Slashdot Mirror
The Failures Of Desktop Linux
PDAJames writes "Maybe Linux isn't quite ready for the desktop after all. After an earlier, very positive evaluation of SuSE Linux Desktop, ZDNet UK has carried out a more in-depth review, running the system in a production environment for two weeks, and found it wanting. A key problem area was interacting with the corporate Windows network. When will this stuff finally be ironed out?"
Well, the obvious solution is to get rid of all the Windows machines on the network. Presto, problem solved!
If all the world's a stage, anyone who says they want better lighting spends far too much time in a dark theatre.
The review is pretty positive, really. They admitted they were testing the most difficult situation -- non-technical people using Linux in a Windows environment -- and were impressed on many counts.
The fact is it's probably never going to be possible to switch operating systems without some minor glitches... switching will always cost money and time, so there's got to be a good reason to do so...
So, does it work well with OS X better than Windows on the network? I should hope so. It's kinda funny. If there's zero Windows boxen on the network, the OS X and Linux users would probably still have to use Samba. Bummer.
A programmer is a machine for converting coffee into code.
Novell is certainly not dead and has greatly fallen to the fud of NT. NDS and Novell provide the best NOS administration environment period! No lpdad is not an answer because its just a protocal and not a solution.
I use to be a fan of Caldera now SCO because of the promissed Novell integration.
Now lets wait for the next release of netware which is rumoured to have a linux kernel.
Relying on active directory is writing MFC programs and expect to port them to Unix.
http://saveie6.com/
This was the only stumbling block that prevented us from getting work done, but it is a serious flaw. The quick-moving open source community may soon solve the problem
Considering the age of Samba, shouldn't this have been fixed ages ago?
Then again, it is trying to implement a
Microsoft Proprirety Protocol , and we
all know how well documented (and static) they are...
"GNU's not Unix....it's Linux" / Kami "kokamomi" Petersen
How does it work the other way round? Is it really a case of Linux not interoperating with Windows networks because of the way Windows is designed, or would it be just as hard to get a single Windows box onto a Linux network?
What I'm saying is: surely the single, lesser box on a network is always at a disadvantage, Macs on Windows, Windows on Macs, Linux on Macs, etc. etc...
Opinions?
Hi, I'm a Slashbot. Linux isn't the problem, Windows is. So fuck Microsoft, but only when I'm not on my Windows partition playing games. And fuck the MPAA, only when I'm not buying the LOTR DVD. And don't forget to fuck the RIAA, but that's only if I'm not buying music in the stores or online.
It sounds like someone was trying to set up SAMBA without reading the documentation or they were lazy in matching the networks. Having used SAMBA in a mixed SUN and Microsoft environment, it was considered a godsend from both the Windows admins and the UN*X/SUN admins.
"All I want is a warm bed and a kind word and unlimited power." - Ashleigh Brilliant
Every day I say "when this comes out, Linux will be ready" and then that thing comes out and I find something else to do just that same thing with. The problem is that if we say that coming to the latest advancements of proprietary OSes is all we need, as we have been there many times, then they (proprietary companies) come out with something else. I say Linux will be ready for the desktop when it can outpace the development of its competition. With as many people working on Linux as there is, I think that this shows good promise. I have seen so much in the two years I have used Linux, it is amazing that we have come this far in only two years. In the short term, I think that Linux 2.6 is very important and if you want to know why, then just read some of the articles on 2.6 and that will explain a lot. I think that the freedesktop.org standards need to be fully implemented and now the the linux standard base seems to have eliminated a lot of the RPM incompatibilities, we are on the road to easy software use and installation.
(Christ, I'm going to get flamed for this)
I just got my hands on a copy of LindowsOS 4.0 (Thanks eMule) and installed it on my laptop.
Wow.
Coupled with apt (I ain't paying for Click-n-run), it is one hell of an OS.
I mean, a Debian install that just *works*.
Episode 1 : Microsoft is a failure /.)
Episode 2 : Linux is a failure <-- YOU ARE HERE
Episode 3 : SCO is THE failure (soon on
When the target stops moving.
Which will be roughly about the same time Bovines achieve lunar orbit.
and found it wanting. A key problem area was interacting with the corporate Windows network.
Well, actually the real problem is that Windows server software is wanting: it fails to conform to standard protocols and formats. If Windows server software was built from the ground up around IMAP, XML, HTML, HTTP, WebDAV, and other such protocols, then Linux desktops and Mac desktops would work well with it. While Windows currently nominally supports many of those protocols and formats, they are second class compared to Microsoft's proprietary protocols.
What's the solution? Get rid of the Windows servers. That also lowers licensing, administrative, and maintenance costs. And Windows clients can talk fairly well to Linux servers running open source software.
In other words, SLD is intended specifically for being compatible with Windows networks.
I do object to the "maybe Linux isn't quite ready for the desktop after all" comment in the
Try sticking a Windows box in a totally Linux environment, and see how that goes.
