Slashdot Mirror
United Linux is Here
pstreck writes "Red Hat watch out! Caldera, Conectiva, SuSE and Turbolinux have made good on their promise and United Linux is here! According to their website 'United Linux is a standards-based Linux operating system targeted at the business user. It is developed, marketed and sold by an experienced partnership of Linux companies.'"
I just don't get it I guess, it just seems like there are already so many standards.
I just don't get it I guess, it just seems like there are already so many standards.
That's the great thing about standards -- there's so many to choose from!
Recursive: Adj. See Recursive.
What standards ? Look at the website, they can't even decide on 1 language...
;-)
LINDON, Utah, PARAISO, Brazil, NUREMBERG, Germany, and BRISBANE, Calif. -May 30, 2002- Linux Industry leaders Caldera International, Inc. (Nasdaq: CALD), Conectiva S.A., SuSE Linux AG, and Turbolinux, Inc., today announced the organization of UnitedLinux, a new initiative that will streamline Linux development and certification around a global, uniform distribution of Linux designed for business. UnitedLinux addresses enterprise customers' need for a standard, business-focused Linux distribution that is certified to work across hardware and software platforms, accelerating the adoption of Linux in the enterprise. Under terms of the agreement, the four companies will collaborate on the development of one common core Linux operating environment, called UnitedLinux software. The four partners will each bundle value added products and services with the UnitedLinux operating system and the resulting offering will be marketed and sold by each of the four partners under their own brands.
Nearly every vendor supplying a piece of the technology infrastructure used by businesses has expressed support for UnitedLinux, including systems and software vendors AMD, Borland, Computer Associates, Fujitsu Siemens, Fujitsu Japan, Hewlett-Packard, IBM, Intel, NEC, and SAP. Independent hardware and software vendors spend considerable effort certifying their products and services on individual Linux distributions to ensure product compatibility for their customers. UnitedLinux will significantly diminish the number of distributions that vendors are asked to certify and will provide a true standards-based Linux operating environment.
Customers Benefit Through Unity
According to research firm IDC, a 2001 survey of 800 North American and Western European companies found that 40% of the respondents were either using or testing Linux in their organizations. UnitedLinux will help further speed enterprise adoption of Linux by providing businesses with a greater choice in the number of applications and hardware certified to work on the uniform version of Linux. Customers will also benefit from the global sales, localization, education, support and services that all four UnitedLinux vendors will collectively provide. The collaboration of the four leading Linux companies will result in an enterprise Linux offering, which is truly global by virtue of the companies' ability to provide local language support, training and professional services, in addition to the support of strategic partners. UnitedLinux will provide one unified Linux code base for IBM's complete eServer product line and AMD 32-bit and 64- bit platform and Intel's x86 32-bit and Itanium(tm) processor family platforms. UnitedLinux supports LSB, Li18nux, and GB18030 standards, as well as enabling installations in English, German, French, Italian, Japanese, Korean, Portuguese, Spanish, Simplified Chinese and Traditional Chinese languages.
In addition UnitedLinux unleashes a massive research and development organization for Linux in the enterprise. Effectively, the four companies involved in this process will shift dollars and resources once allocated to creating and maintaining custom Linux operating environments and divert them to new R&D on Linux enterprise software. UnitedLinux is dedicated to bolstering the enterprise readiness of the platform, but in the same collaborative spirit from which Linux was founded and continues to flourish.
Participation and Availability
While today's announcement outlines the founding members of UnitedLinux, the initiative is open for additional Linux companies to participate. The four partners currently plan to each offer their own server products based on UnitedLinux by the end of 2002. For additional information on UnitedLinux, contact Caldera, Conectiva, SuSE or Turbolinux or go to www.unitedlinux.com.
About UnitedLinux
UnitedLinux is a standards-based, worldwide Linux solution targeted at the business user and developed by Caldera, Conectiva, SuSE, and Turbolinux. Designed to be an enterprise-class, industry-standard Linux operating system, UnitedLinux provides a single stable, uniform platform for application development, certification, and deployment, and allows Linux vendors, Independent Software Vendors, Independent Hardware Vendors, and Original Equipment Makers to support a single high value Linux offering. For more information, go to UnitedLinux.
I've posted this before, and was modded down as a troll, but I'm glad someone else has noticed this too. Doesn't the fact that "there are already so many standards" imply that there is actually very little about Linux that is, in fact, standard?
Don't get me wrong, variety is good. I would be interested in seeing some good consumer flavors coming out of this kind of effort. Not only that, but this kind of organized, coordinated effort might even be enough to give Microsoft some fits in the desktop market.
"Ask not what your country can do for you." --John F. Kennedy
Might be worth a try. I use Mandrake Linux at work now, and I can do pretty much anything with it that Windows users can. The only frustrating thing is the lack of good browser plugins (yes, there is Codeweavers but it tends to be rather slow and I want it native.)
Ciryon
This will reduce the number of standards by four, and will (if done right) produce distributions that adhere to a single standard (LSB). This will have several benefits:
1) Greatly reduce dependency hell since all these distributions will be guaranteed to have a predefined set of libraries.
2) Greatly reduce the storage space consumed by hosting RPMs. Instead of needing storage space for four separate complete sets of RPMs, only one set of RPMs need to be hosted. These will be guaranteed to work on all four distributions.
3) Greatly reduce the effort needed to support multiple sets of RPMs. Now there will be four different companies that will be supporting the exact same binary code base. For those entities paying for support contracts, the support pool and support options will be quadrupled.
There are undoubtedly many more benefits to a standards compliant body of distributions.
VanguardLinux, SouthWestLinux, and QantasLinux!
A bunch of Linux companies making an uber-distro so they can compete with Red Hat should not be called United Linux. To me, it sounds a lot more like Divided Linux.
It will called Suse 9 "(powered by UnitedLinux)". See page 11 of this PDF.
From the FAQ:
"Caldera, Conectiva, SuSE, and Turbolinux will collaborate on the development of the UnitedLinux distribution in order to provide migration pathes from their former releases to UnitedLinux. However, each UnitedLinux partner will still have its own Linux distribution that is "Powered by UnitedLinux." Existing long-term relationships with leading hardware and software companies - as well as the current UnitedLinux partners - guarantee the compatibility of UnitedLinux with relevant business solutions. HW and SW manufacturers have the opportunity to join the alpha and beta test circles, thus reassuring in an early stage that UnitedLinux supports their products."
If I read this correctly, it means that the future versions of SuSe, Connectiva, etc will be forks of the main United Linux distro.
--Kylus
Idiot-proof something, and Life will build a better Idiot.
I think that this would make the market overall MORE fragmented for the following reason:
The 'market' for linux from a strategic perspective only consists of a few big players: Red Hat and ... Mandrake, SuSE .. ? WAIT!! KEEP READING!! SAVE YOUR FLAMETORCH!! Yes, there are hundreds of distros out there - we all know that. But these are highly niche level products and don't represent a real threat to the big profit-seeking boys. These market fragments, so to speak, don't represent significant competition to Red Hat. Do you think a corporation is going to license 10,000 copies of Tinfoil Hat Linux for thier workstations.
