Linux Standard Base Adoption/ Development in the Linux Community

Why does the general Linux community not seem to care about the Linux Standard Base (LSB )and any sort of standards? We need more compatibility across the distributions and to stop the rampant fracturing within the community. With Microsoft closing in on the Linux community, we need to unite as a community, instead of further fracture and become weaker for it.

On Sun, 08 Oct 2017 23:26:02 +0000, Panaman wrote:

> Why does the general Linux community not seem to care about the Linux
> Standard Base (LSB )and any sort of standards? We need more
> compatibility across the distributions and to stop the rampant
> fracturing within the community. With Microsoft closing in on the Linux
> community, we need to unite as a community, instead of further fracture
> and become weaker for it.

What makes you think that the Linux community doesn’t care about “any
sort of standards”?

There are plenty of parts of Linux that not only conform to, but drive
standards.

Jim

Jim Henderson
openSUSE Forums Administrator
Forum Use Terms & Conditions at http://tinyurl.com/openSUSE-T-C

First,
It should be understood that conforming with standards is diametrically opposed to innovation, at the least standards compliance restricts the possibility of better options and worse may cause people to not even consider those possibilities.

In a way the tension between conformity and innovation is fundamental to all aspects around us, we profess to want both when having both in their entirety is impossible.

This is why typically we decide on which principle to follow on a case by case basis but cannot claim to always follow one or the other (or this is how people become brainless chauvinists).

It looks like the LSB is one vision of what Linux as a whole should be, but some requirements are amusingly impossible… Like requiring RPM package management as described in the following Wikipedia article on LSB. It should not be surprising that by essentially excluding the entire galaxy of aptitude distros, this isn’t going anywhere

https://en.wikipedia.org/wiki/Linux_Standard_Base

And, by and large SUSE/openSUSE does seem to be very compliant compared to other distros, apparently SUSE 10 was certified for LSB 3.1 (same Wikipedia article).

TSU

I would disagree with the fact that standards are “diametrically opposed to innovation”. I would say they promote innovation by giving developers confidence that what they build will be available to the entire Linux community with only having to maintain one version of there app or program. And i realize that requiring RPM package management is hard, but the fact that there is a split in package managers and formats doesnt help anyone. And if requiring RPM package management isnt the answer, shouldnt the Linux comunity come together to agree on a package management or create a new one to uphold to? It wouldnt be easy, but i think being more unified would attract a lot more users to the community of Linux as a whole on the desktop and server side.

I dont mean to sound like an ass. Just trying to have a good conversation/debate

From what i have seen, package management is something that needs a cross distro standard. the VLC homepage has 10 seperate download links for Linux. where as it has 3 for windows where 2 of those are builds for windows phone and store. So 1 download for desktop Windows, 10 for Linux. I just think its absurd that to cover most distros you need 10 separate links.

I dont mean any disrespect, i just want to learn and have an open conversation about it.

On Mon, 09 Oct 2017 04:16:01 +0000, Panaman wrote:

> From what i have seen, package management is something that needs a
> cross distro standard. the VLC homepage has 10 seperate download links
> for Linux. where as it has 3 for windows where 2 of those are builds for
> windows phone and store. So 1 download for desktop Windows, 10 for
> Linux. I just think its absurd that to cover most distros you need 10
> separate links.

Open source, by definition, means the source can be built with a number
of options - choice is a big part of why open source works at all.

It also means that if a distribution does something that someone doesn’t
like, they have the option to do their own thing. That’s what’s meant by
“free” software - sometimes called “libre” (to distinguish from “gratis”).

> I dont mean any disrespect, i just want to learn and have an open
> conversation about it.

