[lsb-discuss] Charter discussion (a fresh thread) was: Can we find a fit for LSB and Mobile?

Craig Scott audiofanatic at gmail.com
Fri Apr 13 23:05:37 UTC 2012

On Sat, Apr 14, 2012 at 5:27 AM, R P Herrold <herrold at owlriver.com> wrote:

> On Fri, 13 Apr 2012, Robert Schweikert wrote:
>  2.) One of the core charters, at least for as long as I have been
>> involved in the LSB, was to attempt to provide an environment that is
>> targeted towards the Enterprise market. This is probably not documented
>> anywhere, but this is the paradigm we have been following by the actions we
>> have taken and the specifications that have been released.  ...
>  (snip)
> I guess the day to discuss the matter has come sooner than I expected.  I
> think that we need to nail down "'enterprise' distribution Standards Base",
> or "ISV facilitation Standards Base", vs. Linux Standards Base with a
> charter, and so ask its inclusion as an agenda item
I can't help but wonder how much of that discussion would be made moot if
LSB had not just its current "established practice standard", but also some
form of "closer to the leading edge emerging standard" aspect as well. As
Robert stated and various others have given reasons for of late, the LSB
standard essentially reflects what has become common to all major linux
distributions. The Enterprise distributions are the ones that change slowly
enough for such a consensus to emerge, but the community ones are more
fluid. By their nature (and to some extent, by design), the community
distros are a bit experimental, often being set up as a form of proving
ground for stuff to go into the Enterprise version (e.g. Fedora for RHEL,
OpenSuSE for SLED/SLES). Thus, it would be very hard for the LSB to try to
keep up with or apply rigorously to the community distros, since those
distros don't have the same maturity.

The part that seems missing to me is the LSB folk being involved in shaping
(more strongly) what the community distros provide. Currently, the LSB is
mostly just reactive. It reflects what has already emerged and become
common. What seems to be needed is some guidance to the front runners for
things they should really be aiming to include in their upcoming releases.
This is much harder because:

   1. You have to be really on the ball with changes happening in a large
   number of software packages (or at least in the ones that you want to care
   about). This may even extend as far as influencing what development
   actually takes place, which could be like herding cats for some of the
   bigger packages.
   2. You have to get the community distros to agree with your
   assessments/decisions, as compared to simply identifying what the distros
   have already done.
   3. The community distros need a reason to care about the LSB.

For many of the packages covered by the LSB, they will probably remain
trailing like they do now. The resources just aren't there to cover them
all. However, for some of the more widely used or fundamental packages, it
would seem that there is scope to be closer to the leading edge than the
LSB currently is. Some good examples of this are OpenGL and Qt, but I'm
sure there are others (these two are just the ones I'm more familiar with).

So, in a rather long-winded way (sorry!), what I'm suggesting is that the
LSB should stay a trailing standard as it is now, but there also needs to
be something complementary running ahead of it. The goals of such a
complementary standard would be less ambitious, such as not covering all
the things the LSB does. Maybe it could just target the bigger
modules/libraries that other packages build on top of or depend on,
covering just the core parts and the major toolkits (Qt, Gtk, OpenGL)? No
idea where the resources would come from to do it, but the yawning gap
between leading edge and where the LSB is at is the primary reason so few
ISV's bother to target it, in my opinion.

Reading back over my post before hitting the send button, the one thing I
don't have an answer for is how to give the community distros a
sufficiently compelling reason why they should care. Such distros exist
because people want the freedom to experiment and do things that maybe
others don't want to (or aren't yet ready to) do. Anything that constrains
that freedom is likely to meet with much opposition, so the payoff for them
would have to be very clear and well worthwhile. What people deem
worthwhile will vary (potentially quite a lot) from person to person, given
the different views and agendas for their involvement.

Craig Scott
Melbourne, Australia
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linuxfoundation.org/pipermail/lsb-discuss/attachments/20120414/087604d6/attachment-0001.html>

More information about the lsb-discuss mailing list