company in question out of business.

Reid

Do it yourself

[bidule]

Is the documentation sufficient to write your own interfacing software? If yes, then this is just whining.

Having businesses using linux server-side (or embedded-side) means more job for linux developper. Which means a larger pool of said developper, more hands with larger variety of approaches mucking linux source code, and a more robust kernel in the end. It also makes open-source a thriving industry caught is a spiraling virtuous circle. I think they are already giving a lot indirectly, would you rather they "take no chances" and go the Windows way?

I am sick and tired of people saying they are sick and tired of people whining that nobody is doing what they want done, and I am sick and tired of hearing it.

Was that a retorhical question?

[Anita Coney]

No company wants to provide free support. So if a company can get away with not supporting some customers, it'll do it.

Depends on what you're using...

[CompMD]

I am a major user of software from UGS Corp (now owned by Siemens), in particular, I use their NX CAD/CAM/CAE software, which is heavily used in large scale engineering and manufacturing firms (General Motors is their biggest client I believe). Last year UGS released a Linux port of the NX software, and offered support. Looking at the pricing, both the Linux media kit and Linux support are noticeably cheaper than the Windows version of the software and support. I have used the support and never had a problem with the support techs, in fact, they've been great.

TomTomization

[michaelmalak]

Why do I smell a GPL4 around the corner?

Re:TomTomization

[Anonymous Coward]

Because this is where the goalposts will be moved to next:

"They might be following the spirit and letter of the GPL, but they're not supporting the Linux community!"

or somesuch...

Short answer: Money

[Paul Carver]

I think it was in a Robert Heinlein novel that one of the characters said that the answer to the question "Why don't they . . ." is always "Money". There may be exceptions to that, but when dealing with companies you won't often be wrong if you assume that answer to that question.

You might ask for suggestions on finding Linux compatible devices or strategies for advocating Linux, but asking "why don't they" or "why do they" is pretty much a waste of everyone's time. The answer is quite simply that the people at the company who are responsible for deciding which project will make money and which project will lose money currently believe that writing Linux software isn't in the "will make money" category. If they didn't believe that, you wouldn't be asking the question because they'd be one of the companies that is writing Linux software.

Frankly, your Ask Slashdot post is basically a whine. You want something and apparently you haven't been able to find a company that believes providing it is a viable business model. So you whine about "why don't they" do what you want. If by chance you've happened upon one of the rare cases where something could be done at a profit and nobody is currently doing it, don't post to Slashdot. Run, don't walk, to your nearest library and grab all the books you can find on starting a business. It's extremely rare to find something that can be done at a profit that nobody is doing yet.

TomTom Research in Berlin

[Yahma]

I used to work at TomTom Mobility Solutions in Berlin, Germany. Mobility Solutions is basically the R&D Department at TomTom. At TomTom, there is a distinct separation between the groups that develop the Map Uploading/Updating software and our team, which developed and researched new routing algorithms for the actual hardware. There was yet another Team based in Edinburough that developed the UI and software for the devices. While there is communication between the teams, they are in fact separate entities. The R&D Team I was in had no control over what platform the map updating software ran on. In fact, it was a management decision.

So there you have it...just because the devices run Linux, and the fact that our whole R&D Dept developed in Linux and ran Linux on our desktops, doesn't mean that it makes financial sense for them to put resources to develop and linux version of the map uploading software.

Embedded vs Desktop Applications

[mauriatm]

"Consumer" seems very misleading to me.

This seems really like comparing apples and oranges. Clearly embedded application of Linux is different from the desktop application. Just because Linux seemed ideal for one purpose it does not automatically make it great for something else. I would not be surprised if the developers who provide the PC update software for the device know little or nothing about the internal workings of the device.

Because those are unrelated?

[harlows_monkeys]

When you design a product, you have a target market you are going to support. Generally, that decision is independent of the engineering decision as to what technology to use when subsequently building the product.

Linux is not that hard to support...

[vlad_petric]

but linux desktop is ... way too many configurations/mixes of libraries/etc. Getting it right is much harder than doing the same thing for Windows/Mac.

Re:Linux is not that hard to support...

[Ravage382]

One thing that will be interesting in the future is Ubuntu LTS releases. It provides a fairly stable/static set of libraries and kernel. This may be something that could be targeted in the future by developers for linux applications.

That's an easy one.

[Minwee]

They probably saved millions of dollars using a free kernel and didn't think to support Linux users.

That's because they saved millions more dollars not developing and testing a whole different set of end user software that only a handful of customers were going to use.

Reasons:

[Opportunist]

1. They don't have to. They wouldn't offer you the source if the GPL wouldn't require them. The GPL doesn't require support for Linux, so it's their choice. Now let's see...

2. Supporters for Windows are cheap. Supporters for Linux are not. Simply by supply and demand. There's a ton of people who "sorta-kinda" can do Windows "somehow", or at least learn your standard interface quickly. Smaller userbase==smaller amount of people with experience==smaller amount of people with the needed experience looking for a job.

3. Developers for Windows are cheap...

Re:Easy Answer: 'The Almighty Dollar'

[Anonymous Coward]

I've posted about this before but everyone will just laugh at you for being a hippie that asks why they can't give back to the community. And it's true. 'The Almighty Dollar' is the answer to this riddle.

What about the other way around???

[Em Ellel]

One of the companies I work with develops code on Windows machines that is supported to run in production on UNIX servers only. Despite the code actually running fine on windows boxes, they do not want the (at least perceived) burden, overhead, and reliability issues of supporting Windows servers. I believe they are fully in their right to do so - but that makes the moral stance of forcing people who use Linux in their products to support Linux desktop clients a bit awkward.

Reality is that support of Linux clients and use of Linux are two completely separate issues. I think once there is a sufficient enough base of Linux desktops, companies will support Linux desktops, just like now there is a huge increase of Mac support where a while ago there was next to none. Its all about who the customer is, and for the most part, their customers are currently running Windows and OS/X.

two ways of doing business

[zogger]

1)try to milk out existing markets

2)develop new markets that look to have some potential down the road, where there is little or no competition right now

We have corporations fixated on the next quarter profits,all the way to the point of abandoning R&D and selling off assets, etc, and those looking for the long haul. Sure, you get a fast fat city bottom line that way, but it's *stoopid*

  Detroit in the early 70s vs. Japan, Inc. Who was actually smarter, which set of execs was actually looking out for their investors the best, the old "bottom line"? *Which* bottom line is more important, who's kicking ass now and who keeps having to dodge bankruptcy and junk bond status and so on?

FOSS-you either get it, or you don't, and it really is that simple, and to this day a lot of people even on this site just do not "get it". If you play act at "getting it", you won't receive all the benefits possible. Just try to milk it out short term with no sharing or thought to the users or taking a peek at the long view, again, it proves you don't get it or don't want to get it and in the long run you won't be as successful.

So, to all those folks saying the corporations are only interested in money, sure, I'd agree, but for how long? Do you want to make money for a long time, and just cede potential up and coming markets to squeeze out or cheap out a few extra nickles now in the short run? Is that really all you care about? Is it a good idea to cheap out on R&D, after all, right this quarter it's not "making you any money", now is it? Cheap out on embracing new customers? Slam up a website that bogues out decent double digits of the folks who use "alternative browsers" or OSes besides IE and windows out there, just tell those people to get stuffed?

Choices, business decisions, short range versus long range versus looking at ALL the ranges. Invest in your real business, invest in finding new customers instead of just milking the ones you have now, invest in research and share back because the more who do that the more "you" get back as well. That just seems to be a much better idea than cheaping out for the short run.

It's the hardware

[melonman]

Because supporting your own embedded version of Linux that no-one outside one small room in the basement of your offices is going to modify, on your own hardware, the spec of which isn't going to change, is relatively easy once you've got the thing working - in fact it's probably easier than supporting a proprietary embedded system. On the other hand, supporting any of a dozen major linux distros running on a thousand different hardware setups, using different sets of drivers for each and every peripheral, with the choice of at least two desktops and millions of permutations of modules, before the user started customising and recompiling, and no standard way to distribute your software to all distros apart from a tarball'd set of source files, isn't easier than supporting Windows or Mac end users. Especially given that at least some linux users are going to be more interested in proving they are smarter than the helpdesk team than in getting the product to work, and that a lot of linux fans will use a OSX or Windows when they have to.

And, as others have said, why would you expect one to follow the other anyway? If my company was making money from using an embedded OSS system, I might be inclined to put $$$ or developer hours into helping the OSS development community, but I really cannot see why I would be under any moral obligation to help the distributors of non-embedded distros I don't use or the desktop users who are consumers just like me.

because...