No NFS support, broken kerberos support, no NIS support that I know of, no ssh client or server, no X server so no remote apps. Sure, some of these things can be purchased and installed, but most of the windows versions subpar when compared with the real thing.
This study is like putting Michael Jordan on a special olympics basketball team, and then wondering why it didn't make the NBA finals.
I've had enough abrasive sigs. Kittens are cute and fuzzy.
I've been watching Linux for ages, and about once a year or two, I'll get a copy of a distro and give it a shot. This year I actually tried two, which are supposed to be the more user-friendly ones: RedHat and SuSE. While for the first time I managed to quickly set up a Linux desktop environment which did everything I needed, I still found it a bear to work with. RedHat didn't like my soundcard, the forums weren't much help. It took me two weeks to get SuSE to accept nVidia's drivers (because ONE character in ONE source code was off), and then after a week, it decided to stop using the drivers again. Never got Quicktime and most other video formats working. Opera for Linux isn't as good, and I've never cared for Moz. After a couple months of fighting with it, I finally gave up and went back to Windows. It's CLOSE to being desktop-ready, but barely a day went by that I didn't discover something I couldn't automatically do in Linux, and would require a day's tinkering to get working. And this was, as I said, after trying to different distros. Maybe next year... (braces for flames telling him he's stupid and evil)
Bush: He's Liberal in all the wrong ways.
companies that is.
You have this wonderful multi user OS and you use it on a single PC, arghhh.
Centralised computing is where most companies should be at, cheap disposable terminals on the desktop and a beast of a server under lock and key.
Linux will rule the enterprise desktop when companies grasp the mainfram had the right network architecture. Until then they're just wasting money.
Yes, I can see that meeting now:
CUSTOMER: Well, we're having trouble making our new Linux boxes talk to our large installed base of Windows boxes.
MICROSOFT: You don't say.
C: No, really. We'd really like it if you were to make Windows boxes easier to talk to by publishing your heretofore closed standard.
M: So let me get this straight. You want us to make it easier for our competitors to replace us? If we do this, then our market share goes down because the barrier to switching lowers. But if we don't, you'll keep buying Windows because it's cheaper than doing a complete rollover. Let me consider that.
C: Thanks, we'd really appreciate it!
Matt
Not sure why the poster got so bunged up over two interoperability criticisms.
The yahoo messenger thing and the outdated version of gaim is a bit of pain in the ass for a newbie but not a sysadmin. Good points all the way around on that one.
The LAN integration thing was interesting. I always end of with minor annoying bits of trouble with Windows networks until I load up LinNeighborhood and set the permissions on smbmnt and smbumount correctly for that app to work. We do this on the developer's desktops. We have tried all the KDE and gnome browsing tools and all that stuff. No go. Only LinNeighborhood really fit the bill.
Ok, what Windows browsing tools do you use?
I am using the samba browsing tool with Nautilus on Ximian Desktop2 as a try-out but I am already feeling the itch to get LinNeighborhood back.
What about you?
ACK
I kinda wish that existed.
Microsoft, as much as I hate them, is everywhere. The agressive approach to converting people to Linux - forcing them onto Linux computers - isn't going to work all that well. People need to get over their fears of the alien OS, and, to do that, we need to co-exist, side-by-side, until that fateful moment when the M$ system crashes and we're the only one left running.
Seriously, Linux needs to be there in front of the common end-users' eyes for a while for them to start wanting to use it. That means Linux has to be able to work in Windows environments, and it will be graded based on how well it works with other Windows machines and server setups.
Nicholas Eckert
vidstudent
Just imagine the result if, say, movies were judged on how close they are to the common denominator!
- Well, Gene, Schindler's List was tought provoking and great storytelling. Thumbs down.
- I agree, it's not worth seeing unless they edit it to add at least a gratuitous sex scenes that doesn't advance the plot. A few random car chases wouldn't have hurt either. Two thumbs down.
Feh!
If the only "problems" left with a Linux distribution are that "it doesn't do X like Windows" or "it doesn't interoperate with X of Windows" then it may be time to take a long, hard look at Windows.
-- MG
Doesn't play well with Linux boxes.
The message on the other side of this sig is false.
What good is Open Source if you do is wait for others to fix things?
Do you even lift?
These aren't the 'roids you're looking for.
I don't think you've ever had to support a corporate network in your life. When you get out of school and into the real world, you're going to find that end users and their superiors make most of the software/hardware purchase decisions based on their needs, and not the "bottom line"
My favorite example to cite is a sales department that uses Palm Pilots.
When you purchase a palm, it comes with software, written to interopolate with your outlook. All your contacts, notes, calendars are synchronized perfectly with outlook just by installing the software, connecting the palm cradle to the PC and putting your palm in the cradle.
Linux on the other hand has about 8 groupware solutions out there, so the first question would be what to pick? Then you have to figure what you need so the linux PC can see the palm pilot. Then you have to either push the installations out, or roll your own distro that has all of the components you need to use your palm perfectly with the linux box.
It's a lot more trouble than it's worth when you can just send the palm pilot retail box to the end user and he/she can install it themselves.
