Slashdot Mirror
Latest Linux Standards Base Gets Vendor Support
Neopallium writes to tell us that in a recent announcement at the Desktop Linux Summit the Free Standards Group reports fourteen of the leading Linux vendors have pledged support for the newest release of the Linux Standards Base. From the article: "'The Release of LSB 3.1 is another milestone achieved by the industry and the Open Source Community that delivers ever increasing value to customers,' said Reza Rooholamini, director of enterprise solutions engineering at Dell. 'It enables further uniformity and standardization across applications and distributions that allows quicker deployment of Linux solutions with higher levels of quality.'"
You may remember me, I am old friend. Please don't be a stranger.
Sincerely,
Mr. Comma
"LSB is not to be taken internally lest you wear the robes of Richard Stallman."
"Made up/misattributed quote that makes me look smart. I am on
Soooo... 3.1. The first usable version of Windows was 3.1 also. Coincidence?!
Maybe this WILL be the year Linux arrives on the desktop!
"...Well, there's egg and bacon; egg sausage and bacon; egg and spam; egg bacon and spam; egg bacon sausage and spam..."
Because it has nothing to do with desktops, per se. It is a specification for directory layouts, config files, and required libraries, and its purpose is to make sure that applications that compile on one system that complies with the LSB will compile on all systems that comply with it.
- None can love freedom heartily, but good men; the rest love not freedom, but license. -- John Milton
No matter what the roadmap from the EDA Consortium says, too freaking many of the tools I use at $WORK refuse to run on anything other than Red Hat 7.2 (I kid you not!)
And, yes, they actually check /etc/redhat-release
Lacking <sarcasm> tags,
It's the Linux app ISV's who need to write their apps to this "standard".
So far, that isn't happening.
Nah, it won't be usable until Linux for Workgroups 3.11
Maturity is a double-edged sword. We get all the benefits of uniformity along with all its problems. Let's hope that Free Software can adapt good defenses whilst still providing better app support and uniformity.
Why not "Linux Standard Desktop"?
Because it's existed for far longer then the vast majority of people have even considered using linux for a desktop system (Disclaimer: I have been using linux for my primary desktop for around 6 years)
Meh, shouldn't feed the troll and all that, but LSB set standards for things far beyond the desktop.
There are shills on slashdot. Apparently, I'm one of them.
As we watch competition help the Linux landscape considerably, is LSB a good thing? Gnome and KDE push each other to become better, Java and Mono compete for developers and even Rails and J2EE go after the web market.
Here is a standard that specifes how to package APIs and which APIs to use if you want to have a a LSB complient desktop and application. Isn't that a bit restrictive?
ALL YOUR LINUX STANDARD BASE ARE....
This is a hacked account, for which the owner can not be held responsible.
This is excellent news. Linux application developers write their applications against a particular distribution. Some add code to detect what distro the installation is for and adjust their paths accordingly. All other distro's have to spend time to write an installer to map that configuration to their distribution's file system layout. This is repeated every time a new release of a package comes out. With 100,000+ packages, that is a lot of work. Take Gentoo's ebuild system. Can you image how much effort it takes to maintain all those ebuilds? Any standardization of the Linux filesystem layout will reduce this effort and saves countless hours. Hours now can be used to improve applications themselves. Good news indeed :)
Soooo... 3.1. The first usable version of Windows was 3.1 also. Coincidence?!
:-)
You have a different defintion for usable then me
There are shills on slashdot. Apparently, I'm one of them.
Why not "Linux Standard Desktop"?
Because in the true spirit of free software, it are belong to all of us.
>Why not "Linux Standard Desktop"?
...
>Because it's existed for far longer then the vast majority of people have even >considered
LSD, that is.
There's plenty of concern that the LSB platform is restrictive, but I think it's going to be a huge step forward for 'normal' users who don't want to know how to make an application work - we just want to run it. Linux is still an open model and so LSB is not compulsory, and if an application can be built better in an alternative way, great!
I've had plenty of hassle trying to get various packages to work on older Linux systems, spent endless hours trying unsuccessfully to get services for a wireless network running on the previous version of Fedora. I'm hoping LSB will allow a simple download of an executable that just works - the ability to download an exe and just run it seconds later is probably the biggest advantage of Windows v Linux IMHO.
"delivers ever increasing value to customers - It enables further uniformity and standardization across applications and distributions that allows quicker deployment of Linux solutions with higher levels of quality"
Now we're speaking the language management can understand. All the stuff about "symmetrical multiprocessing" and "system bus throughput" was just a bunch of incomprehensible gobbledygook.
try { do() || do_not(); } catch (JediException err) { yoda(err); }
Because it has nothing to do with desktops, per se. It is a specification for directory layouts, config files, and required libraries, and its purpose is to make sure that applications that compile on one system that complies with the LSB will compile on all systems that comply with it.
Um no. Its purpose is that you can create one LSB-compliant application and have it work on all LSB-compliant systems, but it doesn't follow that anything that compiles on an LSB-compliant system is an LSB-compliant application. This might be fairly obvious and what you intended to say, but it's important that people don't think that just because it compiles on their machine it'll work for others.