[spiffmastercow]

...if you're using Linux you should just read the code and figure it out, right? It's right there on the site, is it not? If you don't like it, fork it!

Why are so many missing the point?

[ndogg]

Everybody is saying, "It's just not economical to support Linux users!"

Nobody is going to argue that. In fact, I think it goes without saying. I think this question is more rhetorical to try to get people that work at these companies to give some consideration to the community that gave them a leg up in the market. It's about being nice to the community that helped out originally.

Linux users just don't get it

[LodCrappo]

Linux is an awesome server. It's great for many embedded systems. It makes sense for lots of "backend" type applications. It's used by lots of companies for these applications.

Linux is not an awesome desktop. It is lots better than it used to be, and it certainly looks promising, but right now, for most average typical users, Linux just isn't a good choice. And unsurprisingly, it isn't used by very many people as a desktop.

Companies, driven primarily by profit, seem to understand this and use Linux for the things it excels at. Linux users, driven by some strange compulsion to use Linux for absolutely anything, are in some cases compromising their own computing experience for the ability to proclaim "I run Linux. Everywhere. No matter what."

It makes perfect sense to me that a company might use Linux in a product and not support Linux desktop users. The company is using Linux for what it's good at, just like they would use any resource, because that is how a company makes money. They may (hopefully) even contribute improvements back to the community, but again these contributions are probably helping Linux become better at the backend task they are using it for.

On the other hand, supporting Linux desktop users is not a way to make money unless linux desktop users would somehow purchase enough of the product to turn a profit. So don't expect a company to do this becasue in most situations it would not make financial sense. Aa company is (supposed to be) responsible to their shareholders. They can't just support linux desktop users because its a cool thing to do.

my $0.02

Solution v 1.1 -

[Lodragandraoidh]

This can be mitigated in several ways without having to break the business model:

1. Expose the APIs used to access the device. This way the FOSS community can build an interface that will get the job done without using company resources.

2. Make the interface non-OS specific using standards. An http interface can be programmed once on the backend, and support multiple OSs via web browser (similar to how commodity IP router/switches are configured today).

These are ways of providing value add for the user, while at the same time saving your company money by only having to maintain one code base.

This is not rocket science - yet few people have the vision to make these things happen.

Distributions, Choices

[sydneyfong]

Aside from the obvious "nobody really uses Linux" argument as mentioned by many other posters, the problem that every "Linux" setup is potentially very different from each other raises the cost by a few orders of magnitude.

When you are writing software that is supposed to run on many client's machines, the monoculture of Windows (and Mac) help. You develop the software on a windows machine, and be reasonably certain that it will run on other windows machines. If not, these days you only need to test for winxp, 2k, 2003 and vista really. Having four different configurations for testing is easy.

Consider doing testing for Linux. You have Fedora, Mandriva, Debian, Ubuntu, Gentoo, Slackware, etc. If you're doing support software for a hardware product, you have to check whether the system uses plain old /dev with real device nodes, devfs, sysfs, udevd, or whatever. You have the various different kernel configurations, together with the patches included by each different distribution (and don't forget not everybody is running x86(-64)). If your product has a GUI, you'd have to make sure it works with GNOME/KDE/XFCE or whatever DE/WM, and since it's a GUI you wouldn't want it to look like crap on GNOME even if it was written with QT.

And unless your product is wildly popular, you'd guess that with good luck there'd be like a dozen people or so using your product for each different configuration tested. There's no way a sane person could convince him/her-self to spend resources testing on so many different systems, with the thought that maybe there's only like a handful of potential users that may be using this particular configuration, and that they probably have a windows machine lying around anyway.

OSS projects have less of this problem since they have a greater pool of testers, and they generally won't worry about getting their ass sued for some obscure glitch. So in early versions of OSS projects, they can simply release and rely on the community to report on oddities on their exotic systems, and maybe even get a few patch fixes. Most commercial companies don't have such a luxury. Especially for utility support software bundled with a hardware product.

Re:Distributions, Choices

[MrSenile]

Which is why when most people want support for the various linux distributions, they provide their own support libraries, staticly link the libraries they need, then release the binary.

This is why you would see a lot of generic binaries for Kernel 2.2.x or Kernel 2.4.x or Kernel 2.6.x and it would 'just work'.

With the graphical tools, X-windows is X-windows is X-windows. Weither it's Xfree or Xorg, it's still the same basic backend set.

As long as you don't rely on QT or other 3rd party tools, you should be fine. If you do rely on them, then it's fairly easy to include binaries for those as well, or at least point out the source repositories for people to get it themselves.

People say 'it installs on windows and just works'.

But pay attention sometime to all the several and insane number of DLL's that get copied around, all usually the same library, sometimes different versions of that library.

Linux could work the same way, just be a lot cleaner with how the libraries are used and much cleaner with how it cleans up afterwords.

The only difference frankly is people don't want to try and find it easier to say how impossible it is.

Having a multiuser application I built myself that compiles cleanly as 32 bit, 64 bit, on AIX, SunOS 4.x, Solaris 5.6 through 11.x, Ultrix, Tru64, BSDi, OpenBSD, FreeBSD, HPUX 9.x to 11.x, Linux (all flavors), SGI, IRIX, and MacOSX's bsd implementation, I can attest it's not impossible.

Posix compatability is there for a reason.

I know why ...

[Gaima]

... it's so obviously simple, we post links to their websites on /. bringing down their webservers.

Bottom Line

[Alari]

If every decision was based on profitability and all everyone ever looked at was the bottom line, these companies would not have a free operating system to run on their devices.

huh

[Anonymous Coward]

what's wrong with providing a closed source statically linked binary ?

Good Point

[bigredradio]

And then... Which distro(s) should they support?

I was just thinking about this. Often people say "support linux!", but that mean what? Supporting the kernel? I work for a company that supports Linux and that means a heck of a lot more than supporting the platform. Each distro with each window manager at each release and version level. It's a lot more than I think most users realize. Not everybody uses the latest Ubuntu. What about the Mandrake users at 8.1 using OpenStep? Saying you support "Linux" is a rabbit hole that goes pretty far down.

Re:Good Point

[Anonymous Coward]

What about distro X, Y, and Z? Fuck 'em. You can't please every single linux user but you can please the most popular ones. Support the top 2 or 3. This is on par with supporting Win9x, 2000, and XP. Same thing with window managers, and package managers. They don't have it? tough. I mean it can't be any worse than what they are doing now in saying "Screw them all!".

Do what a lot of developers who do release a linux version do:

- Support most popular desktop distros (Ubuntu, Fedora, OpenSuse)
- Release packages for those distros
- For everyone else give them a tar.gz and let them figure it out for themselves

Take a look at Nero for Linux. It is an example of a proprietary company supporting a product for Linux. They decided on debs and rpms, they decided on gtk2. What's the big deal?

Oh and to respond to others that say "Well no one is going to buy the software". We are mainly talking about drivers. I really don't think people purchase drivers separately, it is part of the product purchase. As far as drivers are concerned, pick 2.6. Not so hard to do and for gui/command line utilities they want to add use the above criteria.

People have got to stop with this retardation that supporting desktop linux means supporting 5280 different setups. It's stupid, impossible, and a waste of time to do so. What we need are informed people who know what they are talking about rather than the misinformation of a couple of people who bitch because there is more than one type of software that does a, b, and c. Those who aren't using the most popular distributions are going to be the ones who can work with a tarball. Statically linked binaries, while ugly on some desktops is also an option.

Sibling posters...

[Anonymous Coward]

...need to learn to read the entire post before responding. Even if you think all corporations are evil profit mongering behemoths (a complete and utter falsehood), the poster suggested that acting as a good citizenship may be the path to even greater profits.

A long list of reasons

[Anonymous Coward]

1. Because Linux users look out for their own!

Not much of a question.

[fish_in_the_c]

the answer is they don't want to spend time( aka money ) on testing and supporting a small market share. Sure it is probably the right thing to do , but since when does that have anything to do with $$$, and in the end that is all many companies care about.

Even Easier Answer...

[SerpentMage]

Its about the money!

These companies use Linux because it saves them money, because they have to pay nothing.

These companies don't support Linux because they don't earn any money with Linux. People who have Windows, and OSX pay money for their software and applications. People who use Linux are cheap skates, which goes back to the original point that they use Linux because it saves them money.