That’s fine, but I would suggest approaching it from a standpoint of “I
want to learn more” rather than “I have the answers for what’s wrong with
the Linux community”. I know you probably didn’t mean it to come across
that way, but it is possible to learn without saying “just stop the
rampant fracturing in the community” - ie, proposing a solution to a
“problem” that actually is one of the core strengths of the community. :slight_smile:

Jim


Jim Henderson
openSUSE Forums Administrator
Forum Use Terms & Conditions at http://tinyurl.com/openSUSE-T-C

I agree with what you have said. I am sorry for coming off the way i did. I get passionate and i dont always come across the way i should. I mean to say that in my personal opinion, open source and the freedom to choose and do it your way is awesome and i highly respect it and think it has more power than people realize, and that power could be directed by the community at large to make Linux a more friendly place for your average person looking to get away from windows.

Again, i didnt mean any disrespect. I apologize for my tone and ill advised “I know it all” attitude. Thank you for not yelling at me.

On Mon, 09 Oct 2017 04:06:01 +0000, Panaman wrote:

> I would disagree with the fact that standards are “diametrically opposed
> to innovation”. I would say they promote innovation by giving developers
> confidence that what they build will be available to the entire Linux
> community with only having to maintain one version of there app or
> program. And i realize that requiring RPM package management is hard,
> but the fact that there is a split in package managers and formats
> doesnt help anyone. And if requiring RPM package management isnt the
> answer, shouldnt the Linux comunity come together to agree on a package
> management or create a new one to uphold to? It wouldnt be easy, but i
> think being more unified would attract a lot more users to the community
> of Linux as a whole on the desktop and server side.
>
> I dont mean to sound like an ass. Just trying to have a good
> conversation/debate

Your arguments ignore the mechanism by which Linux is supported. Each
distribution is supported primarily by volunteer developers, each one
evolving over time to meet distro-specific goals. Without enforcing
unrealistic release schedules, how would you propose to achieve the
synchronization among the disparate support communities??

IMNSHO I think we already have tools that provide what you suggest. Just take a look at where most source projects are: github. Packaging is a matter of distro choice. In the end the packagers pull the sources to compile from one and the same place.

What you’re suggesting is in fact going back to one single linux distro, and that is not going to happen. Rather consider the Windu way as just dealing with the core ( kernel ). Not going to do that but there are quite some programs ( packages ) for Windows that do have versions for each Windows version.

Standards impose a specific way of doing things to enforce conformity.
And, in your argument, the innovations may not exist in the first place for adoption by many.
Specifically regarding APT vs RPM package management, good luck… As described in the article the aptitude people tried but in the end gave up trying to satisfy your “higher solution” approach.

To understand “Big Idea” issues like this, perhaps it’s best to leave our heads in the clouds and talk about specific down to earth examples…

DirectX vs OpenGL - Although OpenGL has been around for quite awhile longer, it hasn’t been able to measure up to what DirectX delivers which is a major reason why XBox easily overtook other gaming consoles as the leader. Surely you shouldn’t then say that all our gaming machines should run on Windows only (because DirectX won’t run on any other OS).

HTML5 - On the one hand this “latest” web technology standard now dominates how web content is created. On the other hand, unlike earlier HTML standards, it has never been ratified so is not an official standard, so it would never come to exist without the rebels who were willing to be non-conformists. But, turn this a bit further and now that all web browsers support some collection of HTML5 features (yes, they’re all different), end Users like you and me are at the mercy of what can be done in HTML5 including interrogating your machine for Personal Information about you, your machine’s features and some apps running on your machine. Today we have a de facto HTML5 world, but entirely built on zero ratified standards… But, when was the last time you went to a website that wouldn’t work?

SOAP vs REST - This is a bit more technical only Developers who build distributed apps and web developers who build “responsive” apps will know about… First there was the ratified SOAP standard for data connections and transfer across WAN links… But, it was also known to be extremely heavy and complex to implement, not consistent with the lightweight philosophy of technologies like AJAX. Almost as soon as SOAP was ratified, a new technology called REST appeared which was lightweight, simple and could be made as (or nearly so) secure as SOAP. There was a big fight between the “standards compliant” people and the rebel “I’ll do what I want” and today practically everyone does REST.