I don't really want to turn this into a tech support forum, but...
Probably your modem is a winmodem: a crippled piece of hardware which requires Windows software in order to work. See this document (somewhat old) for some help.
In other words, this is the hardware manufacturers' fault, and not the OSS community's.
Organizations that went from all windows 95/98 straight to windows 2k had the same type of issues with things not working quite right, and users not understanding the changes.
That's why you keep rolling up new ghost images that have the latest patches, tweaks, and workarounds needed to get your desktops working properly in a complicated enterprise environment.
As for the Samba problems, most can be ironed out by reading the documentation and checking newsgroups. There are irritating things about linux, but samba isn't one of them.
Ok, there's plenty of posts that say "just get rid of Windows" as a solution to the interoperability problem. However, if I'm generous and give Apple and Linux each 10% of the desktop market, that still leaves 80% to MS. You don't throw out a product with 80% of the market just because you can't get your minority system to work correctly with it.
When will Linux take over? When it interoperates with everything, so that people can get used to using it. Then, you can slowly migrate systems as needed, instead of going all out with one system, then having to re-train all your workers, and iron out all the bugs at once.
Overrated / Underrated : Moderation
For YEARS. Corporations and even schools like mine aren't going to be throwing away our Windows domains anytime soon.
A lot of Linuxheads point to Samba as some sort of Holy Grail. The problem that Samba doesn't yet solve is basic Windows Domain login support. You can't get share drives, rights, or home directories using this service. It merely creates localized Windows shares or lets you connect to them on an individualized basis.
The key for Linux to be accepted in these environments has more to do with network interoperability with MS, than app support. It sounds like an evil prospect, but you know something? I LIKE being able to organize my users on the network with ease. I like remote profiles. I like giving them things like shares and home directories in an organized way.
To my knowledge the only distro that addresses this is Xandros. The big problem here is that their Windows Domain support is closed source - to me, that kind of defeats the whole idea of using Linux in the first place.
When I first explored Linux options in 1999 I was shocked at the lack of this extremely important feature and continue to be. Let's hope the Samba project or something similar (and open source) will fill this in.
"...Well, there's egg and bacon; egg sausage and bacon; egg and spam; egg bacon and spam; egg bacon sausage and spam..."
Reviewers found that green made an inferior purple, and that until green was purple, purple would be the superior purple.
We had similar problems integrating Windows 3.11 PCs onto our Sun and Mac network[1]. They were completely crap at NFS and Appletalk. WindowsPCs will never be ready for the work environment until they can properly handle those two protocols.
Xix.
[1] Killed via CEO edict
"Everything is adjustable, provided you have the right tools"
First - I'm as much of a Linux fan as the next guy. However, after spending the better part of the day today in frustration with my new Linux desktop at work, I feel the need to vent.
My main bitches about Linux (not the kernel, the whole system - RMS's part, Linus's part, all the commonly-installed stuff...) as a desktop OS (and 99% of them are about X):
- X sucks hard in terms of responsiveness. Click a OS-level button (such as, say, the close button on a window) in Windows, that sucker responds. It may still be doing stuff in the background, but from the user's point of view, it's snappy and responsive. I'm running on a 2.4GHz P4 with a 10k RPM SCSI disk and 512Mbytes of memory for god's sake, there shouldn't be signficant UI lag! Win2k was about as snappy and responsive as you could get. I realize this is because MS built the GDI into the kernel, but come on, we're supposed to be better. As a modern business desktop user, I (typically) don't give a rat's ass about running applications on that server in the closet and having them display on my desktop. I want responsiveness...
- Bizarre-ass fonts. I realize this is mostly a configuration issue, but I've never found a distro that provided a decent font setup. Again, gotta hand it to MS, but Windows has a good, no-frills set of fonts that universally look good without taking up too much space. Those who configure X seem to have an unholy fascination with huge widgets and huge text.
- At least semi-standardized look and feel. Windows apps these days all sort of look and feel alike, but X apps are all over the road. This is the result of freedom, and that's not bad in and of itself. However, if we could agree on common places to put certain things, it would really help the user experience.
- And as a side bitch, why does GIMP not have an image browsing plugin? I know, I know, because nobody's contributed one yet. I'd help, but I'm an embedded guy - you really don't want me writing desktop software.
Okay, flame retardant suit on... Sorry, but those are my core complaints about trying to be a simple Linux desktop user today.
Perhaps you're problem was with Mandrake, and not Linux itself? After five years of using Linux and BSD systems as my primary OS, I've found that the most aggravating ones were those geared towards the desktop.
A Government Is a Body of People, Usually Notably Ungoverned
Computer: Libungif.so.012b conflicts with Libungif.so.013a, Unable to install software.
User: What? All I wanted was this cute screen saver with the Linux Mascot.
Linux is fast, Linux is stable, but Linux is far from user friendly.
Do you think any Windows user will understand that the have to use the "Make" command just to install a KDE theme? They just want to double click the installer and run the damn thing. And some of them don't even do that.