The real problem is not a chicken and egg, but to get Linux users to pry open their wallets and start buying Linux applications and so on. You will then very quickly see how people will begin supporting Linux. Yet that will not happen because people use Linux because you don't have to pay! See the problem?

I actually don't see an answer here, and that's why I am very skeptical that Windows or OSX will ever be de-throned by Linux.

Re:Even Easier Answer...

[Bert64]

What will make linux succeed is what made windows/dos succeed in the first place, and what you talked about...

Companies used windows/dos on x86 machines because it was cheaper than buying "proper" computers from the likes of IBM (ie mainframes). Once the inertia of proprietary lockin has been overcome, they will do the same with linux.
Home users will follow suit because they want to have the same as they do at work, just like they did before (The home market used to be ruled by the likes of Commodore).

Re:Even Easier Answer...

[ArTourter]

The real problem is not a chicken and egg, but to get Linux users to pry open their wallets and start buying Linux applications and so on. You will then very quickly see how people will begin supporting Linux. Yet that will not happen because people use Linux because you don't have to pay! See the problem?

Well actually it still is a chicken and egg problem. Linux users will only start opening their wallet to by applications when those applications are available. I would buy a Linux version of Photoshop if it was available, but it isn't. I pay for my version of Slackware because I think it is the right thing to do, and I have payed for crossover office when I needed it. Maybe I am alone here but somehow I don't think so.

One of the other reasons why some companies use Linux in their devices, is also because it is know by many to be stable and secure. that is also why they actually advertise the fact that the device is running Linux and therefore attracts customer from the Linux users. At some point these companies will have to take the risk of supplying software or support to Linux, some companies already do and the fact that OEMs like Dell or Acer are starting to provide hardware with some versions Linux installed as an alternative to Windows, we will probably start seeing some support from software vendors as well.

Re:Even Easier Answer...

[chromatic]

People who use Linux are cheap skates, which goes back to the original point that they use Linux because it saves them money.

People who use Linux don't pay for hardware? TomTom make GPS navigation devices.

Re:Even Easier Answer...

[PDG]

TomTom and all other major GPS providers sell software as well, in fact a lot of it. Virtually all the GPS device companies sell their own software for use on phones, PDA's, and other mobile devices as well as laptops and such.

Most of them don't even make hardware, they just resell PDA devices with their own software installed on them.

Re:Even Easier Answer...

[CommandNotFound]

These companies don't support Linux because they don't earn any money with Linux. People who have Windows, and OSX pay money for their software and applications. People who use Linux are cheap skates, which goes back to the original point that they use Linux because it saves them money.

Aren't we talking about hardware here? Canon, Tom-Tom, etc, do not make money from software, they make money from hardware, which Linux users buy a lot of.

Re:Even Easier Answer...

[the_womble]

The real problem is not a chicken and egg, but to get Linux users to pry open their wallets and start buying Linux applications and so on. You will then very quickly see how people will begin supporting Linux. Yet that will not happen because people use Linux because you don't have to pay! See the problem?

No, I do not see the problem. You might have a point regarding why companies use embedded Linux (to save on licensing fees). However, only a complete idiot would use desktop or server Linux just because it is cheaper. Licensing costs are tiny compared to the benefits of using the best software you can get.

I prefer FOSS for the usual reasons (trust, no vendor lockin etc.) and cost come a fair way down the list.

I even usually pay for Linux. I bought Mandrake 8 and 10, but not Ubuntu or Mandrake 9. I am currently using the download version of Mandriva 2007.1, but intend to buy a Mandriva Club subscription soon.

I have probably paid more to Mandriva than I would have paid for Windows, definitely so if you consider that I have spent the last five years in a country where Windows is de facto free (as in beer). Oddly enough, MS lets you "pirate" Windows, but IBM makes people buy licences for Notes, and Adobe does the same for its major products.

Re:Even Easier Answer...

[mpe]

Aren't we talking about hardware here? Canon, Tom-Tom, etc, do not make money from software, they make money from hardware, which Linux users buy a lot of.

Indeed it actually costs them money to provide software/drivers/etc. If they instead made the interface specifications available, maybe with some basic example software. It wouldn't be long before there would be plenty of software available for all platforms people were actually using.

Re:Even Easier Answer...

[pintpusher]

Head of nail, meet hammer.

I remember the days when a piece of hardware came with the specs:

• printers that included pinout diagrams for the interface and a listing of the control codes
• modems that included detailed information of all the registers and what they did
• heck even joysticks, IIRC, included pinouts

I remember writing little bits of code to control my hardware and it was fun and easy and you only had to deal with the bits you needed to use and ignored the rest. Oh my C64, how I miss you.

Because it's business.

[Kingrames]

You don't see many advertisements for 4-foot by 4-foot printers aimed at the housewives of America.
Because those aren't for them.

Linux is 100% a serious, business-oriented platform. It is NOT for the average user, and if it were picked up by the average user, someone would go out of their way to make a new linux distro that restored linux to its business only status.

It has nothing to do with the shortcomings of the platform.

It has everything to do with the fact that linux targets only the audience with the money and the brains.

Re:Easy Answer (or is that the stupid answer?)

[croftj]

If they used the right tool kits they can easily support Linux as well as Macs and Windows. Qt from Trolltech is just one of many choices out there that would provide an easy way for them at a reasonable cost provide apps that run on Linux as well as the others.

To say it's too expensive to do for the small returns is just a cop-out for being lazy!

Sounds Good but it's Wrong. So's the Question.

[Erris]

To save money. For most companies, linux is too small of market to be worth devoting development time to.

They just wrote the interface in GPL'd code, so you know they already have devoted the development time and might be keeping someone on staff that knows what they are doing.

Their GPL'd code is already "supporting" the user. Using reasonable interfaces and releasing specs is a good first step. Sooner or later this will make it's way to the distribution of your choice and your distribution will have a better copy than anything you can put in a box with the product. One of the great things about free software is the ability to get away from physical distribution and all of the version incompatibilities that plague the non free world. That saves money too. The best support will eventually be telling the user what distribution will be able to use the device without further effort. Next best would be for them to tell you what packages you need to install. The very worst kind of support they could provide is a boxed binary that's obsolete by the time it's bought.

The tide has turned, it won't be long before you are swimming in good desktop interfaces.

To Be a Good Citizen?

[SerpentMage]

The comment "It's about good citizenship, not an extra two cents profit per device." galls me.

Good citizen? Hmm, are not many who talk about the virtues of free software also those that have no problems ripping off the music labels or movie studios? Are not many of those people who talk about the RIAA if they were the scum of the earth?

Now while the RIAA might be harsh, it is after all their right to be harsh! What people seem to have forgotten with Free Software and Open Source is that they actually need to contribute back, and need to respect copyrights. The reality today is that people expect their books, music, and other content for free. And until that changes you can forget anybody being a good citizen. Because after all to talk about being a good citizen you have to be a good citizen yourself.

So I ask have you the poster and reader if you have ever not bought a piece of music? Not paid for a piece of software? Or downloaded a movie?

I personally make it a rule to buy my software, buy my movies, and buy my music! I love Open Source, but also respect the basis of the law that supports Open Source and thus think everybody should as well. I live off selling my IP...

They didnt always have this software...

[Bert64]

Their update software sucks, i found it very unstable when i last tried it...
You also never used to need it, you used to be able to download all the updates manually and just unzip them into the device (it shows up as a standard usb storage device).

Utility

[cybereal]

The answer is that Linux works great for doing what they are doing with it. It's a great little stable OS that can be configured for your needs and openly amended to work properly since it's OSS. However, the client-side of the Linux world is a nightmare to write for and support due to immense fragmentation. You would have to be a masochist and a poor businessman to try and support it directly.

Most of the Linux supporting commercial entities I've come across are either offering very very limited support (i.e. one specific distro) or they don't offer any support, just a build of their client and an "I hope this works for you" attitude.

I like when they do the latter, because at least you have something you can hack in yourself if necessary but it still isn't a real answer to the problem presented by the GNU/Linux ecosystem.

The Wrong Easy Answer

[camperdave]

They would have to spend an additional X dollars to support linux and that decision would only net them an additional Y dollars in income.

Why do people always equate support and giving back to the community with money, rather than information? All they have to do is put a few design spec documents on a web page, and the community will run with it.

Re:The Wrong Easy Answer

[dhasenan]

More so then than if the original company released a Linux driver themselves.

Chicken / Egg

[walt-sjc]

It's amazing how well Linux works on the desktop despite so many manufacturers REFUSING to support Linux for one reason or another.

