[Openchrome-devel] Openchrome status

[Xavier Bachelot]

Hi Benno,

Benno Schulenberg wrote:
> Hi all,
> 
> Xavier Bachelot wrote:
>> - libpciaccess support (Jon)
> 
> Out of curiosity: what extra possibilities does libpciaccess 
> provide?  Or is it no more than a reworked method for registering a 
> device and allocating memory, for newer X servers?
> 
Yup, more or less. It will be used in the next xorg release and is 
already used in Rawhide (which will become Fedora 9 later) and that's 
why we need it now. It currently works well enough to bring X to life, 
but there are some troubles with Xv and probably DRM, etc...

>> - HW cursor improvement (Gabriel)
> 
> With this you mean the HW cursor stuff that is already in the randr 
> branch, Xavier?  Or should that be improved further?
> 
No, just what's in the rnadr branch.

>> - Code clean up (Benno)
> 
> I'd like to continue with this, but am holding off partly to not 
> further disturb the files that Gabriel is probably working on, and 
> partly because other things always seem to be more important.
> 
Both are very valuable arguments ;-)

>>  From what I gather from the bug reports, mailing list and irc,
>> here's a highly subjective list of openchrome's most wanted
>> features, in decreasing order :  [...]
> 
> To which I'd like to add, in first place:
> -  Not to have lockups, crashes, or black screens.
>
Seems a bit too generic, but, yes indeed ;-)

>> Similarly, it would be good to merge as much code as possible in
>> the trunk to get wider exposure. I mostly think of the randr
>> branch, which seems pretty helpful to a a lot of people.
> 
> Agreed.  But it would be nice to have the gamma regression fixed 
> first.
>
Sure, awaiting Gabriel comments on this.

>> I hope better documentation would help with attracting more people
>> willing to answer user questions.
> 
> Better documentation would hopefully prevent most user questions. :)  
> But that's a vain hope -- when openchrome gets distributed more, 
> there will be more questions, and especially more silly ones.
> 
But possibly more people will be able to answer them too.

>> The bug tracker is also not 
>> really helpful as new bugs are not posted to the mailing list,
> 
> Such posts to the mailing list would be very useful yes.  Can trac 
> be made to do that?
> 
No idea. Ivor, Jon ?
I'll take a look at the Trac doc anyway.

>> anonymous reports are allowed
> 
> Very annoying indeed.
> 
I should be able to fix that.

>> and we have a huge backlog of 
>> unanswered and badly/not triaged bugs.
> 
> I would like to attack this, but am reluctant to close bugs, afraid 
> to hide useful information from sight.  But if you tell me "do as 
> you see fit; if a bug is important, someone will refile it", I'll 
> go ahead.
> 
I began to triage the bugs already. I created some components and tried 
to triage the bugs into them. I have not deleted any bug yet and I don't 
thing it's a good idea until we've make sure no valuable information 
would be lost. The main problem is following up with the reporters which 
are too often anonymous.

Regards,
Xavier