7.3 release progress

[Eric Anholt]

I've been shirking my responsibilities as release manager, and sadly the
release has suffered.  I've tried to catch up, and I'm about ready to
push 1.3.99.2, once I can get another approver to review the last two
changes for 1.4.  I'm going to push 1.4, and thus the 7.3 katamari, out
for another week beyond initial schedule to see if we can't get these
issues cleared up, and hopefully get some more testing.  So, 2007-09-05
is the new date.

Here's the status I see.  Bugs marked with asterisk are the ones I
particularly think are blockers.
Current "Blocker" bugs on #8888:
2859:	long-standing issue with radeon for a few people, not sufficient
	to block release for.
4037:	While it's an important cleanup, it's not fixing specific
	release-blocking bugs at this point.
5392:	doc tracker:
	5380: small note about fontconfig suggested, doesn't seem
		blocker-class.
	6193: some drivers need gl.h but don't check during
		configure.ac.  Distributions already handle the
		dependency, so I don't care enough to fix it.
	11566: about-catalog-dir-FPE text, which is a new feature this
		 release.  Anyone want to review?
*6502:	randr 1.1 issue after significant uptime, with patch.
	Review by keithp? 
6842:	crash during VT switch reported against 7.1.  I haven't seen this
	one.  Anyone else?
*7248:	PCI domain issues.  Sounds blocker-class.  jbarnes was looking at
	it -- any updates?
*7447:	Definitely a blocker, has a straightforward fix that needs to be
	written.
*7512:	Memory leak easily reproduced against Xvfb, but not tracked down.
7885:	Subtle DDC timing issues on some 945-ish hardware, doesn't seem
	like a blocker to me.
8885:	Xdmx lacks active maintenance, and is off by default.
9581:	Xdmx lacks active maintenance, and is off by default.
9588:	If we don't find someone with nsc hardware to test, I'm inclined to not
	leave it a blocker
10262:	We've had enough time with i-h in the tree, I'm fine with releasing with
	the few remaining input drivers non-building.
11504:	keithp said the given fix was close but incorrect.  I would like
	to see it fixed.
11796:	patch attached.  Anyone to review?
*11802:	input issue, not sure how common.
*11831:	input issue that many people will see (though I haven't).
11852:	driver-specific multihead issue, not a blocker.
*12055:	Crasher.  Can anyone reproduce?
12145:	zaphod mode cursor issue.  Anyone care enough to fix?
12242:	not sure what this patch fixes.

I'll hopefully get a chance to look at #7512, and I might be able to get
to #7447 if aaronp doesn't (though I hope he can).  The main issue I see
here is the input problems, with daniels currently busy with XDS
organization.  If anyone else could pick up any of them, that would be
great.  Daniel had a couple more concerns for server input stuff:

• XFree86: Eliminate duplicate devices with HAL
  - need a device hook to interrogate drivers if it's already them
    i.e. if you have /dev/input/event2 open via evdev in your config
    file already, don't hotplug it
• Sort out something better with AEI
  - right now, we need Option "AllowEmptyInput", but this is a bit crap.
    was thinking maybe a DDX hook to add the default OS devices, if we
    can't connect to HAL.
• KDrive: Allow DDXes to override NIDR
  - add an OS function to filter NewInputDeviceRequest, so Xephyr (e.g.)
    can reject them if it's already got the host input drivers
• --disable-dbus

Of these, the first seems the most serious to me, and he suggested that
input hotplug actually be defaulted to disabled if it can't get fixed.

-- 
Eric Anholt                             anholt at FreeBSD.org
eric at anholt.net                         eric.anholt at intel.com

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 187 bytes
Desc: This is a digitally signed message part
URL: <http://lists.x.org/archives/xorg/attachments/20070831/7601f414/attachment.pgp>