Did you know, that the early automobiles in London's cobbled streets, they were restricted to a speed of 4 MPH and were required to be preceded by a man carrying a red warning flag? All to not would not disrupt the horses and buggies on the road. That's not just an urban legend, and I'm sure you can see what that did to automobile performance.
How Linux and Linux software and formats interacts with for example Windows is not just a question of infrastructure and sunk cost. It's also about what you can do with Linux in a (desktop) world where 90%+ is Windows. A company can choose that they want to run a car and not a horse-and-buggy. But they can't make everybody else do the same, and if they have to proverbially run around with a red flag at 4MPH (lack of applications, incompatibilities, format lock-ins, suppliers or customers using Windows), what is then the advantage?
Kjella
Live today, because you never know what tomorrow brings
The cornerstone of MS's monopoly is in thier proprietary file formats. If it were software only, we would either be using something else by now, or MS Office would retail for $99 bucks.
If the US goverment would come out with a file format specification for standard documents such as word processing, spread sheets, etc and then mandate that the US Goverment use those standards, you would see the begining of the end for MS as we know it today.
Its time for the US goverment to begin building highways so that ANYONES car can drive on them.
What the poster wants is a pipe dream. Linux is not Windows, and it has it's own set of rules and design guidelines. A Unix network is a totally different beast from a NTLM/Active Directory network. THe protocols used are standard, and do not come in a package.
What teams such have Samba have done is pretty amazing by all accounts. They have gone from NOTHING, to a product which can enable a Linux server to server a Windows network without loosing many abilities.
The otherway is different. Yes a Linux computer can access Windows networks, and of course, no it won't behave just like Windows. But it does a damned mean job of accessing NFS shares.
You have to keep in perspective what we fight against. Creating interoperbility with Windows is chasing a moving target. MS will keep adding new things, like differnet encryption in XP, different encryption in Server 2003, and we will keep playing catch up.
This is a never ending cycle. For Linux to "win" the desktop, we need a clear goal of our own set that has advangates over Windows.
Yes, we need interoperbility, and we have that. It's not hard to set up a Linux SMB server, move Windows shares to it, and map it out over Samba and NFS, but it isn't plug and play, and probably never will be.
What it gives us though is a stepping block in order to migrate other boxes. Once Windows is out of the picture (as it is at my company), 100% interoperbility ceases to matter, and it becomes WIndows that needs to interoperate with us.
My two cents.
"Can you use a Linux system successfully in a Windows-dominated environment? That's what SuSE's Linux Desktop is designed to facilitate. We find that you can, although there are plenty of glitches to iron out."
Thats the article summary. Linux doesn't inter-operate with Windows perfectly just like Windows didn't inter-operate with the TSO editor and job-streams written in IBM's JCL language. So what?
For Linux to succeed it doesn't have to be perfect, and it certainly doesn't have to inter-operate with Window perfectly. Anyone who EXPECTS it to do that has simply made up their minds to continue using Windows already and is doing the "test" to satisfy the requirement that they do some sort of comparison shopping.
Nobody in the Open Source movement will be satisfied with Linux being adopted for any other reason than that it is the best choice. It's hard to imagin how it can be both the best choice and 100 percent compatible with Windows at the same time.
If there were nothing wrong with Windows there would be no Linux in the first place.
I don't think the article is saying "don't switch" they are just saying "it will involve some work". But we all kinda know that. Small shops or small departments switching will be a lot easier. For now, the most important switchers are individuals. As the number of people using Linux at home grows the argument at the office will get easier. Thats the way it happened with Windows too.
Until you try to hook things up to an MS domain. That's where it always falls apart.
People forget, but there was a time when there were other word processors besides Word. In fact, many of them had bizarre and confusing interfaces (Wordperfect for DOS, anyone?) When people had to do the inevitable switch to something new, they may have been befuddled for a while, but eventually, they got the hang of it.
I really don't think the issue is user acceptance near as much as ADMINISTRATOR acceptance. To get that, you're going to have to play nice with the existing infrastructure (after all, it was there first).
People can adjust to OpenOffice - we've done it here. But to replace our domain system with Linux would be near impossible. Forget the investments we'd be throwing out the door - think about all the other things like mapped shares, home directories, etc. It would be a massive undertaking to recreate all of that for very little reward.
I know MS plays their little games of half-assed interoperablility ("Windows 2000 is now based on LDAP and Kerberos! Well... Except for these little changes...") But if Linux is going to want to compete it's going to have to try harder.
Xandros has done this, but it's closed source. Kinda defeats the purpose, no?
"...Well, there's egg and bacon; egg sausage and bacon; egg and spam; egg bacon and spam; egg bacon sausage and spam..."
The problems you encountered (sound card, nVidia) are 100% installation issues.
They have nothing to do with Linux being ready for the desktop. I can install brand new hardware in an XP box and Windows will not know how to handle it. That is, until I install the drivers from the manufacturer. But that doesn't mean XP isn't ready for the desktop, does it?