The over all Linux market share for the desktop is low, but it's not zero. In terms of sheer unit numbers, it's still a lot. As more and more embedded devices use Linux (as well as other platforms (mobile) that are not Windows / IE centric,) the demand will grow for more compatibility / open protocols / etc. and manufacturers / sites / etc. will have to support it. Us Linux users are a patient bunch.

Re:Chicken / Egg

[Registered Coward v2]

The over all Linux market share for the desktop is low, but it's not zero. In terms of sheer unit numbers, it's still a lot. As more and more embedded devices use Linux (as well as other platforms (mobile) that are not Windows / IE centric,) the demand will grow for more compatibility / open protocols / etc. and manufacturers / sites / etc. will have to support it. Us Linux users are a patient bunch.

Embedded devices only need to interface well with the desktop; it is irrelevant what they use to drive the device as long as the connection works properly. If an embedded OS were to drive desktop demand for the OS we'd see a lot of PalmOS desktops; but since what is valuable is the abilty to transfer information from one device to another we see sync capablity developed for Macand Windows instead.

What could speed Linux adoption is greater use of a client - server model since then the ability to run the application is not always tied to the desktop OS but to the browser. Google could be the biggets driver of Linux growth if their Googiffce takes off.

Like it or not ... not enough users

[Lazy Jones]

Why don't more Linux-using shops reach out to the Linux-using community?"

Simple: because there are not enough users. In the 80's and 90's, all less popular computing platforms shared the same fate, they were largely neglected by hardware and software vendors. In order to gain support from vendors who cannot afford to duplicate their support efforts for a small minority, Linux needs a market share of 10-15% or more on the desktop.

lies, damn lies, and statistics.

[westlake]

Guess what - 97% of the population that buys wine on the Internet by the case at 20+ quid a pop does not run Windows. More likely - windows is under 40% and even that runs firefox or opera. Rest are MacOS and Linux users.

I have been trying to find - anything - on Google that backs this up.

Personally, I'd chance a modest wager that anyone buying wine "by the case at 20 quid a pop" is running Windows.

Re:lies, damn lies, and statistics.

[tfiedler]

I'm glad someone else said this.... Just because you pull a number out of your rear doesn't make it a fact. I'm with the person who suggests that most people who buy wine, in fact nearly anything, are in fact not running Linux... but Windows or Mac OS X, simply because they make up the bulk of computer users.

Re:lies, damn lies, and statistics.

[edwardaux]

Personally, I'd chance a modest wager that anyone buying wine "by the case at 20 quid a pop" is running Windows. Only a Windows user would buy wine for 20 quid a pop, when Linux users can download it for free.

Re:lies, damn lies, and statistics.

[mce]

I'm not going to claim that the numbers he gave are correct (on the contrary: they feel "weird" to me). But I can imagine that people who buy wine over the internet are more computer-addicted - and hence more likely to run OSX or Linux - than those who buy wine in a shop. Wether the addicts tend to buy more expensively, is something I can't judge, but it should not be excluded in by means of blanket statements. Real wine freaks don't use the net at all for buying wine anyway. Also, the likes of Mr Gates swim in money and may buy very expensive stuff, but surely don't do it over the net. They have somebody else doing it for them or have long-standing relationships with specific companies.

Any MBA worth his 3 letters will know that one needs to study one's own specific market instead of using very general and irrelevant numbers. Studying one's own market actually implies extracting the browser stats from the logs instead of using the "MS has 70%" number, so that part of the story is actually what they should do. The more tricky part is that one also needs to look at the segmentation of the market: which brower segments generate most revenue, as opposed to just "most items sold", or "most web site visits". This is actually what all those tracking cookies are for. Again, any MBA worth his 3 letters should know this.

PS: I have an MBA. And a CS degree. :-)

Re:lies, damn lies, and statistics.

[kaizokuace]

only hipsters and wannabies buy that wine for "20 quid a pop", mac users embody the hipster lifestyle. Windows users are too busy doing important things like playing video games and looking at porn.

Re:lies, damn lies, and statistics.

[Lehk228]

anyone buying wine online by the case is probably an alcoholic.

Re:lies, damn lies, and statistics.

[ricegf]

Personally, I'd chance a modest wager that anyone buying wine "by the case at 20 quid a pop" is running Windows.

Say, you don't happen to be a moron with an MBA who works for an Internet wine retailer, do you?

Re:lies, damn lies, and statistics.

[arivanov]

MacOs and Safari actually if the idiots at the wine shop in question http//www.virginwines.com/ are to be believed. By the way, I am not talking about MomPoP&Co outfits here. The shop in question is a part of the Branson's empire. Does that answer your question?

Re:lies, damn lies, and statistics.

[Bee1zebub]

If they're only paying GBP20 per case, that must be pretty poor wine. Given that wine in Australia is cheaper than in the UK, and the wines available from anywhere for that price even here are only acceptable after you've already drunk a whole bottle of something else, I would imagine he would be better off with cleaning fluid.

Sooo....

[glwtta]

Why do companies do something that saves them money, and then don't do something else that costs them money?

Yeah, that's a tough one.

The answer is simple.

[AntonDevious]

Linux is a great OS to build things with. You know what goes into it. You can build one version that works just like you want it to and then clone it in your device. But if you want to support Linux for the user, how can you guarantee they will have all the parts needed to make it run? What distro? What kernel? What packages? You have to maintain multiple distros of your product and hope to hit a zillions different variants for what is 1% of the users who want it. Windows, well for 90% of the market, you can put in the energy to support XP and Vista. Even for 9% you can support OSX. I love my linux, but I wouldn't want to produce commercial software for it and have to try and support it.

internal vs external use of Linux

[pak9rabid]

Why don't more Linux-using shops reach out to the Linux-using community?

There's a big difference between using Linux in your infrastructure/implementing devices and supporting users running Linux. Namely you can control how you're using Linux, it's a bit more difficult to control how others are.

Good GPS & map on GNU/Linux = wonderful car pc

[J4nus_slashdotter]

I imagined to build a car pc with a lot of features: music (via wifi, usb, memory cards), videos, connection with the existing car system, radars, alarms, radio, bluetooth,.. The big problem is to find a good software for the gps (map included). Tomtom should be a great choice..

Watchguard

[bondjamesbond]

Watchguard does the same thing. Their firewalls run Linux, but to control their firewalls, they have only Windows-based client software. Go figure.

They should have known

[phorm]

It's the linux users that run stuff on Wine, after all :-)

Two reasons.

[DdJ]

There are two reasons this might happen. One is a good, smart reason. The other is a lame, stupid reason.

Good reason: genuine cost/benefit analysis. "We ran a check of our server logs, and conducted this other analysis (appendix attached), and supporting such-and-such would only benefit so-and-so percentage of our potential users; meanwhile, the cost for developing for this target is this much, and the cost for developing for this wider base is that much. Those are the figures, now decide what to do." "Okay, we aren't going to support Linux or PalmOS for clients right now, but we'll re-evaluate our supported client list again next quarter, as usual. Next agenda item?"

Bad reason: boneheads quoting third-hand statistics and FUD, and then not thinking or listening. "Why aren't we supporting MacOS and Linux?" "Stop being a troublemaker, you know nobody uses those, and it costs way too much to support them."

Both effects happen out in the real world. From the outside of the organizations involved, it's not always easy to determine which is going on.

The one that got me was Watchguard

[51mon]

As someone who used/sold Watchguards firewall products, since they were called Seattle Labs, the days you built your own GNU/Linux server to install it on, I was gutted when they went to Windows only for the admin and reporting software.

True you could still get it to log to a syslog host (and for several sites we did), but they just assumed end users would have a Windows server around (and reliably so, which was even less certain in those days) to log stuff to.

Okay the GUI on the firewall box itself was probably not the greatest security idea, but I was disappointed when they didn't have some sort of cross platform management interface, as they clearly had the skills to do it (had previously done much of it).

On the other hand I don't think anyone can criticize their support of kernel development work to get the features they wanted in the firewall, but these kind of retrograde steps really tie folks to Microsoft, and it didn't do anything for the security of the companies concerned either.

I think AJAX is the free software worlds revenge, most of the big Ajax projects seem to eventually give up on IE6, which means Firefox if they don't have XP.

No suprise

[Anonymous Coward]

So... their GPS units run Linux.

The software to update only runs on windows and mac os.

If the software ran on linux then it would run on.... their GPS units. Then you could update one GPS unit off another, and , then... but... you could i guess update the GPS unit using itself. Then there would be a loud thunderclap, the world would end and the universe would suck itself up into a new big bang.

