[lsb-discuss] Don't blame LSB and standards, please: was: Re: Fedora Plasma Product, feedback please

[Russ Allbery]

Denis Silakov <dsilakov at gmail.com> writes:
> On 01.04.2014 02:27, Russ Allbery wrote:

>> At least, as a distribution developer, my impression has always been
>> that LSB is primarily aimed at providing a standardized platform that
>> third-party commercial application developers can target and be
>> relatively assured of being able to run on any Linux distribution.  For
>> the most part, the actual workings of the native packages of the
>> distribution were out of scope, provided that the interfaces were
>> available to third-party software of that sort after installing the lsb
>> package.

> That's not completely correct, imho. From what I can see (as a
> distribution developer, too) is that a lot of human resources are spent
> on adopting the same software products for different distributions -
> e.g., preparing patches or just updating spec files or dpkg build
> scripts. Maybe it would be more useful for the Linux community if a
> hundred of people instead of rebuilding Firefox for a hundred
> distributions would work on improving its speed?

> And it really helps when upstream code can be compiled in different
> systems without any patches. This source-level compatibility is not
> exactly the LSB area, but source and binary compatibility are highly
> coupled and LSB is able to help here, as well.

All of this is true, but I haven't seen many signs that the LSB
specifically helps with this.  Most of the efforts in this direction are
from the individual packagers coordinating through upstream or sharing
patches and approaches among each other.

I personally find this area more interesting than getting third-party,
often closed-source, applications working across Linux distributions, but
I also think there are a lot of other communities working on this problem:
Autoconf and Automake, all the various upstream communities that care
about portability, a lot of interdistribution communication channels, and
some projects that are specifically aimed at turning the best practices of
a variety of distributions into code (systemd, for instance).  So I'm not
sure there's a large space for LSB here, although there are certainly
things like the FHS that are valuable in that direction.

> I would also note that LSB already provides a significant profit for all
> distribution developers (and in some sense, for application developers,
> too) - this is LSB test suite (a bunch of different test suites, to be
> more precise) that can be easily launched by means of Distribution
> Checker. This test set is one of the biggest one in the Linux world and
> from time to time allows to detect quite tricky issues - I can say this
> on the basis of our practice, we use LSB tests as a part of our QA
> process in ROSA.

That's interesting to hear.  I admit I have very little experience in this
area, and maybe that's a counter-example to my impression that LSB doesn't
have a large role in the space of standardizing software across
distributions.

-- 
Russ Allbery (eagle at eyrie.org)              <http://www.eyrie.org/~eagle/>