If you had purchased a computer with Linux pre-installed, you would not have had those problems. If you had only purchased components with good Linux support, you would not have had those problems.
Those driver issues will only be solved when Linux has 50%+ of the desktop market. That's plain economics. The vidoe card manufacturers don't all support Linux to the same degree.
And claiming something should work with Linux because it is "from the most popular manufacturer out there" shows your lack of understanding. It doesn't matter how popular a manufacturer is. It matters how well that manufacturer works with the Linux community.
What happens when the conversation starts out like this? CUSTOMER: "We're having trouble having our brand new Windows box talk to our network of Linux boxes."
Linux does not need to revolve around Windows. Windows is not the center of the universe. As long as a host of applications run under Linux that satisfy the requirements of the user then there's nothing to complain about.
/Direct3D can be done in Linux as far as office productivity goes.
Almost every function with the exception of DirectX
Firstly, AOL is on its way out. As broadband becomes more pervasive AOL is being cut out. Why pay for AOL on broadband (an extra cost) when broadband by itself is almost as easy to use now?
The remaining customer base of AOL is probably technically inept enough to not even know what an OS is, much less have any desire to install it. AOL is not a target group at the moment, that doesn't mean it won't ever be, but right now, it is not.
Photos.
I think we all seem to be forgetting that Windows didn't start off as the supergiant that it presently is. In fact, when it started, Microsoft was were many of the Linux developers are presently: Underfunded, making good software that goes unnoticed or marketed under different names, and building something that would make the world work a lot more efficiently, except that they can't get them out there!
To compare Linux and Windows to the horse-drawn buggie and carriage may at first seem like a great analogy, but when you consider it more in depth, we easily see that it does not work. The problems listed previously about cost and interoperability are valid, but they're useless.
We've already agreed that 90%+ of the world runs Windows-based systems, whether they be servers, desktop machines, portables or whatnot. To try to take on this market head to head will never work and I think that Linux developers everywhere are beginning to realize this. We can't sit back and say "Things ought to be this way or that things will never work because of Microsoft. To do so will only frustate us and get us no closer to our ultimate goal. Instead, we need to be smarter than that.
We know that Microsoft will do everything in its power to lockout competition by giving us bogus source-code, faulty applications (that we must depend on), etc. In order for Linux to take a stand in this arena, what must be done is Linux must be created as an adaptable, usable and useful operating system that will not only work with what we've got already (saving users and IT guys time and money) but which will also allow us to expand and integrate what will come from Hell's Minions in the future.
Anyway, the short of it is that in order for Linux to work on a desktop level, we (unfortunate) must adapt and accept the Windows environment in its entirety and build our systems in such a way that they can expand and adapt as quickly and easily as possible. We will not kill of Microsoft or Windows for those of you who made cost comments are right. Companies don't want to spend millions training their people or redoing their computers. Time and money aside though, if they could find free (or cheap), feasible solutions that would allow them to conduct business as they have every other day of the week, I'm relatively sure they'd go for it. Linux does not need to play nice. Instead... We need to play smart!
No. We need the Free Software people to take their heads out of their asses and actually make something resembling an honest attempt at making their software usable.
For something that will slash government and school software costs, it's astonishing how much resistance we face
First of all, Linux costs more than windows or mac. The true cost of software is the cost of stuff it prevents you from doing. Or perhaps a better way of putting it, the real value of a piece of software is the total value of the work you can do with it. Right now, someone will be able to get far more work done with a non-Free Software piece of software.
In regards to the resistance, you people have called end users stupid for 30 years. You have repeatedly told them to shut up and read the fine manual. And you have been utterly incredulous when they complain the software is to hard to use. And then you tell them to quit whining about that which they are getting for free. Gee, wonder why there's resistance?
Free Software people have never recognized the Freedom To Get Stuff Done With a Minimum of Fuss as a valid freedom, and no end user wants to be deprived of that freedom even more than they already have been by Microsoft.
I suggest taking all that advocacy that is directed towards the public sector and corporations and redirecting it squarely at the Free Software developers who have repeatedly shirked their duty to make usable products. It's time to turn your guns on your own developers, because those are the people who are holding you back.
Ergonomica Auctorita Illico!
Insightful??? Gee...that's funny...I could have sworn I was using SLIP and PPP to dialin to my ISP from a Slackware box back in the early/mid-90's. I must have been hallucinating.
lack of AOL support != lack of dialup Internet connectivity.
No, the problem is Windows doesn't integrate with Linux.
This isn't Linux's problem, since Open Source projects often adher to open standards while Windows doesn't.
The solution is to fix Windows... oh wait, we can't.
Really? There must have been quite a lot of infrastructure devoted to providing, eg, feed and shelter for horses, not to mention horse shoes, vets, saddles, etc etc. Replace that with service stations, petrol pumps, mechanics, etc etc. Quite a big change in skills!
Coming more on topic, the real question is, why does such a change, from Windows to Linux, require fundamental infrastructure changes?
