Slashdot Mirror
Where UnitedLinux Got It Wrong
An Anonymous Coward writes "A story on NewsForge [ed. note: part of the keiretsu] suggests that the lack of binaries
for UnitedLinux shows disrespect to the community which created most of the software. The author suggests
a better way for handling the business problems that
a lack of binaries is supposed to solve. Some particularly clueless
reader comments say that UnitedLinux has no responsibility
to cooperate with the community. The thought that
UnitedLinux won't even offer a development distribution
to the community does not sit well with me."
This is the exact thing the GPL is designed to prevent. The Linux community *won't* put up with this and UL will engender much public scorn.
This likely won't affect the four people left using Caldera, but for the rest of you, just don't buy it if you don't like their strategy. This is a commercial strategy, so you can actually hurt it by ignoring it.
So they don't want to provide binaries. Big deal- someone will compile it- some member of the community, and pass the ISO along to someone at linuxiso.org. Problem solved. I would argue that they aren't obliged to compile it for you, why should they? Because they use software harvested from the community? Ooooooh. Those UnitedLinux guys owe me lunch! Everything you could need from there is in the source.
Jeeze, get over it.
(No, this is not a troll. I just can't comprehend what is so terribly hard to comprehend about this.)
Working toward a usable PDA environment in the spirit of Newton OS: Dynapad
Correct me if I'm wrong, but it's legal.
Whether it "sits well with you" or not doesn't matter. What matters is how they handle their company and corporate image (as in: how other businesses see them). If they're within the GPL/LGPL licence bounds then quite frankly I don't see what the problem is, and I doubt any corporate United Linux customer will care if the distro offends a few sensibilities.
If they can come up with an innovative way to sell Linux, more power to them. They'll be doing something nobody else has been able to do.
Dragging people kicking and screaming into reality since 1996.
Red Hat, despite occasional faux pas, has been an intregal part of the Linux community from its outset. Where Caldera was primarily a consumer of OSS technology, Red Hat was at the forefront as a producer, giving back millions of dollars in professionally developed software. Most of all, Red Hat has made their distribution easily and freely accessible to anyone with a net connection.
SuSe, Caldera, et al, appear to be ready to board a doomed ship. Folks at work will recommend what they use at home. And most of the time that will be Red Hat. The others, by cutting off the community, also cut off a future pool of admins and consultants--who is going to administer their "United Linux" systems?
Good news for Red Hat. Bad news for the rest.
Oh, you could compile and distribute the GPL'd code, but you couldn't redistribute whatever code the UnitedLinux contributers put into their distro that isn't GPL'd and that's covered by their licenses. Because then YOU'd be the one violating copyright and be very much in the wrong.
Also, you couldn't call it UnitedLinux unless you were releasing the entire distro, which as I already mentioned, you can't, so you're doubly screwed.
They are doing nothing illegal. I'm more than sure that the pack of lawyers any company needs to survive today has already gone over everything they're doing. So get over it.
This would be a non-issue if Ransom Love understood Linux PR. If he answered the question differently, with the same "content", we wouldn't be as pissed. Something along the lines of:
"The United Linux organization will assemble the source code of the product, which will be available to the public. We have decided it is the responibilities of the vendors to compile the product for their specific distributions. Since the raw code is not indended for end user use, UnitedLinux will not expend the resources to compile and maintain a binary distribution of the raw codebase, that is the responsibility of each UnitedLinux vendor..."
The "public outcry" may have been different. Same answer, same question, different spin.
-Pete
Soccer Goal Plans
UnitedLinux is a base from which Linux distributions will be produced.
Want a developer platform? Get SuSE 9.0, or OpenLinux 4.0, or Turbolinux WhateverItWillBe, etc.
You don't develop directly on top of UnitedLinux, folks... You develop on a distribution that is built from UnitedLinux.
Now for pete's sake QUIT BITCHING ABOUT BINARY DISTRIBUTION!! You'll have the sources, and you'll have Linux distributions that are built from UnitedLinux. If you want more, I can't imagine what it could possibly be!
DAMN this is getting old quick!
Oh cripes.... please show me one common distro that is named hacker linux Leet-dude linux or I 0WN J00 linux.... get real. slackware and redhat are well set in the business world.. as well as Mandrake has made it along with for embedded white dwarf linux and Midori... Please oh Please... STOP this FUD campain... the entire decision of accepting linux into the workplace is based completely on the one person that bring it up for a decision.. if that person has any management and sales abilities it will get accepted... if they are the typical, refuses to wear anything but a t-shirt, pierced everything, over-tattooed, wannabe "hacker" they will ignore any suggestion or reccomendation they place on the table... and finally you have the typical Unix-lord.. hair and beard to the floor and wearing an ORIGIONAL Commodore Vic 20 T-shirt.. will get his suggestion to use linux accepted because he will let the board know "WE've been using it for 4 years.. It's saved the IT department tens of thousands of dollars already.. and it runs your pet project from last year.."
Corperations dont want to feel secure... they want to feel money... lots of money, profits, lower operating costs, greater profit margins...
