Minutes September 8 2010

From The Linux Foundation
Jump to: navigation, search

Attendees:

Jeff Licquia (LF), Mats Wichmann (Intel), Alan Clark (Novell), Stew Benedict (LF), Alexey Khoroshilov (ISPRAS), Jiri Dluhos (Novell)

Agenda:

  • Status on LSB 4.1.
  • Status on next SDK.

Minutes:

Jeff: LSB 4.1 - feature freeze is in a month, then bug fixing. We need to either start getting things in or dropping from the list.
Jeff: Any progress to report?
Mats: Mostly bug work, some minor code changes, nothing significant.
Jeff: On infrastructure, no progress in the last week, some issues remaining with libbat etc. still. Once that is resolved, we should be able to autobuild 4.1 stuff, once we create a scheduler.
Jeff: Alsa - still need dtk support, review of shallow tests, incomplete tests addressed
Jeff: symbol inclusion - still need the rest of the symbol adds
Jeff: CUPS interfaces - 9 interfaces still need to be added. Tests would be good, but may not be possible at this point.
Jeff: Need dtk support for the gtk-print-dialog for the appbat app.
Jeff: Perhaps one of us needs to jump in and do the symbol adds and the dtk work
Jeff: Anything else on 4.1?
Mats: We have a set of bugs, but I'm not sure if it's a good description of the release goals
Jeff: A lot of these changes are driven by building an LSB compliant virtual machine
Mats: Right, and java is in a little bit of a questionable state.
Jeff: We do have all the uplifts (cups, gtk, cairo) for 4.1
Mats: Also the xdg-utils issue, needs to move from Trial Use or be dropped
Jeff: Probably need to review all the Trial Use module status

Jeff: Next, sdk status. At least one known issue building things with the current released sdk.Took a look at rollup for 4.0.4 update and 4.1.0 release. 4.1.0 is actually in a better state.
Jeff: There are a couple specific to 4.1, htmlSaxHandler addition - perhaps this should be applied for 4.0 also.
Jeff: Also a change to zlib, bug 2967, marked as 4.1 only. Perhaps this should also be applied to 4.0 also. Any opinion?
Mats: No strong opion. Can revisit
Jeff: OK, re-opened both bugs to block the 4.0.4 refresh
Jeff: Another one for gi18n.h. Opinions on backport? Will give it the same treatment
Jeff: Looking at the 4.0.4 rollup, are there any bugs here that need particular attention?
Stew: Bug 2884 (longdbl) work is going to feed into any 4.0 update
Mats: Bug 2903 is probably the bug that should be linked in there
Jeff: Looks like 2903 is actually targeted to 4.1
Jeff: OK adding blocker of 2884 to 4.0.4 update
Jeff: So we have 7 open bugs for 4.0.4, 3 of which are questions of backporting. That leaves 3 so-called "real" bugs
Jeff: Version independence of pkg-config is one - probably should be addressed
Jeff: Another is 'sdk knows too much about LSB version' - pkg build process
Jeff: Last is 'union _GtypeCvalue', to be applied all the way back to 3.1, some question as to how the header gets generated
Mats: Yes, some issues with the way things get represented in the headers, main goal at this time is if it works
Jeff: Has the sql been applied?
Mats: Yes, ideally backported stuff would get an errata to the spec, that's why it's not closed
Jeff: OK, I think we have a pretty good idea of what needs to be done to get an update out. Biggest real problem is probably 2810, the pkg-config part, which will take some work.
Jeff: Anything else on the SDK?
Jeff: Any other issues?

End Of Call