OK, the answer is obvious, but that simply highlights the core issue here. The technical problems with such a change are relatively minor. Microsoft could probably port the Windows XP shell to run on top of X in a matter of months. But is anyone else in a position to do that? Imagine that you had just invested in a bunch of carriages and horses. Then, the car comes along. Some clever engineer in your company realizes that it would be not so difficult to modify the existing carriages and add petrol engines to them. (Not completely unrealistic; that is essentially what the early cars were, after all.) But, you can't, because the company you bought the carriages from owns all the 'intellectual property' of the carriages, and retrofitting them is a violation of the license agreement. Anyone from 19th century London would have laughed at you!
Today's economic enviroment is based around the idea that it is easier to pay someone else to build something for you, than it is for you to learn the required skills and build it yourself. It is this notion that allowed 'companies' (ie. cooperatives of specialist workers) to form in the beginning. In the past, there was never any laws to say that you could not build your own copy of something that you already own. It simply wasn't needed; for virtually all goods, the cost of building it yourself was (and still is) much greater than the cost of going out and buying another. But this has lost most of its meaning when applied to software. The cost of making another copy is essentially zero. Why should we try to force this new paradigm into the existing economic model? And more importantly, what is going to replace it?
For a general office, Linux is not yet ready. Sorry, but Linux still lacks a great Office Suite. Star/Open Office has made great progress, but they are both slow and most, including myself, find the interface to still be a little clunky. This week I started a new job as IT director at a small start-up. Before this I worked for an Arcitecture firm as the server admin. Before I left, we tested Maya on Linux. The artists loved it. We found that fact that Linux uses fewer resources, we could use that extra power to shave about 2%-5% off our rendering times compaired with XP pro. One of my biggest complaints about Linux has been its lack of focus and how its developers attempt to make it a do all from a server to a Desktop all in one package: it ends up not doing either one as well as it could. Here that feature works in our favor because we can use the 10 or so Linux box as its own render-farm for large projects on the same box as the program. Lots of $$$ saved.
"The problem with socialism is eventually you run out of other people's money" - Thatcher.
Wal-Mart has increased its line of Linux based PC's.
Originally they had Lindows.
They have added Lycoris desktops.
They have added SUSE desktops.
There is a rumor that they will also be introducing Mandrake systems.
When Linux comes pre-installed, it is just as easy for the average person to use as Windows is. Wal-Mart would not be selling them and expanding their line if they were not profitable.
Though I've known this for a long time, it keeps hitting home every weekend as I travel to a small town flea market and sell used systems + offer cheap system repair and troubleshooting.
Most people out there simply want to buy a computer that runs "all the stuff I run across on the store shelves". I've tried selling perfectly good used PowerMac systems and run into this, just like I run into this if I have Linux pre-loaded on a PC that I put up for sale.
You find roughly 1 in 100 people who praise the fact that you're using Linux (or a Mac for that matter), and they typically spend the next 5 or 10 minutes chatting with you about the superiority of your choice, etc. Then they walk off without buying. (They've already got plenty of computer stuff at home.)
To the general public, Linux being "ready for the desktop" simply means it'll easily let them install and run all the "bargain bin" software on CD-ROM they picked up at Costco or WalMart, their copy of Microsoft Office they paid hundreds of dollars for a few years ago, and they really want to buy after they get their new computer.
This is, ultimately, why Linux won't ultimately be ready for "the desktop" for years and years, if ever. Apple still can't seem to pull off even a consistent 5% market share, and they have hundreds of commercially available software titles!
A Microsoft road would only fit Microsoft vehicles. If we draw the compairison to other M$ bloaty things, such as a browser that has a 1G footprint or a text document that consumes 50kB to say "Hello World", a microsoft car would be powered by three horses in a squirl cage, have 6 steel wheels that only fit on M$ patented rails, gets 2 miles to the gallon. Yes it would consume M$ gasoline as well as M$ geneticaly altered hay. Of course only one person at a time could ride in it and they would have no control over where it goes. The driver would also have to prove their identity via tatoed barcode and RFID tags, though the thing is actually leased and owned by Microsoft. Windshields and a roof would be expensive extra purchases. The horse's diet would be so poor that their performance would fail in two years, requiring the purchase of a new car. There is no owner's manual. The rider would suffer daily crashes of horse dung and often the gasoline would ignite and kill both horse and driver. The express purpose of the vehicle would be to keep everyone where they belong and mindful of their property.
There is no compairing Microsoft's hideous software to any practical device. Any physical device that was so difficult to use, performed so poorly, costs so much and worked so poorly with all established hardware standards would never be made. Ford made the automobile cheap and rugged. It was made to run on the poor roads of the day, be easy to fix and purchase by the common man. His express desire was to make it possible for people to get to know their neighbors, city and country.
Nobody is going to design a new road just to be able to run Linux.
No one ever designed anything to run Windoze either, despite the cute little marketing stickers. Microsoft's hand in hardware "standards" has all been negative, Winmodems, the destruction of unified graphics standards, web cams that require NetMeeting or don't work, sound cards that don't work, scanners and other devices that must be bought again on OS "upgrade". Their new software does not run on older hardware and their older software does now work with new hardware.