If you go to your boss and say, "I can save you XXX dollars this year in Server software." he will do it if you have your presentation well thought out and shows all costs involved... he also will hold your butt to it and will gladly fire you if it fails... so if you dont have the guts to stand behind your reccomendations then don't do it...
Me? I have Linux infiltrating over 30 offices next month.... why? because I said it will work and they can have my ass on a platter if it doesn't.(exact words in the regional board meeting) Taking ownership of your decisions and reccomendations not only get's linux in the door really fast.... but it also get's you up the promotion ladder a whole lot quicker...
As for wanting support, where have you been? linux has more and better support than any microsoft product or OS. I can make a call and pay LESS than a microsoft tech support call and get faster,better,competent results.. When was the last time you had one of the IIS developers answer your question instead of a non-trained lackey responding from a canned support script? I have had answers from the apache developer responsible for the section I was having trouble with... WITHIN 24 hours... unlike MS support (I still have an SQL server issue open with MS... it's been 3 weeks now... I could have switched the entire system to linux+oracle in that timeframe.)
please, if you dont have the strength to stand behind your decisions and reccomendations... DO NOT reccomend linux or anything other than a nice safe MS product.. something you can point fingers at when it fails, something to transfer blame to... Installing an Open source solution takes strong leaders that are willing to say "It's MY fault the server crashed. It's my fault that we cant recover the data, It's MY FAULT that the new system failed and let hackers in/ a virus in/porn floods in....
and Yes, It's MY FAULT that we haven't switched everything to linux when the next virus wave rips throught the outlook hole and cripples the network.. and that is EXACTLY what I will say at that meeting... use everyopportunity to get linux in.. but only if you have the ability to.
sorry about my crusade-rant here.. but I am sick of people claiming they know what corperations want... they dont and until they get off their butts and learn what corperations want they will sit all safe in their cubicle making assumptions..
you want to be your bosses boss? you cant be comfortable.
Do not look at laser with remaining good eye.
One good example, and a project that I've made some small contributions into, is SDCC - Small Device C Compiler. Like many projects, there's a CVS server, nightly snapshots (source tarballs), and even nightly binary builds for linux and win32.
Do the linux builts work on Redhat 7.2? Yes, I have Redhat 7.2 on my system, and so do other active developers.
Do the win32 builds work on Windows 98SE and 2000? Yes. None of the developers uses windows, but many users regularily download the win32 builds on these 'doze systems. When bugs have turned up one these systems, the developer mail list and bug tracker have filled with bug reports.
Do those win32 builds also work on windows XP and windows 95? Good Question. A number of non-developers have probably tried it on XP by now, but are any users really still running it on '95? Who knows.
Long ago, the code could compile with Borland C. Developers using linux-based systems started cross compiling and building under Borland was broken for a very long time until someone reported it. Borland support has never been restored (no active developers care, and mingw as a cross compiler can be fully automated by the nightly builds)
Does the code compile and run on PPC? Yes, one of the developers has a PPC box, so this definately works. Which distro works for PCC... better ask Michael.
Does it work on Debian? Yep, several active developers use Debian... and there was recently an announcement that it's available with apt-get. Debian and RedHat 7.x are the only two distros that are really gaurenteed to work.
Does it build on MacOS-X? Nope, but a couple of OS-X users have joined the developer mail list and it's probably a matter of time until the OS-X issues are worked out, but at least it is know that there is a problem on OS-X (I think someone submitted a patch but it hasn't found its way into CVS).
Will it work on UnitedLinux?? Who knows? Nobody involved with this project has UnitedLinux, and it doesn't look like anyone will. RedHat 7.2 works great for me, and since these "business oriented" linux vendors aren't going to provide me with a cheap/free cdrom or network install, why would I bother. They're focused on serving the "business server market", so there won't be anything interesting about UnitedLinux to make it worthwhile for me to bother installing (not to mention paying for). I'm certainly not going to waste my time to compile an entire linux distribution. I'd much rather spend the time getting to understand the register allocator code better and make more significant contributions to SDCC. I'd be very suprised if any other developers lifted a finger to test UnitedLinux.
So the subject, "big friggen deal" couldn't be more appropriate. As a developer (primarily firmware, using tools like SDCC), I'd say "big friggen deal" about UnitedLinux. Caldera, TurboLinux, et all probably say "big friggen deal" with respect to SDCC, since their customers aren't developers, and they certainly aren't developing firmware for low-end 8-bit microcontrollers.
So if with want something "business oriented", supposedly with "world class support" from the likes of Caldera (or you just want to pay them a lot for something that's roughly equivilant to Redhat), then maybe UnitedLinux is for you.
But, if instead, you're interested in using the linux distribution with the most "third party" software that's tested and known to work, your best best is going to be with RedHat or Debian. (FWIW, you might pause to wonder why Windows is so popular).
By not building "mind share" among software developers, how can they ever expect third party software to be tested with their distrbution?
PJRC: Electronic Projects, 8051 Microcontroller Tools
It's always been mightily confusing to me when people use the term "engineer" and talk about OS advocacy in the same breath.