If some companies came together to form another big-boy player of the game where real corporate money is at stake, then the amount of real 'for-money' competition has INCREASED! Those niche players can come together and create a real presence which is worthy of competition. This would not create any 'unification' in the linux market as a whole. It just forms another sizeable fragment.
Otherwise the big boys don't really have as much competition and most of the market fragments are so small that they do not matter.
And this, my friends is a good thing. Just like you said, fragmentation is a sign of a healthy market. The market fragments would be big, visible and jarring against one another.
OK, distro makers: Are you ready to RUMBLE?! Let the best innovations win!
Hmm, if I read the presentation correctly, it seems that each Linux partner will still be pricing their product separately, and each have their own support teams. This seems to me to be a drawback that will be a big challenge to overcome, especially in the support area.
Hopefully there is good communication channels between the partners, so they don't step on each other's toes!
Attention all planets of the Solar Federation! We have assumed control! - Neil Peart
Item 9 in the FAQ states:
"Will users be able to download free versions of UnitedLinux for non-commercial uses, similar to how Linux is freely available today?
Yes, UnitedLinux sources will be made available for free download as soon as version 1 is released."
What on earth does this mean? How are they restricting the commercial use? Will one 'commercial' copy be sufficient for multiple installs, or will UnitedLinux bring in a per-seat or per-station licence free for commercial use?
I am not reassured.
In my opinion, so long as Redhat stays focused and continues catering to big business, I don't see them losing ground to this team.
I may be one of the few on this side, but I won't be trading in my Redhat CDs for United Linux for quite a while. I've been using Redhat for a couple years now, and for the most part, I'm a happy customer. It would take either a HUGE advance on someone elses part, or a big nose dive on Redhat's part to get me to switch.
If you actually took the time to check the website before asking the question, that is actually explained there:
Next Steps
* Each UnitedLinux partner continues to sell Linux under its familiar Linux brand and product, "Powered by UnitedLinux"
* Caldera OpenLinux "Powered by UnitedLinux"
* Conectiva Linux "Powered by UnitedLinux"
* SuSE Linux Enterprise Server "Powered by UnitedLinux"
* Turbolinux "Powered by UnitedLinux"
* One core development team benefiting several partners * Other Linux companies invited to join
Furthermore:
Competition
How will Linux companies in UnitedLinux still compete?
Pricing: Each company will set its own product pricing
Channels: Retail stores, reseller channels, direct, etc.
Support: Each company runs its own support team
Education: Independent training and certification
Professional Services: Custom implementations
Applications: Management, administration, messaging, etc.
OEM: Industry partners still choose products to bundle
Rather than yet another!!
Although in name they will all still exist. The will all be 'Powered by UnitedLinux' and have a couple of things tacked on. At the core they will be the same distribution. See page 11 of this PDF.
Why not get the /. community to create the 'standard' distribution?
T.
SoftLogic Solutions
http://www.softlogic.8m.com
you mean the "take cobalt linux we got from the acquisition and rename it to sun linux" solution, right?
Main components * Kernel 2.4.18 or higher * glibc 2.2.5 * gcc 3.1 * XFree86 4.2 * KDE 3.0 * Acrobat Reader From when has Acrobat Reader become this important? And to mention it in the same breath as the kernel and gcc is blasphemy ..ok i could accept emacs ;)
To a large number of people who aren't utter Linux fanatics and don't read /. or other Linux sites to actually understand why, Linux appears fragmented and inconsistent because of the sheer number of players. To remove 4 from the pool of 15 odd companies and come back with one united brand would help against Redhat and Microsoft.
Essentially saying "well you can go with Microsoft OR Redhat or United Linux". Rather than "you can go with Microsoft OR Redhat or Calerda or TurboLinux or Suse or ..."
Avantslash - View Slashdot cleanly on your mobile phone.
Standards That I accpet:
RMS will always attempt to rename any oS using GNU tools.
Linus will always be the head at Linux
RedHat Linux will always Rock!
Sun and IBM still will not get it!
Don't Tread on OpenSource
Hey,
I just wanted to congratulate them and to write them some wishes. When clicking on 'contact' on the unitedlinux-website, there are four adresses.
The mail to unitedlinux@suse.com seems to have reached the support: I recieved a support-ticket number. Hmmm...
The mail to unitedlinux@caldera.com was replied automatically too, but without subject or sender. It said "I will be out of the office traveling between May 28 - June 9. I will be checking e-mail[...]"
The mail to unitedlinux@turbolinux.com did not reach anybody: "Mail delivery failed: returning message to sender"
Should I laugh or should I cry...?
You forgot to mention:
Geography: each company already has more customers in some part of the world.
Unless this new distribution offers something significantly better than what Red Hat already does--and it looks to me like it doesn't--then this means nothing. A large part of the business world is still uneasy about migrating to Linux; those who do decide to use it will undoubtedly pick a well-established name like Red Hat over a consortium of distributions with much weaker market presence.
If you actually took the time to check the website before asking the question, that is actually explained there
Took the time. Couldn't wade past the marketing.
Given what you've quoted above, I predict that this effort will end up a minor footnote in the history of Linux that lasts for maybe a year and then disintegrates.
I think it should be called "Motherfucking Linux" and they should have Samuel L Jackson advertising it. I'd like to hear Mr Jackson say "Motherfucking Linux will whip Microsoft's ass". I'd pay good money for that.
It seems to me that UnitedLinux will look a lot like SuSE 8.0 ( very fine distribution):
Kernel 2.4.18 or higher
glibc 2.2.5
gcc 3.1
XFree86 4.2
KDE 3.0
Acrobat Reader
I'm happy to see some companies actually promoting standards and interoperability. I think that this is very very good for the future of Linux.
Nobox: Only simple products.
It's easy to take something, go your own way with it, then when it's sufficiently advanced and distinguished, call it your own standard.
The thing is, you can call it standard all you want, doesn't make it any more a true standard.
You need to build mindshare with all of your users, clients, etc., get some partners to help you along and support Your Way(tm).
That part looks good for these United Linux folks.
I still prefer The Debian Way, though, and I doubt they will be able to change that.
However, it will be good to have an alternative to Red Hat in the minds of the Common Folk.
Slashdotters have zero business sense.
.zip file that they download will install properly, but each windows variant may look and feel radically different.
United Linux is a fantastic idea. Ostensibly, the "UnitedLinux" tag would mean little more than "this distribution adheres to the LSB", but practically, it means much more.
When you download a package for any permutation of UnitedLinux, it will work. Period. No tricks, no shenanigans.
This means that if I'm using SuSE Linux 9.0 or TurboLinux 10.1, I can install any package labeled "UnitedLinux Compliant" without fear or hassle.
I am free to choose which ever distro fits my fancy, based upon its _features_, not whether or not it will run certain packages OK. This is a huge benefit. It means that the UnitedLinux vendors will have to compete with one another to build a better mousetrap, or risk losing out to another vendor in the coalition.