These examples are only three that illustrate different consequences of standards conformity vs innovation.
You have to break restrictive conformity to create innovations that improve, yet as I also describe the chaos of unfettered innovation needs acceptance and adoption which can lead to standards, ratified or not. These two diametrically opposed principles must simultaneously exist although in each’s purest form the other cannot.

HTH,
TSU

On Mon, 09 Oct 2017 07:36:01 +0000, Panaman wrote:

> I agree with what you have said. I am sorry for coming off the way i
> did. I get passionate and i dont always come across the way i should. I
> mean to say that in my personal opinion, open source and the freedom to
> choose and do it your way is awesome and i highly respect it and think
> it has more power than people realize, and that power could be directed
> by the community at large to make Linux a more friendly place for your
> average person looking to get away from windows.

For a lot of people in the OSS and Linux communities, being a place for
Windows users to move to is secondary - learning Linux does involve some
effort, and that effort is something that’s well-rewarded.

Indeed, there are distributions that are distributed in source-code form
only (Linux from Scratch, Gentoo) that are geared towards helping people
learn a lot more about Linux. Debian-derived distributions that use
aptitude for package management have their own approach to doing things;
RPM-based distributions like openSUSE/SUSE/Redhat/CentOS have theirs.
Neither is really better or worse (there are pros and cons to each
approach), but because people are involved, people have their own ideas
about “what’s best”, and follow their passions. That’s the core of what
open source software development is all about.

> Again, i didnt mean any disrespect. I apologize for my tone and ill
> advised “I know it all” attitude. Thank you for not yelling at me.

You’re welcome - and you’re not the first to start a discussion like
this. It comes from a lack of understanding about what the purpose of
open source is. For those of us who have been involved for a long time,
it’s easy to forget that “we were newbies once” as well, and some will be
upset by that approach. As I’ve gotten older, I try to remember what it
was like to feel like I knew it all. Still make that mistake from time
to time. Experience is a great teacher. :slight_smile:

Jim

Jim Henderson
openSUSE Forums Administrator
Forum Use Terms & Conditions at http://tinyurl.com/openSUSE-T-C

On 2017-10-08, Panaman <Panaman@no-mx.forums.microfocus.com> wrote:
> Why does the general Linux community not seem to care about the Linux
> Standard Base (LSB )and any sort of standards?

I believe the RPM GNU/Linux community care about the LSB and the DEB GNU/Linux community do not. Since the DEB GNU/Linux
community is much bigger, this gives the impression that the general GNU/Linux community do not compare about LSB.
Since the enterprise GNU/Linux community however do care about the standards, LSB will always be supported and that is
one of the wonderful consequences of openSUSE being related to SLE.

> We need more
> compatibility across the distributions and to stop the rampant
> fracturing within the community.

Bryan Lunduke regularly gives talks called `Linux sucks’ (although I believe his 2017 recording may be his last) making
this point. The GNU/Linux community is a community not a commercial venture. As a result a huge amount of time and effort
is wasted on the parallel duplication of similar technologies (e.g. Qt vs GTK, systemd vs sysinit/openrc, KDE vs GNOME)
and often the divides are the basis of idealogies that appear fairly arbtrary to pragmatists. The idealogical GNU/Linux
zealot cares little for fracturing in the community and this isn’t going to change anytime soon.

> With Microsoft closing in on the Linux
> community, we need to unite as a community, instead of further fracture
> and become weaker for it.

Well, no. Remember the Linux community includes Android which is 100% Linux but 0% GNU, and Microsoft are not closing in
on Android. That’s why it’s important to distinguish Linux and GNU/Linux, and existence of such a distinction
constitutes yet another fracture. The fact that the Linux Standard Base is much more about GNU than about Linux yet
doesn’t even have GNU in the name proves it isn’t fit for purpose as a unifying strategy, but only serves as common
specification for commercial vendors (such as Intel) when compiling their binary backages with pre-established
dependencies. This is makes LSB useful, but not in the manner you propose.