Gotta watch those infinite paradoxes.

It goes deeper

[Orion Blastar]

Linux is used in cell phones that only have Windows based software to access features on them.

Many PC makers use Linux for their web servers, but refuse to support Linux on their desktop PCs.

Tivo uses Linux, but the software to allow your PC to communicate with your Tivo box is Windows only.

Many video game Internet servers run Linux, but the games that connect to those servers are written for Windows only. Sometimes the Mac users get lucky and there is an OSX port, but not always.

PLC and PIC controllers run Linux, but the software to access and program them is Windows only.

Many PDAs like the Palm series run Linux now, but the software is only written for Windows or Mac OSX, and the Linux port was reversed engineered and not supported by Palm and written by open source hackers.

Re:It goes deeper

[Megane]

Sometimes the Mac users get lucky and there is an OSX port, but not always.

You forgot about the part where the Mac users end up having to use different servers and can't play with their Windows-using friends.

One more reason for empathy

[beetle496]

There is a long history (as in decades) of game publishers using Macs during development but not releasing products for Mac customers. I always thought anyone involved with in that, especially the Mac users at those companies, should be publicly outed and shamed. Like what we do to closeted gay politicians who campaign against gay rights. It may not be a crime, but it is a sever moral failing.

madethemanoffertheycouldntrefuse?

[Antiocheian]

Made them an offer they couldn't refuse???

That's from the Godfather, but what's the point in tagging?

GPLv4

[audi100quattro]

Thou shall not forbid interoperability between any two software configurations using the same code base.

Re:GPLv4

[adrianmonk]

Thou shall not forbid interoperability between any two software configurations using the same code base.

That's hilarious! I was going to make the exact same comment. There is only so far you can go in forcing people to be cooperative and to be good citizens.

And...

[sterno]

Not only that but the complexity of supporting Linux users is substantially higher than Windows or Mac. There's many more permutations of what a "Linux" system is than a Mac or Windows system.

Because its a pane to support linux

[DirtyFly]

HDesk : Hi, you are talking to Susan, How may I help you ?
User : Hi, im Joe user, I have a problem installingcoo the new timbuctu street map.

HDesk : So, what operating system are you using ?

User : Linux.

HDesk : Which 'version' are you running.

User : Ehrrmm , fedora core 6.

HDesk : ok, So you should do this and this ...

User : but, I cant see the xyz menu ...

HDesk : Oh, you are using KDE ? ok , just hang a moment.
...

HDesk : Im sorry, I cant find any one that is versed in the _insert_crazy_combination_of software_here_

This intro, although a little fetched, is not very far from the truth, Linux users are but a small percentage of users, and a very troublesome percentage, there is almost no standatdization of stuff, many distros have diferent setup methods for installing programs, diferent UIs and just to many variables to take into account, then, linux is NOT for the usual badly paid HDesk person, remember that many Support lines hire poorly qualified people, so hat would be the point of hiring well paid knowlagable pros (the kind that analyze the problem instead of following writen procedures) to support a handfull of users.

As for the provide a no support Linux package, with this I couldnt agree more, I could live with a package that had no official support, but that worked on linux.

Re:Because its a pane to support linux

[Shados]

User : Ehrrmm , fedora core 6

I had several people a -DAY- telling me they had Windows XP even though they had Macs (Macs were fully supported btw), vice versa, people who couldn't read, and so on...

Once Linux is fully supported by many company, we'll have to expect it to be only specific distros, and even then people will lie about them and such to go around that... Just figuring out which combination of software people have installed is gonna be -insane-, nevermind actually helping em with it at that point.

"Which distro are you using?", "Which version?", "KDE or Gnome?" (dang, Enlightenment, oupsies!), which videocard drivers? (official or Free??), which KERNEL (good lord...). It will be totally insane.

Even in the Windows (and even Mac!) world, large scale companies have to define certain configs they support and don't, otherwise it quickly goes out of hand. Nevermind in Linux... Its not impossible, but considering how insanely expensive call centers and such are (not as much with outsourcing, but still incredibly expensive), its almost a definitive net loss in most cases, even if you consider the enhanced company image.

Are Linux/GPL advocates being hypocritical?

[AHumbleOpinion]

"But why should a company support linux just because their gadget has linux running inside it?"

Because they are benefiting from a mature, open source, and well understood pre-established operating system. If there was no Linux they would have to spend much more development costs in building their own OS for their devices.

I am sensing some hypocracy here, not with respect to this poster but Linux/GPL advocates in general. When BSD folks complain about GPL folks not respecting the spirit of FOSS and "giving back"(1) there is a strong sentiment from the GPL advocates of "too bad, the letter of your license allow us to take and not give back". However when corporation comply with the letter of the GPL and do not "give back" beyond source code GPL advocates complain.

(1) For example in a scenario where a GPL developer takes BSD code, incorporates it into a GPL based project, makes minor fixes or improvements, but does not update the original BSD code with these fixes or minor improvements. Absolutely legal with respect to the BSD license but against the FOSS spirit of giving back to those whose shoulders you stand upon.

Re:Are Linux/GPL advocates being hypocritical?

[FishWithAHammer]

You are either a troll or a moron.

The GPL devs are not "giving it back" to the BSD devs, because the BSD devs can't use it in their projects

I know it's tasty and makes you feel good, but step back from the RMS Kool-Aid.

Re:Are Linux/GPL advocates being hypocritical?

[ClosedSource]

BSD code is already available to everyone forever. What the GPL does do is force new code created by others to be licensed under the GPL if it is added to the original GPL'd code.

We can debate the merits of this approach, but it is incorrect (and possibly dishonest) to state that putting BSD code under the GPL increases the freedom of the original code in any way.

Re:Are Linux/GPL advocates being hypocritical?

[jetmarc]

> The GPL devs are not "giving it back" to the BSD devs,
> because the BSD devs can't use it in their projects

A GPL dev CAN give his changes back to the BSD community, effectively "dual licensing" his changes. His main "product" may be GPL'ed, but a couple of bugfixes can be his "giving back" under BSD license, too.

Re:Are Linux/GPL advocates being hypocritical?

[Anonymous Coward]

When BSD folks complain about GPL folks not respecting the spirit of FOSS and "giving back"(1) there is a strong sentiment from the GPL advocates of "too bad, the letter of your license allow us to take and not give back".

How about:

Too bad, now you guys have been arguing for years that the BSD license is sooo much better than the GPL because it allows us to take and not give bad... And then you get p*ssed when we do what you preach.

If you wanted the GPL, you should have used the GPL instead of preaching that the BSD is better for allowing *exactly* this.

Re:Are Linux/GPL advocates being hypocritical?

[ClosedSource]

That's only a theoretical scenario (how often does this actually happen?) and one that is unlikely to be practical for more than a short time. Once the GPL'd code diverges from the orginal BSD, changes will be dependent on GPL'd code written by third parties and consequently it would be a GPL violation to back port them to the original BSD code base.

Re:Are Linux/GPL advocates being hypocritical?

[AHumbleOpinion]

That's only a theoretical scenario (how often does this actually happen?) and one that is unlikely to be practical for more than a short time. Once the GPL'd code diverges from the orginal BSD, changes will be dependent on GPL'd code written by third parties and consequently it would be a GPL violation to back port them to the original BSD code base.

You seem to be rationalizing the original GPL coder's decision not to live by the spirit of FOSS and give back. Furthermore, a bug fix is often so simple that it can not be copyrighted.

Re:Are Linux/GPL advocates being hypocritical?

[Anonymous Coward]

If they want people giving back, then they need to switch to the GPL, plain and simple. As long as they use the BSD license, they deserve that treatment. ie. "live by the sword, die by the sword". The GPL is a better license. Period. End of story. Anyone who refutes me is an idiot.

Re:Are Linux/GPL advocates being hypocritical?

[ClosedSource]

I'm not sure what you mean by "rationalizing" in this context. Sure, a bug fix that is made very early after GPL'ing the code may not be a problem, but it's more likely that a bug will be found later and the fix may no longer appropriate to the original code or may require the inclusion of third-party code that wasn't dual-licensed. I don't think I'm going out on a limb to say that the vast majority of GPL'd code (regardless of its origins) is never dual-licensed with BSD.

In any case, I think a bug fix is a pretty narrow interpretation of "giving back". I'm sure a lot of commercial companies who don't use GPL'd code would be willing to if all they were required to do was to distribute any bug fixes they made to the original code.