(Don't like the way SuSE works on the desktop? Swap it out with Conectiva. The packages you need to run will still work.)
This would be like Microsoft releasing the source to windows, and allowing every OEM to rebuild it, suited to their own taste.
Users could still be sure that any random
Same idea. A good idea. Promotes healthy competition.
I just don't get it I guess, it just seems like there are already so many standards.
What are you talking about?!
This will be the base and it will be LSB and Li18nux Compliant. Where do you get the "more standards" from? This will actually remove 3 and possibly more "standards"!
Instead of having to support Suse, TurboLinux, Caldera, Connectiva and possibly others, you'll now only have to support UnitedLinux. Basically, they have decided to work together in a number of areas to lower the development and distribution costs for both themselves and for others. When the next version of KDE arrives, they only need to make one set of binaries for people to download - not four or more!
I think this is great!
I find it interesting that people bitch and moan that Windows is a monopoly, but turn around and get uninterested because there's variants of Linux. This means two important things:
1) Linux is strong and has more support, especially since this will cause whole companies to rally their technology together.
2) This gives everyone more competition. If I remember correctly, SUSE or Mandrake was THE Linux distro to get. Red Hat came along and pushed the bar. Whether or not you like Red Hat or not, they have made a major impact in the Linux world.
Personally, if what they saw is true about making a business distro holds true, I want to see what they produce. Here is the simple truth, I was once a tech. I loved learning all the arcane commands and symbols, but I don't have time for that anymore. I need stuff that helps me work faster, better and *simpler*. That is what Windows *does* have in its favor currently. Most things are just a few dialog boxes away and I'm done. I'm waiting for that in Linux and I hope with a decree that they are going after business that they will realize that business isn't interested in the arcane. They want simple, fast solutions to common tasks.
My two cents,
Bel, the mostly sane.. "Of course I can't see anything! I'm standing on the shoulders of idiots." -- Me
Of course, with a Q4 release, who is to know? As they say, the proof is in the pudding.
I will be more interested in this discussion when the 1.1 version of this new combined distro is released. I've not actually used any of these distros for any period of time. I purchased Caldera back in 1999, but found it to be weak, so I went back to RedHat.
It will be difficult for these guys to break into the business market unless what they have is really overwhelming--and their support has to really ROCK! I admin over 100 servers using Redhat 7.[12] now, and its very smooth. I don't look forward to having to rewrite my admin scripts for a new distro, get used to a new way of doing things, etc.
DFossmeister
---
Think your webhost is fast?! Check out mine.
No Not Again! Its whats for dinner.
This just sounds like LSB with really good internationalization support. I might be glossing over some important things here, and if I am, please, someone tell me, but that sounds like that's all there is to it. Oh, yeah, and they enforce the use of KDE 3, which means I'm not interested, thanks. I guess I can see why they're picking just one desktop; it would seem to make sense, but I just can't stand KDE. ;) (could Kontrol-center get just a FEW MORE USELESS PREFS?!!?)
:) On the other side of things, there's United Linux, Mandrake, Lycoris. and Lindows... that pits some serious muscle against some serious muscle. While I'm rooting for GNOME, I'm excited no matter what the outcome, because it can only mean a better desktop for all users!
I'm really glad they're pushing for LSB compliance, but RH has promised they will be releasing a LSB 1.1 compliant distro this year. Since 7.3 isn't it, that means it'll have to be what will undoubtably be called RH 8.0 and will probably be released this Fall/early Winter, at least based on their past release patterns.
As an aside, the GNOME/KDE thing is about to get very interesting... GNOME 2 is like a couple weeks from release, and it's going to be the default desktop for Solaris, HP-UX, and (of course) Red Hat. All of these are major "enterprise" players. (I wouldn't be *too* surprised to see AIX follow suit.... any IBM people care to comment? Heh... CAN you?
The Free desktop that Just Works
Why is Acrobat Reader listed among the main components?!! First, AFAIK, Acrobat Reader is only "Free Beer"-software. Second, what makes it so importand that it has to be listed as a "main component"?! I don't get it...
My other account has a 3-digit UID.
The trouble with de facto standards is that no-one can be held accountable for not adhering to them. Debian uses a non-standard cramfs initial ramdisk by default, it doesn't include a gawk-awk softlink (or didn't, the last time I checked), yet Debian is one of the top three Linux distributions. I would never rate it that highly but I'm only a software developer. The users are the ones who decide which distribution wins and which one loses, and thus which de facto standards should stay and which should go.
The fact is, software developers can code to anyset of standards, so long as it sits still long enough. Moving targets are very hard to hit. If the distros sit still, the software devlopers will have an easier time but the act of sitting still might cost the distributions some users.
If the distributions would get together and write an API to let software developers figure out more easily where everything is, rather than expect the developers to customize their apps every time to make allowances for all those eccentricities, it would be enough, IMO. If Debian, Red Hat and SuSE would sit down at a table and hammer out the format of /etc/my-b0rken-distro.conf, I would have far less trouble on my hands on a daily basis.
This looks like a reduction of standards... something of a consolidation, even as it adds Yet Another Linux Distribution (YALD).
Notice the pattern of these four vendors. They nicely cover four continents, albeit Caldera covers N. America rather weakly, relative to the others in Asia (Turbo), Europe (SuSE), and S. America (Conectiva).
Geeky modern art T-shirts
Few question. How many of the current redhat/enterprise users are going to change the distro just because few competiting companies are now making up some standards ? Havent read anything about the case but.. what standards ? LSB ? Isnt redhat also supposed to follow that also ? blaah. This is just marketing hype...
yush
IMHO, it is mainly a move to save the costs of maintaining 4 different distributions. Let Connectiva do South America, SuSE Europe (they are big here!), Turbolinux Asia and Caldera North America.
Bye egghat.
-- "As a human being I claim the right to be widely inconsistent", John Peel
There's some points in there you may find interesting...
1. United Linux is based on LSB and LiN18ux standards. NOT on their own, like so many slashdotters seem to think. This is a very GOOD thing.
2. They will be offering business level certifications for "UnitedLinux" that will certify you on all of their distros for taking just the one set of certs.
3. They have the backing of a shitload of major players... IBM, INTEL, and AMD to name a few.
4. They are willing to accept any other company into this "United Linux" conglomerate who wants to join. Open doors are good.
5. This is not a single distro. Each vendor is selling their own versions of Linux as they see fit, but the rub is that all of these distros that are "Powered by UnitedLinux" will be compatible with eachother.
Basically, this is just an agreement by these companies to make their Linux distros interoperable. That's it. This is a big push towards standards, and it is EXACTLY what Linux has needed. I hope it takes off, and that they get RedHat on board as well.
Hell is being intelligent in a world full of idiots.
One good thing about this is that the Linux Standards base version 2.0 is now a lot easier. It only has to deal with Red Hat (and Mandrake which tries to be a more or less superset of Red Hat), United Linux, and Debian. This means that a lot more can be standardized.
Emphasis and addition mine.