Live today, because you never know what tomorrow brings
*slaps forehead*
/yoda
Indeed.
Trolled successfully I was.
There are shills on slashdot. Apparently, I'm one of them.
Linux was (and is) supposed to be a Unix-type kernel. Nothing more, nothing less.
There are objectives that you would like to see achieved and there are avenues to achieve those objectives.
So the question becomes, what objectives will the LSB achieve and whether you believe those objectives should be achieved.
The LSB was, originally, an attempt to make it easier for ISV's to port their apps to a "standard" that would run on any Linux box that was LSB "compliant".
One problem was that one "compliant" box did not have the same libraries and such as another "compliant" box.
Another problem was that the various distributions had more financial incentive to push their own partnerships with ISV's rather than wait for the bugs to be worked out of the LSB. Which is why Red Hat and Oracle work so well together.
The third problem is that the LSB does not set a standard. It merely documents what some of the distributions are doing. In effect, they write down what the "de facto" standard was from 6 months ago and publish it 6 months from today.
Eventually (maybe), Linux (the kernel and all apps including desktops) will have matured sufficiently that there will not be such a difference in a year. Until then, the LSB isn't going to be much use.
The only way to reliably guarantee binary compatibility (especially with the test suites the LSB use) and compatibility to any important degree is to have the same binaries, and henceforth, the same distribution. It is actually possible to pass the certification for the LSB with one set of hardware and fail it with another.
LSB compatibility is a nice badge to put on your software boxes (management love accreditation logos!), but whether it will mean anything to the ISVs who should be taking notice of it and anything practical for end users is another question.
D'oh. That's what I meant, thanks for clarifying. The application has to be LSB compliant to work on LSB systems. You could certainly write a program that works on one LSB system but not others.
- None can love freedom heartily, but good men; the rest love not freedom, but license. -- John Milton
> Um, wasn't Linux supposed to break away from standards and uniformity
:)
No--where on earth did you come up with that silly notion? Linux has achieved most of its success through leveraging existing standards (e.g. POSIX, TCP/IP, ISO language standards). The one that tries to "break away" from standards is MS, because standards don't promote customer lock-in. If you follow standards, then customers may be able to look at other vendors that follow the same standards.
Standards in Linux are not mandated (because you have the freedom to do whatever you want with the code, pretty much), but are greatly respected and generally followed when possible/reasonable. Standard-breaking Linux projects (and I admit there are some) are almost always completely outside of the mainstream.
> or is that just breaking away from Microsoft standards?
"Microsoft standards?" Isn't that an oxymoron?
What MS mostly has is ad-hoc, undocumented arbitrary code which the rest of the world is just supposed to accept as-is without questioning. The main notice they take of standards is when the see an opportunity to embrace-and-extend to subvert a standard (see ISO C, HTML, Java, Kerberos, etc., etc.)
> Sarcasm if you didn't get it.
Um...does that mean that you're a troll, rather than just a very clueless person? If so, then count me as trolled, but my post is really addressed to those who are clueless enough to think there's some validity at all to what you posted.
I'm still reading the latest spec to see if this has been or is going to be addressed. When/if it is, then I'll be very happy, because it will mean finally the end to confusion about using the "right" RPM repositories for your distro: if the distro is LSB compliant, then any RPM repository for that distro should work with other LSB compliant distros, with the dependencies for packages containing Base libraries being met or at least consistant accross the distros.
Until that happy day, the LSB doesn't add a lot of value to me as an end-user. As a developer, it does have some small value, in that it provides me a consistent API, but that's about it...
“Our opponent is an alien starship packed with nuclear bombs. We have a protractor.” — Neal Stepnenso
No, it's not a coincidence, considering that nothing is coinciding. Linux is already more than usable, and windows is still being majorly re-thought. As for linux desktops not having "arrived" yet, that's a ridiculous concept that only lives on for people who still use GNOME.
As for linux desktops not having "arrived" yet, that's a ridiculous concept that only lives on for people who still use GNOME.
Come back when Linux driver support by manufacturers reaches the same level that Windows 3.1 driver support reached, that is, when at least one manufacturer of each type of PC peripheral makes a commitment to support for use of its products with home editions of GNU/Linux, up to and including including a working Linux driver on the bundled CD. Scanner maker Microtek sure hasn't.
Yeah, Mac users will have to wait for Leopard's virtualisation layer to get a usable Windows.
echo -e 'global _start\n _start:\n mov eax, 2\n int 80h\n jmp _start' > a.asm; nasm a.asm -f elf; ld a.o -o a;
I would probably be put under the clueless catagory on this one. I replied before I thought. Oh well, life goes on.
Click Click Bloody Click PANCAKES!
As long as the standard requires rpm files, its going to be a non-starter for many. The rpm format sucks the big one. I would much rather almost anything else than be stuck in RPM HELL. Give me slackpacks, apt-get, or even tarballs...but please save us from RPM Hell!
ttyl
Farrell
p.s. I don't like rpm, can you guess?
