[lsb-discuss] Using lsb_release for other purposes

Wichmann, Mats D mats.d.wichmann at intel.com
Tue Apr 8 21:11:33 PDT 2008


Robert Schweikert wrote:
> I think we should not over load the information the lsb_release
> command returns. If we need more info lets have a separate command.

Not arguing one way or the other here... just want to observe
that a new command is harder than extending one distros already
have, and may indeed be impossible for some distros that are
quite near to release (pointed reference to Ubuntu Hardy) while
I could see such possibly later releasing a patch to slightly
extend an existing command.

> As far as packaging is concerned we are working on the packaging API
> which will hopefully give us programmatic access to information about
> installed packages, and with respect to hardware we need to get going
> on putting some proc info into the LSB.

for the first part,  API suggests an interface in a particular
language (normally C), but LSB is now five languages: C, C++, 
Posix shell, Perl, Python.  We haven't pledged to support everything
in every language, which is an interesting question of its own
(that is, should Python/Perl bindings for LSB features like 
OpenGL, Gtk+ and Qt exist if the feature itself is in LSB), but
a supported-(sub)module query is one that certainly should be
available to all supported languages in some form. Perhaps the
term "API" is too loaded for my taste...

for the second - yes, this seems to be somewhat of a need, but
I think we're still looking for a concrete proposal that can
be circulated to the appropriate parties to see if it fits the
need, and at the same time is supportable and testable. /proc
is another one of those areas that has been on the table
forever, needs a champion who'll bring it to a decision point.

Maybe this is fodder for discussions at the LSB meetings
Thursday/Friday this week?  


Which brings me to a point unrelated to the subject line of
this message:  for folks who are NOT gathered here at the
Linux Foundation Collaboration Summit, and have thoughts on
issues being discussed for LSB 4, it would be really good
to send out email with those opinions on those to lsb-discuss
REALLY SOON if they haven't already been heard; we're getting 
close to freezing the feature targets, I belive Ted wants us
to close down on that at this meeting.  




More information about the lsb-discuss mailing list