I think you will find that outside of America the picture is quite different, with SuSE and Mandrake doing well in Europe and TurboLinux doing well in the far east.
Also interstingly enough, SuSE 8.0 was fully LSB compliant, the first commercial distro to be so.
My guess is that the base OS will be SuSE. Who's management tools they decide to go with is still anyone's guess. Maybe it'll be a whole new set entirely.
When last I checked, Q4 starts in October. It is now pretty much June and we are only 4 months from Q4 which is not even close to a year.
Ans when was the last time a company announced a product available 'Q4' and actually shipped it before the last week of December? In this business, Q1 means March/April, Q2 means June/July, Q3 means Sept/Oct, and Q4 means "God I hope we get it out the door by Christmas!"
Hey kids, there's only 5 days left 'til Yak Shaving Day!
What does it stand for?
Gnu is Not Unitedlinux
Name: Lame
Committee method of design: Lame
Idea that if it isn't prepackaged and retard-proof, it has no place in the business world: Lame.
Caldera is involved: Lame.
Is it june in your part of the world already?
it's in my head
So you're saying the fact that IBM allied itself with this little known company called microsoft didn't have any effect?
Arguably, that wasn't an alliance, that was a purchasing decision. If, however, you want to call it an alliance, I'll call it the exception that proves the rule.
The X Consortium is probably a better example of an exception, but note that it never led to the sort of innovation that X on Linux did - I mean, c'mon, CDE? Is there anything that screams "This is not your a desktop system" more than CDE?
In reality, I see this as something IBM might jump on to, and if they do this could be one of the best things to happen to linux since apache.
I'm sure IBM will endorse it. And RedHat. And if IBM starts shipping UnitedLinux, then RedHat will become LSB and Li18nux compliant and become UnitedLinux, at which point we're back to step 1.
This just seems to me like a whole lot of marketing around compliance to existing standards, and like any multi-company marketing event, it will sooner or later collapse for lack of substantiality.
I sorry to say but... I see the parent comment as another instance of someone not doing their research. This is not exactly another distro.
This is more a combining of already existing distros under one umbrella and the implementation of and adherance to the LSB Standard.
In short... This is a GOOD THING(tm).
Competing with RedHat can actually be easy once it becomes known that the new consortium/distro follows a set standard - LSB. The standard means that software written to the LSB should ALWAYS run with no problems on a compliant distro.
Codifex Maximus ~ In search of... a shorter sig.
3. They have the backing of a shitload of major players... IBM, INTEL, and AMD to name a few
Quoth unitedlinux.com: "Majority of enterprise system and software vendors including AMD, [...] , Progress Software, and SAP, support effort to create standard Linux platform".
I wouldn't call this "backing". Backing implies a commitment of money or resources which is not mentioned AFAIK.
Let's hope they do a better job of being open than Caldera or SuSE did.
Notice on their FAQ there's a question that says: "Will users be able to download free versions of UnitedLinux for non-commercial uses, similar to how Linux is freely available today?" And of course the answer is yes.
What bugs me here is their implicit use of the phrase "for non-commercial uses". As far as I know, "how Linux is freely available today" is free as in Free. You can use the distros for any purpose you see fit, including making millions, for free. You only pay if you want fancy CDs, manuals, support contracts, tech support, etc.
Is this just a bad choice of wording, or does this mean they'll try to impose some licensing or distribution scheme aimed at making "free" only apply to non-commercial use?
11*43+456^2
The US may think of themselves as the center of the world, but sometimes that just boils down to ignorance (or arrogance ;-)...
What you perceive as arrogance or ignorance is neither, usually. Many times people in the U.S. fail to take a world view of things like this because it doesn't matter. SuSE may be very, very popular in Europe - but it's a bit player here. It really doesn't matter one damn bit to someone in St. Louis or Seattle how popular a brand is in Berlin or Tokyo if it's marginal here.
The same goes for Windows vs Linux. Unfortunately the U.S. is far behind other contries in adopting Linux. The unfortunate reality in the U.S. is that if you're looking for a job or planning an IT strategy Microsoft products are going to be a large part of that compared to a business in Europe that might be able to ignore Microsoft completely. That doesn't mean someone's ignorant of the situation in Europe - it simply doesn't matter to someone in the U.S. because it's not the situation here. Unless it applies to you, it's basically just useless abstract knowledge.
I'm not saying that the average person in the U.S. couldn't stand to know a bit more about the rest of the world, but you're just pursuing a personal bias here.
RedHat will become LSB and Li18nux compliant
I don't know about Li18nux, but Red Hat has already said they will release an LSB-compliant distro this year.
The Free desktop that Just Works
Turbolinux pretty much owns the Asian markets with the minor exception of Hancom in Korea.
When an ISV has to port a software package to Linux, it has to port it so it'll run on multiple different platforms. As much as most of the folks here want to say "just port to Red Hat, that's what everyone uses", the fact is porting to Red Hat locks out a large chunk of the potential market for software. Porting to all distros doesn't make sense either, though.
Try as you may to state otherwise, the fact remains that glibc is glibc, libm is libm, the kernel is the kernel, etc. Red Hat, Caldera, et al differentiate themselves using little chunks of code like package managers and installers, but when the system is installed it's all just Linux. The idea that an ISV has to port something to lots of different flavors of the same OS is silly. It's not like the old days of UNIX where the different flavors actually WERE different. Linux systems should be able to run Linux apps without going through ridiculous gyrations with filesystem heirarchies, RPM versions, etc.
Currently, ISVs can't rely on the base system of all Linux distros being consistent. An effort like United Linux offers a predictable, consistent base on which ISVs can port their software.
UL won't take over the entire Linux market, but it should. There's absolutely no reason companies like Red Hat, Sun, and HP can't use UL as the base for its distro (remember, Red Hat doesn't shine because it uses a special Red Hat glibc; it's the installer, the package management, and the branding that Red Hat is known for, not the base system). With the rising cost of creating, maintaining, and supporting these Linux distros that so many of you love to get for free, eventually companies like Red Hat, Sun, and HP will need to consider cutting what is literally a pointless duplicated effort.
All ISVs and OEMs care about from a technical standpoint is the base system and its ability to run apps or run on an OEM's hardware. That's it. The marketing folks love brand names, but what's under the hood is spooky magic to them so they don't care. If all Linux distros (at least RPM based distros) adopted UL as its base, Linux may actually have a chance to take over the IT world.
In my view, the folks who DON'T adopt UL are the ones that will be fragmenting Linux. No single point of failure, support organizations worldwide... UL just makes sense.
cat > /etc/motd
F1rst P0st!
^D
So you're saying that Debian doesn't follow "standards" (presumably meaning "whatever RedHat decided to do this week"), and that you don't like moving targets?
;-)
Debian can hardly be described as a moving target , given it's 18-month release cycle.
If you want to measure consitency of behaviour over time, I think you'll find that Debian would win hands down.
If you want to measure some sort of "least surprise quotient" when a random *nix user comes across a distro for the first time, I think Debian would win again.