Re:Are Linux/GPL advocates being hypocritical?

[Frank T. Lofaro Jr.]

Sometimes you have to restrict freedom to protect it.

Re:Are Linux/GPL advocates being hypocritical?

[Rakarra]

However, the person who owns the code - the copyright holder, can release his GPLed code to the BSD project under the BSD license whenever he wants. It's a bit trickier though when the code is owned by multiple people.

Re:Are Linux/GPL advocates being hypocritical?

[mr_mischief]

So once http://212.12.47.19/~joebob/projects/bsd/whatsit/foo isn't found any longer and the only fork is inside Windows, how exactly is that "available to everyone forever"? The GPL enforces that "to everyone forever" through the law, while JoeBob would need to make plans to have his stuff mirrored. It's true that this type of problem is increasingly rare these days with all the cheap hosting and project sites out there, but it's not gone quite yet.

The GPL and the BSD are two different things with two different ideas behind them. The fact that the word "freedom" is bandied about for both of them doesn't mean it's the same freedom or that one guarantees more or less freedom than the other. When we talk about the BSD and the GPL, we're talking two different sets of freedoms and two different sets of promises.

One thing the GPL doesn't do when BSD code gets released as GPL is force additions and bug fixes to the BSD code to be released under the GPL license. It just creates a GPLed fork, some parts of which might also go back to the BSD branch. For all the fuss the BSD folks make over it, one might think people were taking away the BSD freedoms.

IMHO, releasing BSD code under the GPL does increase the freedoms associated with the code in some ways, but it also diminishes the freedoms associated with it in other ways.

Re:Are Linux/GPL advocates being hypocritical?

[FishWithAHammer]

But the likelihood of this is tiny. It simply doesn't happen often. So why bring it up?

(Mind you, it should be happening every fucking time a GPL developer uses BSD code, because it's the right thing to do. But, see, the GPL isn't about fairness, it's about keepin' dat code FREE!)

Re:Are Linux/GPL advocates being hypocritical?

[geminidomino]

Insightful? Who the fuck gave Bush mod points?

*DING* *DING* *DING*

[sterno]

And now you know why the "Linux on the desktop" bit has yet to happen after repeated prognostications of its ascendancy. It's not about ease of use, or any of that clap trap. It's about network effects. Why write software for 2% of the market when you can write software for 90% of the market?

The reason Linux has done well in server environments is because it doesn't have the same market share problem. There, you've got a far more diverse environment, and Linux, sharing much in common with Unix, was naturally at home in those established markets.

Linux did *not* make their profits possible

[AHumbleOpinion]

But wasn't that part of the point of the summary -- they saved a ton by using a premade OS rather than building their own. What's so hard about giving back to the community a tiny little something. After all, it is that very community that made their profits possible in the first place. It's about good citizenship, not an extra two cents profit per device.

Linux did *not* make their profits possible, nor was the alternative building their own. Linux was merely a convenient tool. If Linux had not existed BSD may have been an option, or perhaps the incremental cost of a commercial license (QNX?) would have been largely irrelevant. The point is that Linux makes a minor contribution to the profit, not a deal winning or a deal breaking contribution, but it is non-zero so why not pick it up.

With respect to good citizenship, I smell a little hypocracy among the GPL advocates. I don't want a redundant thread so I'll just refer to a different response for those interested. http://slashdot.org/comments.pl?sid=310539&cid=20775403

The obligation should not be to Linux's users..

[LordByronStyrofoam]

but to Linux developers/development. Many Linux/*BSD/OpenOffice users are moochers, contributing nothing but load to furthering the goals and needs of free/open source development. If they're going to give something back, let it be usable, modifiable open source code and interface specs, not canned apps. Then if someone wants to make a client side interface to TOMTOM, they can see the API clearly and scratch their own itch.

Canon ImageRunner series

[Anonymous Coward]

lease for $200 to $700 per month. They're good for no more than 200,000 copies before they're junked. That's nothing for even a mom and pop operation, about 15 months of service at most. Kinkos and OfficeMax can get better deals.

They're very Linux compatible, Canon sells (leases) a classy looking Linux mystery box for a $75/month upcharge. It's just SAMBA turning HP PCL into Postscript. The ImageRunners accept Postscript. We don't need it of course, we use Ubuntu, but your average letter shop is flummoxed. And Canon sales droids won't tell their customers to just output on a generic PS driver on their HP 4300 laser printers.

Canon makes huge dollars leasing and selling Linux boxen. I took one of these boxes apart, it's just an Intel P3 with a half gig of memory, a 500 Meg!! drive to spool, and a network card of uncertain parentage.

What about fucking gratitude?

[jotaeleemeese]

Just a thought.

Re:What about fucking gratitude?

[murdocj]

Here's a thought - you want someone to "give back" to you when they use your software? Well, you could require them to pay you for the software and not allow them to redistribute it. In other words, you could follow the closed source model.

What amazes me is the hypocrisy of people who scream about "closed source" and "proprietary" and then bitch about companies that use open source software in compliance with the license. If you want "open source" live with the results.

Re:What about fucking gratitude?

[JasterBobaMereel]

The point is that if it was closed source then there would be nothing anyone but the manufacturer could do about it

Since it is open source there is at least the possibility of the community writing the interface to Linux themselves

the linux community is a joke

[Anonymous Coward]

why is it that you handful of linux fanbois still think that you're a force to be reckoned with? maybe linux is on it's way up but the community is going to come down to those who advance and make cash playing the big boys games and the idealists who are left behind kicking out millions of lines of code while not making a dime. this isn't unique, this same thing happens to any marketable commodity that rises out of the hobbyist sect.

No money on it? Like advertisment you mean?

[jotaeleemeese]

Any company that plays nice with the Linux folk gets a bunch of loyal clients that will spread the world about how nice those guys supporting Linux are.

That is priceless, even if the percentage of costumers is not the biggest one.

Well, they are wrong.It is advertisement for geeks

[jotaeleemeese]

MBAs have this pesky habit of not being perfect. Also they are perfectly happy to spend money in something that delivers dubious results for the company like advertising.

One example ....

[Anonymous Coward]

I race homing pigeons. I recently purchased a Benzing M1 electronic timing system, the unit runs linux and has a usb port. Guess what? There is a Windows driver and Windows programs to talk to the device, no luck however from a linux machine. The provided windows software (pidexx) for the M1 can export data to winspeed files. Then one can run the Winspeed program, a Windows only program available from the American Racing Pigeon Union, to manage their pigeon races. From Winspeed one can generate an output file for the national database. If you try to upload your file to the national database from a linux system you will get an error message stating that you need to use IE. Arrrrgggghhhh!

An even better question: RS232 and NMEA0183

[kindbud]

It's not like RS232 and NMEA0183 are hard. Why isn't there an OSS mapping/navigation application that handles GPS waypoints and mapping using these industry-standard, publicly documented and royalty-free protocols? Every GPS device supports these protocols. The only reason I can think of is that there aren't that many GPS map users on Linux for that itch to have been scratched. So if there aren't many Linux GPS users, why should TomTom release a Linux version?

Why? Here's a novel possibility...

[uhlume]

Because (and this may come as a surprise to many Slashdot users) many rational people consider Linux to be a software platform -- not a fellowship.

Easy: Support Economics

[gsarnold]

This is an easy one - it's economics. The cost of supporting a continuously variable platform like Linux is inherently higher than commercial proprietary software.

Let's face it, when you say "I run Linux" you may mean kernel 2.4, kernel 2.6, Ubuntu, SLES, RHEL, DSL, Slack or any of a thousand variations thereof, not to mention home-brewed custom kernel builds and DIY distros. With Windows and Mac, you have maybe a half-dozen variations of each. If a dependency fails on Windows, you run Windows Update and update an ODBC driver or your VB libraries. On Linux, maybe you need to update a driver, or maybe the alpha driver the package maintainer slipped in there is preventing the software from working and there's no way to update it or roll it back automatically because it's alpha.

The point is, Linux varies widely from install to install. Until it becomes very VERY standardized, companies will generally continue to consider it not cost effective to support. When they do support it, it's only the commercial flavors like SLES and RHEL because they are standardized enough with a large enough installed base to train support staff to troubleshoot with scripts.

P.S. I love Linux, but this is the reality. That's why Ubuntu is so important.

Which Linux do you support?

[Runesabre]

I'm an independent game developer building a non-violent family game MMO service with a game client that runs on Windows, Mac and Linux.

Supporting all flavors of Windows all the way back to Windows 95 is rather trivial; there's one binary build of the application I have to distribute and support for all flavors of Windows users.