In short, M$ blows and it has given everyone a terrible impression of home computing. People are afraid to install and use software much less write any to do useful things. Because Windoze is so touchy, ureliable and sensless, they imagine free software to be a thing of vast complexity impossible to set up, grasp and use. Idiots like these ZDNet people perpetuate these negative impressions when the reality is that free software is extensively documented, configured with text files, extreemly robust and far cheaper to run and use. Because of M$'s bad reputation, people continue to purchase $2,000 computers that are little more than $400 generic computers with Windoze installed and "configured".
Friends don't help friends install M$ junk.
Friends don't help friends install M$ junk.
'When will this stuff finally be ironed out?'
When interoperabily with MS tools is no longer a concern. And MS format standards are no longer a moving target. With the lock they have on lobby groups on capitol hill- never.
OH THE SHAME I fell off the wagon and use sigs again!
Since the SAMBA team has been trying to get M$ to release the specs to SMB for years, without success, I would say that the limited LAN support is a problem with the Windows network - not the Linux desktop.
LRJ
i switched from Windows to Redhat 9, then to Suse 8.2. I'm never going back. So what is Suse isn't that great on Windows networks. Who the hell wants to be on/run a Windows network anyway? I'll take OS X along with Linux, but that's it. ;)
= chiisu
Right, wow insightful, like that new happens on Windows networks. After your description I'm sure that bug will get fixed right away. Are you sure it wasn't a problem between screen and chair? Or maybe the network was designed to not let you have access unless
Sheesh. What utter crud. Expect more and expect it often.
Microsoft's lack of open standards with stuff it develops?
Novell is certainly not dead and has greatly fallen to the fud of NT. NDS and Novell provide the best NOS administration environment period!
What Technicolor (TM) world are you living in?
My experience with Novell is 600 user environment at a large government organization.
I started collecting notes on the UI flaws that drive me nuts. Flaws in everything from GroupWise taking up half my screen to empty space in the Compose window (and if you change your UI preferences you stop getting e-mail because the recipient's view of your message is changed to match your own and your users can't find the Reply button!) to the stupidity of ConsoleOne. My notes are now in a three-ring binder that I wish to publish as a book before Novell's inevitable and too-long postponed death.
Another great problem with Novell is that it tries to simplify things too much, so you end up with people who don't know that you can telnet into a mail server (Oops, I'm sorry, GWIA) in order to see if it's responding properly. The whole idea of point-and-click administration is okay for small networks, but when you've got a mail admin at the helm of a 600 user network who thinks it's perfectly normal to be dealing in terms of proprietary binary mail spools that trap your information within their application like a Ph.D thesis written in Word, you've got a problem. Say "firewall". What's that? Oh... wait a minute, we have to remember that in this little patch of the world, it's called BorderManager.
Hell, they even think rebooting servers is a *normal* thing. (My boss, for example, is a Novell fan, and he flatly refuses to believe that the 115-or-so day current uptime of my webserver is possible.)
If Novell were a car, the hood ornament would be a 9-foot-tall big red N blocking your view through the windshield. Despite the road being clear to the horizon, you'd be unable to start the car until you cleared away a warning message on the dashboard saying, "Are you sure you wish to start your car? There's a tree 11 miles away and you might crash into it." When you finally manage to start your car, you need fear the tree less than the fact that all the passenger seats fall through the floor.
Unfortunately, they also have a rabid fan base, primarily composed (from what I can tell) of people who don't know any better. All the zeal of Apple fanatics but without the core of real superiority that Mac users can take comfort in. We're talking Novell golf shirts being worn proudly everywhere, Novell coffee mugs, etc. They keey on sending them to me, and I keep on sending them back to Utah. (All except the Novell golf shirt... I took that to an embroidery shop and had the big red N surrounded by a red circle with a line crossing it out.)
Novell and Corel are in the very rarefied position of being the only two companies in the world that I would thank Microsoft for running into the ground.
Fire and Meat. Yummy.
Actually it takes about two weeks to image 1,000 boxes with a new image. Did that myself this last month.
I've been on a crew that rolled out to enterprise class installations seven times. Here's what I know:
It doesn't matter what you're setting up -- the image is everything. In the enterprise a gang of really smart people put together the image of one ideal operating "environment" for one PC over the course of a few months. That's their full time job, and they're good at it. It is not trivial putting together a Windows image that works with itself, let alone all of the enterprise's custom apps. Getting a proper Linux image is equally non-trivial. They add things and test them and make sure everything the people use gets tested with everything else. They figure out which service packs break their required components and omit them. They have meetings and brainstorming sessions and teams and "pilot" projects.
Then one day their image is ripe for installation. They hire a gang like ours to go around and put the image onto the hard drives of the day's scheduled customers, and manually set some settings according to a script. Usually hardware upgrades (if any) are installed at the same time. Our crew travels to all their sites and performs similar operations on a regular cycle. Someone on their side performs server-side migrations if necessary. Their helpdesk crew are prepped for issues and ready to roll.