As for the awk link, mawk has been providing such a link since Mar 1997, and gawk since before Dec 1995, so I don't know when you last looked, but perhaps you should look again.
Debian: GNU/Linux done the Linux way
I've just read the United Linux FAQ.
It's obviously written by a marketing person who hasn't read the Cluetrain Manifesto. The answers all read like ``United Linux is wonderful, the sun shines out of it's arse''.
There is no discussion of questions that no doubt will be frequently answered, such as:
I agree with you completely but I have a few things to add:
I have had fewer problems with RedHat and hardware configuration than most other distributions (try using YAST 1 or 2 to configure two different ethernet cards from different manufacturers in SuSE 7.0 and 7.1).
That being said, I think that the joint expertise of these four distributors will be immensely important. Initially, it may not be that important in the US, but TurboLinux is big on the Pacific Rim, SuSE is strong in Europe, and Conectiva is strong in South America. This initiative may enable these companies to cut R and D costs somewhat and still offer a much stronger product, which can compete with Red Hat.
This is the thing I like about the Linux communtity-- when the competition gets tough, the community comes together. A community can always beat a 300 lb gorilla, and it will, be it Microsoft, Red Hat, or whoever.
But I also agree with you-- Red Hat has been immensely poisitive in the industry, but they are smart and will survive, even if it means marketing and assisting United Linux.
Will United Linux be the Linux defragmentation tool?
LedgerSMB: Open source Accounting/ERP
Anybody remember the OSF (Open Software Foundation/Oppose Sun Forever), the anti-Sun proprietary unix alliance? Interesting parallels...
It was pretty much marginal, except for killing openlook and NeWS in favor of Motif and X (is that a good thing?).
I'm constantly surprised by how few people who have heard of AIX have heard of the Linux compatibility that IBM has been stressing on ever since AIX 4.3.3 . Their repackaging of the Linux applications is referred to as the AIX Toolbox for (sic) Linux. Indeed, AIX v5 is called AIX5L, and the 'L' in it is to emphasize Linux compatibility. 'rpm' runs natively on AIX5L. So in that sense, IBM has already determined that AIX is going the Redhat way.
Of course, this is a packaging choice - rather than a full fledged subscription to whatever standards that Redhat follows. But it does make it more difficult for IBM to consider any standards that conflict with Redhats standards.
There is no such thing as luck. Luck is nothing but an absence of bad luck.
yes bash for a shell and *.so in the right place can make or break it but the real nice thing is GCC
they have put up the fact they will use 3.1 guess whats in redhat 8.0 (-;
and what will be out quicker redhat 8.0 with all the same features and compilent with LSB or United Linux ?
now if people would sort out the BSD's and Debian to use gcc 3.1 then we would be rocking.....
regards
john 'I dont care about gcc for the PDP' jones
UnitedLinux seems to finger the enterprise for its customers but fails to clarify which markets.
Is this effort targeted at the server market or the desktop market?
Since, they seem to finger Redhat I would assume the server markets?
NexuSys - Linux support by the best
I assume you're looking at the main components on the release plans page. I don't know that UnitedLinux enforces the use of KDE, so much as ensures its availability. Nonetheless, I think it's premature to standardize on a desktop environment for Linux. While the GPL release of Qt was terrific, GTK+ is released under the more business-friendly LGPL. Developers of non-GPL Qt products (e.g., Opera and Kylix) have to pay a hefty fee to TrollTech.
More to the point, why in the world is Acrobat Reader listed as a major component of UnitedLinux? It's a proprietary product of a DMCA-wielding company. Perhaps more egregiously, it's ugly. The recently released 5.05 for Linux (with no search support, mind you) has no place on a KDE desktop.
Hmm. Maybe that's because redhat has traditionally enabled services rather than diable them. It also installs too much bloat requiring
more to be locked down.The install is just too coarse - you don't have the fine-grained control over installs that's really needed.
I have been a long-time RH user and still use it (including 7.3) on a couple of my machines.
HOWEVER, I'm a recent debian convert. Debian rocks. Apt rocks. up2date is lame. Redhat is going out the door.
I don't know what UL is planning to use for package management, but I sure hope they consider apt. It's the best system linux has. "apt-get dist-upgrade" is just awesome, and you don't even need a super high-speed net connection to do it.
This will not work. When two people sell exactly the same product (think lettuce) it becomes a commodity and the margins fall to nearly zero. So manufacturers have a mandate to add distingushing features so that they sell a non-commodity.
It was tried many times with posix, open88 and other group of standards put together by two-bit players...
"does anyone find it interesting..."
Not really. They have to have a default desktop (like Red Hat has Gnome), and they chose KDE (propably because Conectiva and SuSE (and Caldera?) are KDE-centric distros.
Of course, you can still run Gnome on United Linux
Lesbian Nazi Hookers Abducted by UFOs and Forced Into Weight Loss Programs - -all next week on Town Talk.
How long until the strong headed geeks that make up these companies disagree and go their separate ways. I think this will be more along the lines of dropping four Siamese Fighting Fish into the same bowl.
In order for them to have any hope they desperately need a tie breaker. "The vote is 2 for and 2 against" just does not work.
Get a free ipod.
If one thinks of turning Linux from a niche product to a widely accepted system, one must not forget the code developers. If it's hard to find good, trained, sysadmins for Linux, when it comes to developers the situation is even worse. For a Linux company, the way to go is to convert Microsoft developers to Linux, and that is much easier to do with KDE than with Gnome.
What turned me definitely to KDE was the API. I was used to MFC, so the C++ toolkits in Qt and KDE seemed natural to me, being at the same time simpler and more powerful than MFC.
On the other hand, Gtk reminded me a lot of the Motif hell one had to go through when coding GUIs for VAXen and Unices. I don't know how it has evolved lately, but the last time I tried to create something using Gtk there wasn't even an easy way to write inclined lines of text. My first try at a GUI program in Linux was plotting a xy graph, and in Gtk I couldn't find any obvious way to write the vertical axis label. Also there wasn't any good IDE for Gnome, while Kdevelop compares favorably with Visual C++.
Anyway, for my taste, KDE is better than Gnome as a final user as well. For instance, every time I have to open a file in Gimp I realize how much better the KDE file open dialog is.
look at the homepage. There are positive comments from practically every MS competitor/group that has a legitimate gripe with MS.
They all appear to see this as "The Way" to finally 1) shirk off the *need* for MS with their business customers because, hey, we have a "Standardized" Linux here... (insert finger tapping PHB's looking impressed) 2) do this Linux thing easier!.. multitudes of distros, RPMs, GUIs, yada, yada, yada.... smeg that. 1 standard distro with infinite possiblities! (if you don't LIKE KDE, you can ditch it... try "ditching" major portions of Windows)
this is a positive move for everyone.
Simplicity that costs a ton of money and gives you no options and sends back your information to MS without your knowledge is a bad thing.. its Windows.
Simplicity is a good thing, not a bad thing.. its Mac OS.
