[Dextrose] [minutes] Dextrose bug triaging meeting | 25th Apr 2011

[Rafael Ortiz]

On Mon, Apr 25, 2011 at 9:16 PM, Anish Mangal <anish at activitycentral.org>wrote:

> ==Attendees=
>
> bernie, tch, dirakx, icarito, m_anish
>
> ==Logs==
>
> First part of http://meeting.sugarlabs.org/sugar-meeting/2011-04-26
>
> ==Discussion==
>
> 1. m_anish : merge recent patches and (1) release intl build before
> edujam and (2) publish a set of updates for py
>
> 2. Spending time in eduJAM [input needed]
>        - getting started on XS
>        - big push for DX3
>        - testing sessions with people from ceibaljam
>
> ==Bugs/issues==
>
> 1. [dx3] http://bugs.sugarlabs.org/ticket/2800
> + add support for joining a hidden wifi network
>
> 2. [dx3] proxy-configuration **(more details/open bug)**
>
> 3. [dx3] http://bugs.sugarlabs.org/ticket/2799
> + register to XS automatically, in the background
>
> 4. [dx3] http://bugs.sugarlabs.org/ticket/2617
> + not included in the latest kernel yet  (too late for dx2)
>
> 5. http://bugs.sugarlabs.org/ticket/2155
> + suggested debug strategy: put a gdb on the process while it's
> running, then install debuginfos for the library where it stops
>
> 6. http://bugs.sugarlabs.org/ticket/2064
> + recommened debug strategy: recompile the python rpm with a lot of
> debug stuff turned on, then download a big file to trigger the bug
>
> 7. http://bugs.sugarlabs.org/ticket/2184
> + (related) http://dev.laptop.org/ticket/10131
> + suggested strategy: dunno. ideally, we'd want to break into the
> python interpreter and look into the palette object with
> introspection... but how could we do this? it's also hard to trigger
>
> + **tch**: could also be a palette logic bug, like destroying the
> palette object before time..
> + **alsroot, silbe** ideas please ;-)
>
> 8. http://bugs.sugarlabs.org/ticket/2023
> + pain point here, needs fixing :-)
>
> 9. http://bugs.sugarlabs.org/ticket/2367
> + lets try fgrose's recommended timeouts
>
> 10. http://bugs.sugarlabs.org/ticket/2593
> + dirakx looking into it (recommended porting f14 xulrunner before
> they switched to firefox4, grab it from the updates repository of
> fedora 14)
>
> About this one as I've tested on devel environments, upgrading to
 xulrunner-1.9.2.10-1.fc14.1.i686
seems to fix this issue. ;).

11. http://bugs.sugarlabs.org/ticket/2659
> + dirakx will look into is, he's looking into other browse issues as well
>
> <IRC>
> <bernie> dirakx1: i think it's done somewhere in the prefs.js file
> <bernie> tch: the disk cache should be LRU, with a limit on the size.
> <dirakx1> bernie:  hmm /me should dig on the js part of browse.
> <bernie> tch, dirakx1, m_anish: ah, i bet I know what could make it
> grow to 400MB! Try downloading a huge video :-)
> <m_anish> bernie, or lots of huge videos or songs :P
> <tch> i am pretty sure that was the case
> * bernie bets 1 bitcoin that xulrunner cannot download a file without
> going through the cache first.
> <dirakx1> but then it should warn the user somehow..
> <bernie> dirakx1: maybe i'm wrong... in sugar files are downloaded
> directly into the journal. you should try and see what happens during
> a download. where the file is being stored.
> </IRC>
>
> 12. [psst] cloud-journal, discuss during eduJAM
>
> It's be great to have inputs on some of the points above :)
>
> --
> Anish
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.sugarlabs.org/archive/dextrose/attachments/20110426/15f28491/attachment.html>