Supporting Mac is a little less trivial but still relatively doable with a little extra effort.

Supporting Linux, however, is quite challenging. First there's the sheer number of different distributions all running a different version of the kernel, different and even modified library versions. You can't simply say "we support Linux" with any real confidence.

Second, there's no guarantee of any sort of backwards compatibility with Linux. What you finally get working today will likely stop working on the next major release of a given distribution. Most distributions change every 6 months. You know when you get things working on Windows, it will work on nearly all Windows platforms past, present and future. In Linux, it's a constant deluge of re-fixing things over and over and over.

Third, everything is customized on Linux. It's part of its philosophy which is fine except when it comes to trying to support a commercial 3rd party product. The average consumer doesn't want the solution to their problem to be compiling source code to get your application to work on their system nor is that a feasible or economical way of supporting 1000s of customers.

If Linux really wants to become mainstream, it's going to need to greatly improve 3rd party support, distribution and start standardizing things like desktop shortcuts, menus and installation process.

Ironic

[Anonymous Coward]

That you bitches that have a *free* operating system are complaining that a company is getting something for nothing, and should give you something for nothing because you, er, what? You sit there and download linux for fucking free? So a company that makes and sells things should give you something, as one post said, out of "fucking gratitude"? Out of gratitude for sitting on your fat fucking arses and whining and bitching on slashdot?

Really? Are most of you linux users working on kernel tools? After all, Linus et al are all doing this for free aren't they? Oh, wait, no, most of the work is paid for by corporations.

Tell you what, why don't you bitches start showing some "fucking gratitude" and start paying us developers. You're all paid up on your redhat support and ubuntu donations are you boys and girls? No, thought not.

Re:Good GPS & map on GNU/Linux = wonderful car

[Macfox]

Licensing fees to use maps on display sizes larger than the commonly available TomTom and PDA's is expensive. Why it's like that...Who knows? But this is just one of a few stumbling blocks why TomTom and Destinator are refusing to support platforms that they can't guarantee maximum display size.

Re:answer - lots of lies and exaggeration here

[bit01]

They think supporting linux desktops is too expensive to be profitable.

And that's why we've got all the lying astroturf trying to mold perceptions here.

The fact, not the fiction those astroturfers and others are spouting, is that the major and minor app's that have been ported to both Windows and Linux are an existence proof that porting between the environments is not as big a deal as they are implying.

Practically every single "fact" those astrotufers are pushing is highly exaggerated. Different linux distributions? Open Office, one of the largest and most complicated applications around, doesn't seem to have problem. They have precisely 3 binary downloads for Intel Linux and and 2 for Intel Windows- with Java runtime(W+L), without java runtime (W+L) and debian (L). Changing kernel versions? Standards conforming digital cameras have no problem, they're using a USB memory driver (gosh, they're using standards! What a concept!). Level 1 support is impossible? Fine, don't offer it but for a standards conforming product it's no big deal to write a script that will eliminate all hardware problems and many software problems. Won't be fantastically profitable? So what, they might make a minor profit and they're establishing a market and creating mindshare all over the world. Companies create loss leaders and support and grow niche markets all the time for a host of different reasons including good will and/or mindshare and Windows/Vista/whatever-it's-called-this-week isn't going to be with us forever.

End of discussion.

M$ would certainly like it if the discussion ended there.

Next question!

Yes, we definitely don't want people thinking about alternatives.

---

Monopolies = Industrial feudalism

Admin Tools

[Mista2]

The one that annoyed me was when Novell were moving system to Open Enterprise server. NDS would work in Windows, on Netware and Linux, but the oneconsole portal, running in Apache and Tomcat, using java, only supported IE6! Huh? If I'm running 'nix server and I install E-Directory on my SLES 10 server, surely I should be able to admin the servers using Firefox?? I dont know if this is still the case though. VMWare - Nice system, ESX has a service console that is based on Redhat. We use all sorts of unix like tech to manage these, we wrote bash scripts to fetch patches automatically, SSH to terminal to them securely and transfer files, NFS shares for sharing data for the service consoles, but the stuff that really makes ESX fly is Virtual Center. This requires a Windows server, running MS SQL server for the database and the VI client software is Windows only. No Mac or Linux client, no syslog daemon for central event logging. Obviously aimed at Windows administrators. But wait, we are now running more Linux servers as guests because they take much less memory to run then Windows servers doing the same job, and we needed some Linux technologies to assist in administration, they boot faster and are supported by VMWare. So now we have more Linux skills in our admin team. I'd like Virtual Center to have a linux client for their next release or be able to run on 'nix, then I wouldnt have to run Windows at all for my job. As much as people say QT shouldn't be used because it's closed source, I think it's great as the same code works well in Windows, Linux and Mac. Why not have the VC run in Java? They already put a web engine on every host and on the VC server but it's not as nice or as fast to use as the full client. My 2c rant.

Why we don't care

[jroysdon]

I use Linux exclusively for some time as my desktop (5, 6 years?). Everything personal works just fine for me with one thing or another (mplayer, gnucash, etc.). When things don't work (usually Work-related products issues), I just fire up my VMWare GuestOS running WinXP and grab what I need. Linux users don't complain, or at least not loudly, we just find a work-around and move on.

For instance, Cisco Unity Express and Cisco CallManager 5.x run on Linux, but require (or required, the latest may have fixed it) Internet Explorer to configure.

Re:answer - lots of lies and exaggeration here

[Lord Ender]

My use of the word "think" was not by accident. The community colleges and Devrys of the world are churning out tons of windows developers willing to work for $50k/year. Developers knowledgeable in multiple platforms, or, even better, know how to code to standards, cost more to hire. Training customer support on multiple platforms costs more.

I can't tell which is more profitable. But thinking non-windows support is less profitable is not an idea wholly without evidential support.

When YOU own a business and middle management tells offers you the option to take $50,000 out of your pocket to invest in a linux development, QA, and support infrastructure, which might possibly grow your marketshare by half a percent or so, you may understand the question a little better.

I recently bought auto insurance, and when pricing it, I ruled out companies whose web sites did not render on Firefox/Ubuntu. But I didn't presume to know whether supporting these systems was in the business owners' best interests.

Because There Is No Connection Between Them

[Narcogen]

Why don't more Linux-using shops reach out to the Linux-using community?"

Why?

Because there's quite literally no connection whatsoever between these two things other than the word "Linux". That an embedded device uses Linux because a company has determined that this is the most reliable and cost-effective way of designing it has nothing to do with what market research tells them the majority of their end-users will be running on the desktop.

It's also possible, perhaps even likely, that the teams or individuals working on the embedded system in the device are distinct from those developing the desktop software run by end users.

There is also likely little in common, in terms of codebases, between the software run on the device and the software run by the end user. They need to communicate to each other, but they need not run on the same operating system to do so.

In fact, if having both the device and the end user interface run on the same operating system was that much of an advantage, in terms of development resources, it's likely that companies would produce their client software on Linux.

However, I'd say it's not, evidenced by the fact that they don't.

Seems to me the only argument to be made against this is on moral-- and rather shaky-- grounds. That since a company chooses to embed Linux in a device, that in addition to complying with the free software licenses that are applicable, that they should also cater to users of Linux on the desktop, even if their analysis of the market does not lead them to believe this to be cost effective.

Thank you! (+1 Funny)

[RandyOo]

You just made my day... I laughed so loud, I fear I may have woken up others in the house.

Usually, my moderation points go wasted and unused. Oh, how I yearn for them now!

Maybe they _did_.

[Anonymous Coward]

Perhaps they used a commercial embedded Linux distro (MontaVista, for example) or out-sourced the development to some company specializing in Linux. It is unlikely they did _everything_ in-house using an off-the-shelf Linux distro.

Sending money to a commercial embedded Linux vendor, a bookshop selling Linux books, or whatever is still a good thing, right?

Re:OpenTom.org is for Linux

[debstar]

You can find usefull informations about TomTom for Linux users on OpenTom. It seems the people who run this site are from Maintech.de and they are hired by TomTom to maintain their software. For the story about OS client : At the beginning they only provided a Windows client. Then equinux, a reference for Mac software, wrote TamTam. TomTom threatened with a lawsuit equinux which immediately halted the sale of TamTam. That's after TomTom released a Mac client.

You make me sick...

[denmarkw00t]

kind of. You see, the thing that got me into Linux - and that really makes it special IMHO is that the community will find a way to do it. Hack the firmware, reverse engineer it, etc.