Simplicity with modifyability is not a good thing, its a kick-ass thing... its a standardized Linux that can be changed to suit your needs OR can be left alone to be simple.
How is this not a good thing? I don't get it.
guns kill people like spoons make Rosie O'Donnell fat.
There are de-facto standards which are imposed more or less by a single company (such as RedHat), and there are standards which are developed by independent organizations (such as the Free Standards Group). I think an open (in every sense) operating system such as Linux deserves a standard that has grown dynamically out of the collaboration of a multitude of people, from companies as well as from independent organizations. If a company like RedHat is powerful enough to impose de-facto standards on the Linux community, this might hurt the evolution of Linux in the long run.
While basically all major Linux distributions contribute to the Linux Standard Base, not all of them follow the specifications equally well. The general opinion seems to be that RedHat in particular does not implement the LSB very strictly. One goal of the United Linux project, as far as I understand, is to create a distribution that follows the LSB very closely, and has enough market share to compete with RedHat. So instead of imposing yet another standard, one should see the project as an attempt to strengthen an already existing standard.
If you look at the main components, it could very well describe Red Hat 8.0, except for Acrobat Reader. Red Hat 7.3 already ships with KDE 3.0. Red Hat 8.0 is supposed to be LSB compliant.
It depends on the conditions of joining. To advertise "Powered by UnitedLinux," does your installer have to force the installation of KDE? This could be a deal breaker. Do you really have to ship Acrobat Reader? This would definitely be a deal breaker.
I have been trialing GCC 3.1 and the C++ compilation is a major improvement over past version of GCC. Being C based the GNOME 1.4 and GNOME 2 libraries and most applications compiles and runs without too much hassle. However KDE 3.0.1 is somewhat more problematic, even when neither debugging support nor strict syntax checking is enabled..
This is not the fault of either the KDE or GCC developers. KDE was coded to support the "older" C++ style of pre GCC 2.9x and Microsoft's compilers and the GCC Team is following the new C99 & ISO 14882 C++ standards.
After kludgeing around the defects in the older GCC C++ template and library implementations, GCC 3.1 C++ is real joy to use. It makes it possible to program C++ in a completely new styles, that IMO can be far more productive.
It is difficult layering one type of programing style over another, the older C++ style libraries certainly make Windows programing a pain.
Would it not be better to wait for the KDE team to port KDE to a pure GCC 3.1/ISO 14882 style?
At the very least the debugging support is required for GCC's Profile Driven Optimizations which can greatly improve application performance.
GNOME 2.0 is due for release soon enough, at the very least the GNOME libraries and core should also be included at a United Linux "main component".
Actually according to the white paper, there will be RedHat compatible libs/symlinks ,so in theory 1 RPM will work on 5 distros. I'm not sure I like this however, as it'd effectively make RedHat the "standard" to which RPMs would be targeted, as then you know it'll work natively on redhat, and with the compatability mode on UL systems.
Each of these companies has done a lot of interesting stuff. I want to know which parts of each distro is going to go into United Linux. My biggest gripes with RedHat is lack of XFS support and painful upgrade routes. I think United Linux will have a chance in hell if...
1. They adopt apt-rpm as a layer in their installation and upgrade process.
2. They include advanced features like XFS and ACLs from the base installer.
3. They keep YAST.
4. They support up and coming platforms, like ia64.
5. They make it easy for third parties to add proprietary features to their distro.
Reading between the lines of the poster and from the various comments i see alot of people favoring RedHat. Nothing wrong with this except it's also evident that they are as fervently pro RH as they are pro linux. Instead of Linux versus Windows (which isn't correct as well IMHO) it's Red Hat against all the others..
For alot of people in the US RedHat may be synonimous with Linux, it isn't the other way around for the rest of us. RH has a strong base in the US but take a look to the rest of the world will you? There is Red Flag linux in China, Mandrake and SuSE are pretty popular in Europe where RedHat isn't an obvious choice.
In fact, globally looking this could be what Linux was holding back. Like several others have pointed out, it's a pain to write for Linux because you have to take into account all the various distro's with there slightly different way of placing conf files etc. United Linux could be changing all this. An ISV can now just write the program once and that's it. No more extra development time to see if the app will run on both RH or Suse. No more pain in the ass to see which lib version is used by Mandrake istead of the one used by Caldera. The list could be as long as you want.
The fact that RH isn't in the list of participating distro's doesn't mean they couldn't be a part of this, they can join whenever they want and without any problem. This is not meant to compete with RH. This is meant to UNITE all distro's to comply to a standard base to give linux a big boost into corporate minds. Why is Linux still marginally used in office's? Not because it's lack of power and stability. It's the apps department. Why aren't ISV's working on linux versions? Because they don't know which distribution to target. RH, SuSE, Caldera, Mandrake and Debian are all too diferent from each other to just port an app once and then be done with it. You have to write different versions. This costs developerstime and thus money which they rather spend at doing something that will give them a steady revenue without too much hassle..
Please do not start bashing other distro's because they do something different then the one you favor. You should instead be focussing on learning why they do this and if what they do will help Linux or not. See more then just your personal feelings or needs. try to see the big picture because before you know it Linux wil be dismissed as another platform that's not viable to be used in corporate environments. And if you don't want it there perhaps try to look for a new os to toy with then..
Read the whitepaper - GNOME 2 support is in there, and in the PDF is says actually only KDE Base is mandantory. Some of the details seem a little inconsistent right now, I think they're still figuring out exactly what it'll be.
SuSE, Caldera, and TurboLinux also use RPM. Connectiva uses APT over RPM. I don't see anything in the UnitedLinux pages that specifies a package format, but I'd be surprised if it's not RPM.
Please chew your food before you begin typing next time.
Moon Macrosystems. Sun's biggest competitor.
Actually, distributions also differentiate themselves by adding patches to things like glibc, GCC, and the kernel.
As I mentioned in "Red Hat's little forks," there are over 100 patches in kernel-2.4.18-4.src.rpm, including a 20 MB whopper from Alan Cox. As I recall, SuSE incorporated ReiserFS, JFS, and LVM before they were in the Linus kernel.
Wearing your optimistic programmer hat, it should still just work. Wearing the pessimistic hat of a user or a tester, it has to be retested. It will be interesting to see the extent to which a "Powered by UnitedLinux" distribution is allowed to add patches.
- a kernel recompile
- a buncha utilites (at least two different packages)
- a specialized login script somewhere
- changes to the GUI login
This shouldn't be hard, but (long story short) it's never worked for me.I could really use a more automated, standardized setup for this that would be suited for an office-wide implementation. Especially desired would be a way to send Linux-useable login scripts from Novell NDS Administrator. I wish Novell would lift a helpful finger too. It may be pragmatic to support hardly anything outside Windows, but I would imagine their experiences with Microsoft would drive them to support other desktop options.
This feature would earn UnitedLinux my heart. Currently I'm Red Hat at the office, Mandrake & SlackWare (tho I'm not smart enough for SlackWare) at home.