"I don't know that atheists should be considered citizens, nor should they be considered patriots." - George Bush
Have you ever even tried to build a working distro completely from sources without even a pre-existing binary image for a minimal OS to work from?
The assumption that a developer working on one tiny area of the OS, or more particularly working on a user space application will always want to install a full distribution piece by piece from sources is absurd.
That's not to even begin to approach all the secondary issues. Seems that this is a fairly transparent attempt to use the amount of work involved in putting a distro together from sources as a weapon against the community that wrote most of the code.
GNU/Linux mostly gets into the server space because it is championed in, usually by someone who is not a developer, but is a user. This will come back to bite UnitedLinux when they find that all those same advocates sneer at their distro because of the impossibility of familiarising yourself with it by using it at home.
It's a shocking business mistake, but that is not the grounds on which to reject it, but rather the sheer contempt for the community from whose work they seek to profit
The anti-management sentiment on Slashdot is disgusting. Look, we ALL love Dilbert, and while there is a LOT of truth in Dilbert, it does ignore the other side.
Businesses, large and small, want to make money. Within large organizations, you do have some empire builders (managers that only care about building up their clout through headcount), but senior management normally looks at the bottom line.
Sure, CEOs may not realize that there is a strange policy in one small division that makes no sense, but they are overseeing the general policy decisions.
Slashdot users seem to think that all managers are like the assistant manager at the fast food joint they work at. Senior people work weekends. They check email and take calls all weekend. They take risks.
Middle management avoids risks, but so do engineers.
There is, however, a BIT of truth in the corporations want to feel safe. However, that situation is in large companies where IT is a small percentage of costs. If 15 minutes of network downtime costs the company more than the IT Departments monthly budget, then they are interested in feeling safe.
It all depends. If you are a high tech company, IT is a big chunk of your costs. If you are a manufacturing/distribution company, IT is likely a smaller percentage. However, if a network outage shuts down a factory where 2000 employees are doing nothing for 2 hours, they aren't going to be happy to learn that you passed on the redundant hardware Sun solution and hacked up a beige x86 box to save $15,000.
Corporations want to reduce costs and increase productivity. Part of this is lower cost solutions. Part of this is more reliable solutions.
Where Linux hurts is in the reliability camp, and that's largely a hardware issue. x86 hardware simply doesn't play in the same space that Sun's high end Unix Servers and IBM's mainframes do.
The fact that Linux is more reliable than Windows isn't that impressive... Intel was still (as of 2-3 years ago) still running their manufacturing processes on VAX systems.
However, saving $100 may seem like a lot to the users here posting from the parent's house, but it isn't a big deal IF it exposes you to greater problems later on.
Alex
They're distributing the source.
They're not providing binaries.
And? So? The problem is...?
Jesus Christ, what a pathetic bunch of wankers. (That goes for the author of the article, as well.)
"Developers won't have binaries!" Huh? If "developers" can't figure out what to do when they have the source, but not the binaries, no way in hell do I want them "developing" on a system I intend to use.
Not that I intend to use UL.
You're assuming that developers will jump through hoops for the privledge of having their software work with UnitedLinux. You're probably wrong. If UL makes it too hard for develoeprs to test their programs on their platform, it won't happen. It's difficult enough to get developers to test on anything other than x86 RedHat or Mandrake, let alone any BSD or non-x86 UNIX-like OS.
I used up all my sick days, so I'm calling in dead.
"Because how can you possibly succeed through giving and sharing?"
This is why these developments are repugnant to a lot of free software people: speaking for myself, I think it's wonderful if these companies are prepared to abide by the letter of the licenses, but they're setting up a situation where they'll be saying, 'the reason you should give us money while we avoid doing more than the minimum we can get away with is, you can't go around being altruistic'.
This is not an argument that I feel needs additional support. To me, the argument, 'great things can happen when people cooperate on something altruistically' is the one that deserves the support. It's true, it's produced the body of open source software, and it's challenged every single time some bozos at a company decide to do Their Special Thing on top of open source, and combine their proprietary software with the free stuff.
Because they will ALWAYS say, 'we're value added', and they'll say, 'You have to expect to pay for the added value, of course', and they'll say 'you can't expect to be given things for free', which is exactly what was done for them. It's ingratitude, and it's distracting, and it's capable of confusing people as to what's really going on- even to the point where they think all the IMPORTANT work must be done by companies for money, and the more important it is, the less anyone will be willing to give it freely for the general betterment of all.
And that is why these business guys are a problem, even if they are obeying the letter of the rules. It's wrong to discount generosity and cooperation. And having examples out there of 'no, you can't have that, shame on you for wanting to deprive us of profit' is bad. They should just write proprietary operating systems if they want to go there- oops, no, they can't! They should take a hint from this.
No, a GPL application developer will still need this.
I write an application- get a bug report from a UL user.
I test it out- it works for me.
How do I know what thier binaries are? I won't use non-free software nor do I feel obligated to pay them for the use of my own work.
- Serge Wroclawski