And while vendor support wouldn't be bad, you should also consider that some companies, such as M$ and Apple, have more of a working relationship with hardware manufacturers so they're more likely to get support first, and likely last as well. And just as well - if you're a Linux user you're used to this kind of treatment in the world of hardware and drivers, so you do your homework. I don't go to Best Buy looking for a DVD of Rush Hour when I own a VHS deck, so why pick a new device that isn't compatible with my primary operating system? (granted, TomTom isn't just something to interface with your computer, but then again your whole post is about the device's [in]compatibility with Linux, so...yeah)

No, it was Theo de Raadt being an ass

[Per Abrahamsen]

I'm convinced that if you ask the following question to free software developers:

"If you take some free software, make some minor modification, and redistribute it as part of your own project, is is then considered good style to contribute the modification back to the original project, under the original license?"

almost everybody will answer yes, no matter what the original license is, and what free software license they themselves prefer.

But, on the other hand, if you accuse someone from doing something illegal by not contributing back, most people will act defensively.

Re:No, it was Theo de Raadt being an ass

[AHumbleOpinion]

"If you take some free software, make some minor modification, and redistribute it as part of your own project, is is then considered good style to contribute the modification back to the original project, under the original license?"

almost everybody will answer yes, no matter what the original license is, and what free software license they themselves prefer.

It is not "good style" to follow this policy, it is good ethics. Comments made by Theo are irrelevant, they occured after GPL fans decided that only the letter of the license, and not the spirit of FOSS, was to be supported. Unlike in the comment that started this thread where the spirit of FOSS is invoked to say that a manufacturuer should target Linux users. It makes the GPL community *look* hypocritical. I realize most members are not hypocritical, but the way Theo colors your view of BSD, comments like the original color the view of GPL fans.

Tom Cleese

[Anonymous Coward]

Soon after I bought my TomTom One I bought the Tom Cleese voice addon (yes, it really is him). His commands are full of all sorts of snarky comments, like occasionally telling you not to slow down in case the locals hijack your car.

If you want to manage tomtom maps under linux...

[Builder]

If you want to manage tomtom maps under linux, have a look at my brief howto at
http://www.penguinpowered.org/documentation/tomtom_maps.html

Linux users save the MS tax

[Anonymous Coward]

so they have more money (and more nouse) than their market share would suggest.

So they have more money and they are more technically inclined, so you don't have to ask "is it plugged in" "have you turned it on" and so on, making support easier. When you ask "does the udev system spot it" they will generally be able to answer.

Because the Linux users will cry...

[Anonymous Coward]

about the software and then write their own anyhow. :shrug:

xul

[Anonymous Coward]

Considering that the TomTom desktop software is written with XUL / XULRunner, and that the stuff that goes to the device is just files, Linux support is probably easy, and probably has been tried out behind closed doors. I'm sure that when the demand is high enough to pay for the extra support staff needed they'll do it.

It's business, but at least the TomTom guys have some foresight.

Mac os?

[ebvwfbw]

I never did understand why companies make stuff for the Mac and not Linux. They used to claim there were more users, clearly that isn't true now. They need to move on from the Mac.

The same goes for any embedded OS

[dan the person]

You don't see many QNX or vxWorks, symbian, etc etc devices that support those operating systems on the desktop.

Likewise windows mobile devices require windows XP.

On the desktop they support the common desktop OSes, and that doesn't include linux.

companies don't support linux for there employees

[Anonymous Coward]

Why do companies invest in a large server infrastucture, but refuse to support linux desktops for there employees? Maybe getting employees accustomed to linux desktops will help spread the use of linux and provide a larger market for linux software that is commercially supported.

the LOGICAL answer

[EdelFactor19]

because that would be far too easy obvious, intelligent, conscientious, and logical.. not to mention self-surving.

DUH

noob.....

[Anonymous Coward]

Why Do Commercial Offerings Use Linux, But Not Support Linux Users?

STFU NOOBs..... go RTFM.... duh.... its linux the way linux was before it grew up and got a job.

Some companies do

[KingVidalia]

Not to pull an advertisement, but our desktops (even our $299.99 ones), feature a 3-year standard warranty that includes 100% Linux support assuming you're using the version of Linux we installed when we ship it to you. We will support you with driver issues, gui problems, software issues and other things as long as it's designed for the OS we have installed (Currently Ubuntu or Fedora). Again, I'm not trying to pull an advertisement here--but my point is that our company obviously 100% supports Linux and I know we're far from the first. (And we are proud to say we support the GPL and other open-licenses) I think you just have to do some looking, support companies that do what you want (I am an avid nvidia fan for how good their tech support treated me over the phone with me running linux. They called me back and offered to fix my problem for free if I shipped my PC to california). Support those companies and others will have to follow suit if they want your business. If they don't want your business, why buy their products?

Linux

[jantman]

First of all, if the device is running Linux, the development team was probably using Linux, too. So you can bet that somewhere, their internal dev team has an update utility on Linux. Sure, it might be a command-line hack with no man page, let alone documentation, but it works! Two main points- 1) The real issue here isn't supporting one OS vs. another, it's supporting multiple OS's. If TomTom stopped coding their desktop client in VB or C++ or whatever they use, and wrote it in Java or Python or Perl or anything else, they could have it up and running on any OS the users wanted. Yes, it's proprietary code. But I know that both Java and Python have the capability for pre-compiled, obfuscated code. 2) The fact is that, for the most part, the majority of the Linux community is still fairly technical. I know that I don't speak for all, but if I could get a Linux map update client for my Mio, I wouldn't ask for much. I don't need a fancy GUI like the Windows client. How 'bout a simple CLI program, something like "mapUpdate /dev/ttyS0 /media/cdrom"

"official" Printer drivers suck

[OrangeTide]

I'd take CUPS and Gutenprint over drivers from Canon, HP, etc. Their drivers on OSX and Windows seem to be bloated and unreliable. (especially HP's) The open source guys are always a few products behind in support, but the support is generally less likely to hang or crash your system.

These days if you want a printer you're better off buying a used office quality laser printer for $200-$300 than one of those cheap ass brand new Vista-only printers with wireless and SD memory slots that you don't need.

$$$ and technical reasons too

[FranTaylor]

There's a difference between allow and support. Allow means "go ahead, but we won't support you". Most of these appliances have a Windows setup disk for the truly disinclined, but mostly you can just ignore all that and fire up a web browser. The setup disk makes support a lot easier. Some devices require use of the setup disk and that's a problem.

I don't think that making a nice autorun setup disk is as well-defined a procedure on Linux as it is on Windows or OSX, especially if you want it to run on SuSE, RedHat, Fedora, Ubuntu, etc. "automagically". LSB is nice and all, but it doesn't go far enough.

Having some sort of semi-automatic Internet install is nice, but it doesn't work for the person who buys a nifty laptop kerjigger on the road and wants to use it right now without messing with the Internet, especially if it's a network connectivity device.

because it's free, and support is not

[Anonymous Coward]

A huge motivation to use Linux is it's cost, which is usually $0.00

Another common motive is zealotry, i.e. people who hate Microsoft and just "vote for the other guy".

If you resell a product that you acquired for FREE, that very few people will use, why on earth would you then SPEND money to support it?

Free means free, duh!

[tjstork]

People who expect that a GPL consumer will somehow "give back" to the beyond what the license requires are foolish. Businesses do not exist to deliver welfare to developers. If you want software to be free, then you have to understand that you aren't going to get any money for it. You just aren't.

I imagine that in some future GPL, we'll start to see some sort of a redistributive tax for software, based on the logic that, well, the software is existing, therefor, people should be required to support the free community. There will be some uber commission designed to dole out revenues to the most worthy projects (aka politically connected and politically correct).

And, all along, people that who have said that the GPL is a socialist system will be redeemed.

Pareto's Principle

[DaveV1.0]

Businesses are in business to make money. It is not cost effective to support Linux in the way you suggest. And, as so many fanboys point out, Linux is "by geeks, for geeks". It should not be necessary to provide Linux users with that level of support.

I'd mod this up but it's already 5

[sys_mast]

This is exactly the 1/2 way that I personally think is reasonable. A major cost of "supporting" a client OS is the phone support. If they would release info on building support, or better yet, the program as 'unsupported' that would be perfectly acceptable to me. Even if it was binary only, a nice way around that is hosting the forums, so they can have a general feel for how well the binary works. That would even allow them to 'monitor' the forum and keep it on topic. All of which I could live with if it gets the device to Just Work!!!