Of course a stripped-down internet appliance is easier to make under Linux. Linux can be stripped to its bare core and manipulated in every which way possible. Hell, they have distros that can be installed, and ran, on a single 3.5" floppy. So yes, on that front, you are absolutely 100% in the know.
The problem comes when you need more than a stripped down internet appliance. When you need a desktop, a true environment that allows you to run office apps, play games, launch and easily install applications, and also be a server if need be.
Linux can do all those things, but to do them, you normally have two or three different GUIs and/or interfaces that do the same thing. Where was the sound volume located? Where can I set my screen saver settings? Why is it so hard to choose a new resolution? I'm not talking about the intermediate Linux user here, I'm talking about Grandma Thelma who just bought her machine at Wal-Mart and wants to install Pro Knitting Tournament, the latest in Grandma fun.
Lets keep on this little rant and just assume that there is a Pro Knitting Tournament game for Linux that Grandma Thelma wants to install. After she glances at the directions and does what she is told, an error comes up saying Grandma needs a new version of perl. The troubleshooting section says only to go to CPAN and find it. On her way there, assuming that she somehow magically made her internet connection work, she gets lost, finds a link, downloads something but isn't sure what, where it is, how big it was, or what she's supposed to do now. Now she's totally frustrated, her joints ache, and her teeth hurt. She turns the machine off never to bother again, the darned old thing too frustrating to mess with. Another Linux user lost thanks to the thousands of window managers, package managers, source distrobutions, and bash script installs.
This is the first step in fixing this problem. Maybe it won't happen this year, or even the next, but soon graphics will get prettier, GUIs will be more efficient and easier to find (ie, you won't have three different "System/Settings" menus in your "Start Button" (or whatever you want to call it), and packages will be easily installed with great documentation and links will be created on the users desktop as well as their application menu (ie, Start Button).
But that is the future, and this is the start. The journey of a thousand miles has begun.
That's not true at all. The standard most businesses are still choosing and developing for is not Redhat ... it's Windows. Yes, I know you were really speaking in terms of the Linux context. But it's still the same problem. Consider an extreme scenario where Linux manages to oust Windows to say, less than 10% of the market. What will we have then? Linux? No. It will be Red Hat. And if it isn't sanctioned by Red Hat, it won't go anywhere. And in time, someone at Red Hat will become greedy. It may not be the people who are there today, but if that scenario ever happens, you can be sure some very greedy people will be working to take control of Red Hat.
The real problem is not which distribution of Linux is more powerful. No, the real problem is that a distribution of Linux is more powerful. And business in its infinite stupidity is demanding just that. Software developers want just ONE distribution because they are too lame to develop something like portable code that can be compiled on more than one platform.
The whole principle is about choice. While I don't expect the developers of an e-commerce web database system to port it to an embedded Linux setup, I do think it is reasonable to expect them to port it to at least a few different systems so I can have a choice. The problem today is too much software is Windows-only. And given the greater difference between Windows and Unix, I can see how it is hard to cross-port. But when people start talking the very same talk, about having ONE distribution of not just Unix, but of Linux, then that tells me the real problem is not the true difficulty of porting software, or the difference between Windows and Unix ... but of the lameness of commercial applications developers.
now we need to go OSS in diesel cars
I'll bet it works great with IE, though.
In Windows it often takes more than one dialog box just to find what you want to do, or if you know where it is already because you have done it a lot, its still more than one to get there. I work faster, better, and simpler, by just executing the command I need to have done without having to spend the time opening up lame dialog boxes. And if I need to do the same thing in a hundred different contexts, instead of having to open one hundred, or twoo hundred, or three hundred dialog boxes, I can drive my one command in all those hundred contexts by wrapping it in a while command in the shell ... still interactively. I find that speeding up business really works better by automating with smarter computing, not by flashier prettier computing.
now we need to go OSS in diesel cars
If your RPMs can't work on all RPM based distributions (assuming the distribution isn't broken ... and I assume these distributions are not) then something is wrong with your RPM. If it doesn't work on some specific distribution, then why not? Is the blame because that distribution did something specifically wrong? Or is it because your library function calls are depending on undefined semantics seen only in a specific version of the library?
Join The United Front Opposing Lame Binary Packages!
now we need to go OSS in diesel cars
I had the strangest premontion yesterday that something like this was going to happen.
What's in a Sig?
Why would you be placing the trust of your business into the software of a company that doesn't have the competency to make it work on more than a couple Linux distributions?
If OEMs built different versions of Windows from the source, you can bet that application software developers would find a reason to not support them all. They would end up supporting just one or two flavors of Windows. Hell, they even did that anyway with versions that ran on 98 but not on NT, or visa versa.
I think we need to take a closer look at why some programmers are making lame programs than only work in certain distributions. My bet is they jump to conclusions about how things work instead of trying to understand how things really do work across the scope of several distributions and system types. I've seen some very lame programs that run on Red Hat and not on Mandrake, even when the programmer tries. And I've also seen some programs that were written entirely on Windows and recompiled just fine on Unix and ran not only correctly, but actually faster, too.
If there's a standard we really need today, it's a standard that says no lame programs, and no lame programmers. If your code doesn't work on at least a few systems, you did it wrong.
now we need to go OSS in diesel cars
Why the hell do they need certification if this is going to be based on standards? Why the hell do they need certification if their systems are supposed to work well? Sounds more like a scam to extract more ca$h out of people, since most businesses don't pay extra for certification of any kind (except CCIE).
now we need to go OSS in diesel cars
One core development team? Now I see the scam here. The whole idea is to allow these companies to fire more developers and increase the profits for the greedy.
now we need to go OSS in diesel cars
Some of us may remember the fuss when Corel put their beta out in binary-only form. Am I the only one to notice that UnitedLinux's faq says they will only release source with version 1.0, almost 6 months after the first alpha releases?
I'm glad to see that crucial pillar of an operating system, Acrobat Reader getting a mention right next to the kernel, gcc, XFree86, etc.
Main components:
* Kernel 2.4.18 or higher
* glibc 2.2.5
* gcc 3.1
* XFree86 4.2
* KDE 3.0
* Acrobat Reader
Lol...
Get your own free personal location tracker
From their FAQ:
Will users be able to download free versions of UnitedLinux for non-commercial uses, similar to how Linux is freely available today?
Yes, UnitedLinux sources will be made available for free download as soon as version 1 is released.
I think there is danger in this message, something that came from the dark side of the force:
Download free versions: means that will be some non-free version
For non-comercial: means that comercial user will have to pay (looks like M$ stuff)
Similar to how linux is available: means that will not be like linux, but similar, no GPL? not free software? is like M$ closed stuff?
So, people, as this messages continue on their WEB SITE, as they are not GPL and because they want to earn money over their copy's, let's stick with red hat, mandrake and debian!
let them burn!!
I'm sorry. I thought this was slashdot...