CAN-CON 2019 - Ottawa's only book oriented Science Fiction Convention! October 18-20, Sheraton Hotel, Ottawa, Canada h
I sure hope Caldera is one of the others!
(SCOre: 5 Ironic)
I'll be waiting for LinuXP, seeing that no lagging date is implied.
How are sites slashdotted when nobody reads TFAs?
Second... rpm is a deal breaker.
It's dpkg or nothing.
Spoon not. Fork, or fork not. There is no spoon.
As long as LSB95 and LSBXP don't follow :)
http://rpmforge.net/manifest.php
This will be the beginning of the end of these issues.
If RPMForge can succeed then 3rd party developers can use their lessons-learned (build tools) to release packages or create repositories that work on as many systems as possible.
Also I hope (as mentioned by another poster) that some base RPM dependancy names are adopted for truly cross-distribution packages.
I think the best solution would be for RPMForge could create distribution-specific packages that add these symbolic entities into the RPM database.
So like on SuSE the lsb-3.1.i386.rpm would have as its dependancies packages that come with SuSE distributions whereas in the RHEL4 tree it would have a similar but slightly different dependancy list... and both would provide 'lsb 3.1' or whatever. Any required "patches" would be in that requirement list... so if SLES 9 has some issue with complying with LSB 3.1 rather than wait for Novell to patch it RPMForge could release rpmforge-lsb-sles-9-compat.i386.rpm which adds some files or runs a script to make it compliant.
Tie this together with yum...
THIS THING CAN TURN ON A DIME, MACROSSZERO STYLE ALSO FUCK BETA, ~NYORON
The latest version of Oracle is supported on not only Red Hat, but most Debian derivatives including Ubuntu.
Or maybe you mean desktop apps like just about anything based on Eclipse?
I will concede that Microsoft doesn't support Office on any non-Redhat version of Linux. But, then again, they don't support Office on Redhat Linux either.
If Microtek don't support Linux, don't use Microtek products. Whats so difficult about that?
The fact that I already own a paid-for Microtek product and cannot even get an interview for a job despite my qualifications is so difficult about not using a Microtek product.
Well written. Even if the "sarcastic" parent didn't need the smack with the clue stick (is there an equivalent "don't be such an ass" stick?) you make a good point about educating others.
What MS mostly has is ad-hoc, undocumented arbitrary code which the rest of the world is just supposed to accept as-is without questioning. The main notice they take of standards is when the see an opportunity to embrace-and-extend to subvert a standard (see ISO C, HTML, Java, Kerberos, etc., etc.)
I would add that they also "embrace" standards when they feel the need to do so to please requisition departments. POSIX and CS2 certification in NT4 comes to mind... POSIX in NT4 was a joke and if I remember correctly CS2 certification was only achieved by running an NT 4 server that was not plugged into a network.
Creationist Textbook Stickers Declared Unconstitutional by CowboyNeal
Er, it is the first time that I know Mark is the Chief Developer of Ubuntu....
Considering most if not all of the commercial linux companies specialise in the server market and pay little regard to the Desktop, it would make more sense to have one desktop linux distro and base their products around it. It is a more efficient way of concentrating resources by maintaining just one packaging and linux system to the LSB standard. It would also mean that more resources could be diverted to projects that are both benficial to the commerical and home desktops like KDE, Gnome etc. The only downside is that the financial support for Community distro's would most likely be cut.
...from the start!
I hope 3.1 addresses my main gripe with RPMs: an RPM built for Fedora won't install into SUSE because of dependency issues, or vice versa.
Application distributors aren't supposed to build RPMs for Fedora or SuSE or Mandriva; they SHOULD be building RPMs for LSB. ALL RPMs built for LSB should have exactly ONE dependency: the LSB package. The old LSB troll on here from supposed Debian fans alternately makes me chuckle or want to smack then upside the head. Basically they say "but RPM sucks because of dependency hell and DEBs are way better because you can use apt-get--those idiots at the LSB should change the standard to use DEBs instead". HELLO DUMASSES, the whole idea of LSB is that there ARE no dependencies except the LSB standards! Why go through the bother of changing something as fundamental as the package/deployment format just because it is better at handling a problem that would be non-existant in a standards-comliant system!
What is exciting here is that LSB has not really been that "useful" because it was too narrow in scope. With LSB 3.x they have finally given serious attention to the DESKTOP by specifying base requirements for things like X, GTK and (optional at the moment) Qt libraries. Until now, if you wanted to distribute, say, a modern GTK-based GUI application as LSB the install package would have to link in all of that crap and it'd be HUGE. LSB-compliant packages were really only useful for console-based apps or server applications which severely limited the appeal of LSB. Now packaging and distribution LSB apps for both client and server, including rich GUI applications using modern toolkits, is reaching the point where it is actually practical to do so.
Most of the big OS distributors have been on board for awhile now (yes, including some Debian-based ones). Hopefully more big APPLICATION developers will clue in quickly to the fact that LSB is starting to hit its stride.