Thanks Benoît for the link to SUN's JDS patches.
If anyone ships this Totem patch, they get to fix all their bugs themselves. As if X worked that well with threads, they're making all the display locking no-ops. If it hangs, your libraries are broken, and you're just 4 years late fixing it.
Update: Damien kindly posted a link to the original bug report. And it confirmed my above assumption. They have a broken libXi. For reference, the original Totem bug report.
Jeez.. whats with all the Sun hostility today on planet.gnome.org. If the patches are bad why not bring it up with the developers of the patch and work with them to get the proper solution upstream. Benoît's blog was particularily nasty (unfortunately I couldn't figure out how to post a comment there). Its a GOOD thing for everyone that Sun has developers working on Gnome why not work with them instead of flaming.
ReplyDeleteDisclaimer: I work for Sun although I have nothing to do with the JDS group.
Simple answer to your original questions: because I only learnt about the patches today.
ReplyDeleteAs for the other problems. There's no explanations of what the patches do, no mention of who the author is, and certainly no mention of whether they tried to push those upstream.
They obviously didn't use their best judgement, or ask the developers about the problems they encountered, otherwise this patch wouldn't be there.
I make crappy patches as well sometimes, but I always report those (and the good ones) upstream, where I get feedback.
If you reacted like this to a patch I sent you some time ago I'd know which project I'd avoid in future. Come on, if something isn't okay inform its developers about the problems, or at least stay silent. We don't need stupid rants and flames against *anybody* producing free code.
ReplyDeleteThe patch was added to fix http://bugs.opensolaris.org/view_bug.do?bug_id=6535546
ReplyDeleteYou can see the bug number in the spec file: http://cvs.opensolaris.org/source/xref/jds/spec-files/trunk/base-specs/totem.spec
and the ChangeLog entry by the patch author http://cvs.opensolaris.org/source/xref/jds/spec-files/trunk/ChangeLog
A pstack of the core file is in the bug report (unfortunately not in a section exposed on bugs.opensolaris.org).
Damien Carbery
Hendi, notice how that patch was never posted upstream, or the problem reported. I fixed the bug in Xfree86 4 years ago. The bug is known, the solution simple. But nobody bothered mentioning it, and instead pushed an obviously broken patch to tons of people, giving my own code a bad name.
ReplyDeleteOk, well I agree that perhaps the JDS developers have fallen short by not pushing these patches back upstream for feedback etc. It just seems to me that instead of publicly blasting them a quick email to desktop-discuss[at]opensolaris[dot]org would be more productive.
ReplyDeleteAnyway thats just my 2c.
Anonymous Sun poster.
Nothing a good ol' bit of whining for people to give you contact details. At the same time, I'm still available for contact, just as I was before, and that if people want to do stupid things with my software, they're welcome to. They just shouldn't expect me to have to chase them, or get help if they're not coming to me.
ReplyDeleteI also think this attitude against Sun engineers does not help; it's not constructive. What's more, it damages the reputation of the GNOME developer platform and makes us look silly (check the sponsors of GUADEC).
ReplyDeleteThe majority of Sun engineers come from the proprietary development world and I would expect open-source development is not their second nature yet.
The way this and similar developer issues should be tackled, is to have a contact that will sort out situations where one cannot handle himself. I believe for the case of Sun engineers, you should get in contact with Glynn first.
/not affiliated with Sun.
This comment has been removed by a blog administrator.
ReplyDelete