It's not the power of XP that is putting it in the majority of businesses. It can't be, because there is less power in XP. The answer is, it's the hype. That's right, Hype. It's HYPE that drives most of the businesses decisions, and the corporate executives don't even know that the marketing people (in other companies) are the ones who are really driving their decisions through brainwashing hype. Microsoft happens to be the most successful at this.
now we need to go OSS in diesel cars
Likewise, "Hey, look at the new Frogon Linux", and the other guy says "No, the Grubar Linux is better." How could they still act as separate, competing companies if they all just called what they sold a "Linux" system?
No, they call it Red Hat Linux, Slackware Linux, Mandrake Linux, and now SuSE United Linux. They simply tack their name onto the front of Torvalds's trademark
My understanding is that you are suggesting that they drop the "Caldera" or "SuSE" and just call "the" product "United Linux"--is that right? My objection to that idea is that the companies, which do not exist to improve the public's impression of Linux, but to make money for themselves, will lose their brand identity if they drop their name from the product they sell.
In order to make things better for the non-dominant companies, they all got together and agreed to a standard. Presumably you will be able to get a software package that will run on any "Unified Linux" system. That gives big vendors like Oracle one target. If they have to chose between fifteen targets, they will say "Ok, RedHat it is.". If they can say, "Hey, if we also support United Linux, we'll be able to hit all of SuSE, Caldera, ans whoever else's installed base as well", they might decide that's worth doing.
I doubt very much that any of these companies are doing this to try to make the user perceive Linux as being less fragmented. They don't want people to lose track of the idea that if you buy Caldera's distro you get something different than when you buy SuSE. They are very interested in keeping that distinction there, and convincing people that Caldera is better than SuSE. But they also want a single target for a software vendor. They get that with "United Linux". If they don't call it "SuSE United Linux" they lose their brand. If they just call it "SuSE Linux" there is no indication that a package for "SuSE Linux" will also work on "Caldera Linux".
"SuSE United Linux" and "Caldera United Linux" indicate both their crucial similarities and their crucial (at least in the companies' minda) differences.
(Just don't mention "SuSE United GNU/Linux" or you'll get modded down as flamebait (Yes, really!) lol)
Liberty uber alles.
It's like a bunch of nations getting together to fight terrorism using cowboy tactics. There will be a lot of fire and show, and we'll still have terrorists.
now we need to go OSS in diesel cars
That's what 'de facto' means, or didn't you get the memo? ;-) De facto standards come from the most popular Linux distributions.
/usr/bin/gawk that points at /usr/bin/awk, which may in fact be a link to /usr/bin/mawk, then you're wrong.
/usr/bin/gawk because you are relying on features unique to gawk, then you would be disappointed to find that you were actually running mawk, true?
/usr/bin/awk.
:-)
[aside: Debian is one of the most popular distros --- does that mean that Debian is the de facto standard?]
So, let's take something that you presumably would have considered to be a de facto standard at the time. Linuxconf was the standard configuration tool on RedHat for a while. So, do RedHat get to change that "standard" just because they originated it, or were they non-standard then, or are they non-standard now?
I'm not criticising RedHat specifically, since the other commercial distros do similar things, but
if they change their defaults over time, then is the de facto standard limited to the logo on the startup screen?
Debian on the other hand does not have the organisational structure to allow it to chop & change rapidly, so any change ends up having to take account of all the people that will choose not to follow the change, resulting in a more robust distribution.
Chaos brings forth order.
Consistently difficult to install. Consistently Balkanized. Great. No wonder Debian is so popular. It is the Archie Bunker of Linux distributions: you can always rely on it to behave predictably.
Well, thanks for that ringing endorsement. I didn't ask you to use it did I?
Perhaps you should see the dozens of e-mails I received from Debian users reporting the "'which gawk' returned an error" message from my software. Perhaps you should tell those users that gawk was indeed there when each one specifically told me it wasn't. It wasn't there when I installed Debian in late 2001
Ah, sorry, you're complaining that gawk wasn't present when awk was? I read you to mean the reverse.
If you want gawk on Debian you can run:
apt-get install gawk
but you have the freedom not to, and there are other alternative implementations (i.e. mawk) so there's no need to do so.
If you are saying that someone should provide a link
If you are invoking
If on the other hand you are not using gawk specific features, you should be calling
Either way the problem you describe is not a fault in Debian, it is either a bug in your code, because you were saying gawk when you meant awk, or it's a failure in your package dependencies (if you packaged it) or in you readme, which should have indicated that your program depended on gawk.
I notice you didn't bother tackling me on cramfs.
You were saying that Debian was non-standard --- I don't think that what a particular version of a particular other distribution is doing counts as standard, so I didn't bother considering it. I don't like initrd/cramfs anyway --- it's just another thing to go wrong at boot time, so I don't use it on any of my machines. We use it on the install floppies/CDs to make things fit, but I don't see that that's something to get overly stressed about, and why anyone should care, I don't know.
If you're so excited about this de facto standard of yours, what do you think of using file system labels in fstab, instead of devices? The majority of RedHat systems that are running don't use that scheme, so are RedHat breaking the standard by introducing change, or are all historical RedHat installs now non-standard? How about Grub vs. Lilo? Do you want me to go on?
P.S. Read your Gun control post. --- glad we can agree on something, anyway
Debian: GNU/Linux done the Linux way
That's the great thing about Linux standards. There are so many to choose from.
=brian
This to me, sounds like a good thing. Not all Distros will be the same. SUSE will still be different from turbolinux etc.
What it does mean is that each distro will have things in common. Things that companies can count on like standard libraries, a standard print system etc. It does not stop distros from putting in extra libraries and print systems just guarantees that they will also have the standard ones and, I assume, a easy way for people to know when they install the distro that they have picked the standard features.
One of the things that can really frustrate a developer is to spend a lot of time developing a software product only to find that he has to keep half a dozen versions in order to facilitate the majority of Linux users. It can be a real maintenance nightmare.
I hope that companies like Redhat and Mandrake join in supporting the standard.
The race isn't always to the swift... but that's the way to bet!
....is that there are so many to choose from. ;-)
sigs are a waste of space
Considering how Taco categorized the announcement as being "...from the who-really-cares department", no, I don't think he's joking. It seems he doesn't see a need for UL.
If he does see a need, his categorization of this announcement as well as his jab at the end will be that much more confusing.
OK, I can see that the gawk thing pissed you off, but it's actually pretty simple to either make your package depend on gawk (if you package it) or put a note in the README saying "This package depends on gawk" if that's the case.
You were saying that Debian should have provided a symlink from awk to gawk though, which is just plain wrong I'm afraid. If the user chose not to install gawk, then you don't get gawk. End of story.
Were you really using gawk specific features? or would awk have done the job for you? If the latter, you should have been invoking it as awk.
Having said all that, I'd imagine that it's pretty easy to miss that subtlety, in which case it would be pretty bloody frustrating, so I do understand where you're coming from. Just a shame nobody bothered to point this out to you earlier really.
Cheers, Phil.
Debian: GNU/Linux done the Linux way
Well, if RH is compliant that's good. I'm all for it! I use RedHat myself but was unaware that it was compliant.
Thx for the info.
Codifex Maximus ~ In search of... a shorter sig.