[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

LSB; was: Re: starting Fedora Server SIG

Sorry to come late to this discussion and break fedora-devel ML threading, as I am working from the web archive (I had tossed the underlying pieces already)

Dan HorĂ¡k at Wed, 12 Nov 2008 18:16:00 +0100:

IMHO LSB Core conformance should be required in minimal install

Sadly, as a long time participant in the LSB process, I would offer that there is not a well defined subset such as 'LSB Core' (and adjuncts or extensions such as: 'LSB Desktop', 'LSB Mobile Device', 'LSB Headless') are not a way that the LSB decided to go in 4.0 initial (the next scheduled release, due out later this year). This is in part from a lack of developer mass and consensus.

LSB has a weekly conference call open to all contributors to the LSB; its mailing list is open as are its archives {some process happens 'out of sight' which is out of scope here}. Mats Wichmann and I have been at OLS, and the LSB BOFH or presentation for the last few years has drawn few attendees, and sadly either just the 'usual suspects' or people looking for a seat in a reasonably empty and quiet presentation room ;)


Earlier, Peter Robinson at Wed, 12 Nov 2008 11:53:21 +0000

For some reason redhat-lsb pulls in qt, qt-X11 and qt3. redhat-lsb is certainly something you'd probably want in a server environment but no idea why it would need qt

This being an 'redhat-lsb' lsb in the 3.x series.

LSB 4 is in release testing, and 'slops' all sorts of 'goop' in to a 'conformant' installation: sound, printing, X, and GUI widget sets, ... and more, in. The 'use case' was that the ISV's needed each to be predictably present.

I predict that when LSB 4.0 releases, there will be much wailing and gnashing of teeth, because no-one from Fedora has 'been at the table' arguing (and getting the consensus and running code in place for) the 'Server SIG' need case.

A year ago, and periodically, lone voices contributors to the LSB in the server wilderness, cry out:
	Enough to the LSB

but no-one steps up to suggest the subset [and then to do the repackaging (breaking off the one GUI dep that 'Requires in much unwanted (in a server environment) GUI 'goop') in support within the distributions] to elide such cruft not wanted in a 'server' environment.

LSB is willing and interested in having a sensible discussion, and reaching specification of such a 'subset'; this seems like a natural for people in a Fedora Server SIG to collaborate and participate with this 'upstream' at LSB.

In the weekly conference call today, it was clear that such an effort do advance a rational subset would be entertained (not in the initial 4.0, probably -- it is past feature freeze there, but later in that series)

Please join in.

-- Russ herrold

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]