Our gang's favorite upgrade is naturally software-only. No trekking flat panel displays up the stairs and monster CRTs down. As yet we haven't done any Linux migrations, and (gasp) I hope we don't.
I know Linux is better. It's more stable. It's more efficient. It's more compatible. I prefer it. I use it at home. But if our customers find out they can have all of their software for free, and push upgrades down from the server without our help I get these negatives: no more software only upgrades (less work), hardware only upgrades relegate me to delivery boy status rather than tech.
Fortunately for me, our salespeople are unlikely to push a solution that kills their future business (hence violating the goose rule).
Help stamp out iliturcy.
I had an AVI whose codec Intel made propriatery so only 2000 would play it (XP would send me to a page to purchase it) on Linux it playes great with MPlayer.
"Computers have become so popular with normal people because the average person can just figure it out because it has been designed to be easy to use."
No they have not. You believe they are easy to use because you have spent years using them. I've had to teach people who have never used a computer how to use one. There was one woman who needed two hands to work the mouse. One to hold it steady while she clicked with here other hand.
The reason your computer works so well with Windows is that you are installing hardware that the manufacturer wrote drivers for Windows for. Is this a difficult concept?
Now, try installing Windows on a PowerPC.
How about trying to put one of these old PCI video cards I have in a Windows XP machine? No luck. They don't have Windows2000 or XP drivers. But they were top-of-the-line when they came out. Too bad the manufacturer stopped supporting them back in 1998.
What was that you said? I shouldn't be using old cards? I should buy new stuff that works with XP?
Well now. It seems that your XP installation has the same problems your Linux installation had.
It doesn't work without supported hardware.
I am aware that such has been your experience. But don't blame Linux for your experience. You chose the hardware to use. If you had chosen supported hardware, your experience would have been completely different.
And don't complain that the hardware you chose worked with Windows so it should work with Linux. If you want to play that game, then why don't you get XP running on a PowerPC? The reason the hardware works with Windows is that you are specifically selecting hardware that works with Windows.
Buy a PC pre-loaded with Linux and you'll see 99% of your problems vanish.
"An infrastructure is not ripped out and replaced in a day -- or even two. I doubt that we'll see Linux being used for wholesale replacements of corporate desktops in the near future. Until that day does come, Linux needs to play nice with the current prevailing technology."
Desktops:
For certain desktop aspects? Yes. Linux needs to play nicely, users aren't going to compromise.
That said, I think it's very fair to argue that with the same corporate setup (IT guys doing all software administration for 90% of employees), Linux does fine for the vast majority. The hardest part of using Linux is getting programs installed and working and configured, which is all an admin job. Like it or not, the rest is generally pie. Most Linux applications (Galeon, Gaim, AbiWord, Evolution) all have very easy-to-use GUIs. I don't see Linux lacking here.
Servers:
I'm sure some folks will correct me, and I'm sure that even more will nitpick here, but how well did Windows work with UNIX? Can it do NFS well? Kerberos? All the client software worked with UNIX I would assume, but somehow I doubt that NT was extremely UNIX friendly upon its introduction. Somehow I don't see the "new technologies must be friendly" in this instance. Can anyone (intelligently) point out a flaw in my logic? (it's late)
Cheers
~Dalcius
Rome wasn't burnt in a day.
"With linux, you need them, and if they , at C|Net didn't have them, it probably means they are not that obvious to identify and/or find for the average end user."
What end-users at a typical company these days know how to manage a Windows network connection?
Your analogy doesn't add up.
And who the hell looks for Linux apps at C|Net anyway?!? I have a hunch this might be your problem -- maybe a little ill informed but making an opinion anyway?
In addition, claiming that the "right tools" aren't "part of the system" (differentiate between a tool and the OS) in Linux is ridiculous. Windows has a number of services and tools built in, but nothing so specific that you'd have to hunt for a Linux equivilent in any modern distro.
For example, a typical Red Hat installation (or anything else, really) has enough software (already installed and configured) to log onto a Windows network and happily share files and print. This isn't opinion up for debate, this is plain fact. I've been doing this since I got into Linux, actually, about three years ago.
Just to help the logic-impaired, don't go flaming about "Ooh it's so hard to get SAMBA working." For one, these days, it's as easy as doing it on Windows. For two, again, what end-users do you know in a corporate environment that manage their network connection?
Sorry for the rambling.
~Dalcius
Rome wasn't burnt in a day.
Exactly, the German army, albeit the most modern and surely the most mechanized army of it's day in Europe, relied heavily on horses.
But in 1940 their panzers were inferior to e.g. the French, unfortunately the French used them in a way that was outdated (only in support of infantry, WWI style) while the Germans used them in a superior way (as the spear point of their army).
There is a known bug in Mozilla (known for AGES now) where it cannot copy'n'paste large amounts of text, due to a screwup with their implementation of the X clipboard protocols. Unfortunately the guy who "owns" that module, has never done anything with it. The bug still remains unfixed.