Re: Some code changes wanted in Rygel

["Zeeshan Ali (Khattak)" <zeenix gmail com>]

Hi,

On Mon, Sep 27, 2010 at 2:30 PM, Vincent Untz <vuntz gnome org> wrote:
> Le lundi 27 septembre 2010, à 09:30 +0200, Kjartan Maraas a écrit :
>> sø., 26.09.2010 kl. 23.52 +0300, skrev Zeeshan Ali (Khattak):
>> >   All these changes are to fix these 2 bugs:
>> >
>> > https://bugzilla.gnome.org/show_bug.cgi?id=626196
>> > https://bugzilla.gnome.org/show_bug.cgi?id=629378 (this patch was
>> > reviewed by Andre Klapper for me).
>> >
>>
>> Then I guess it's the first of those I don't really understand yet.

   There are device description files that provides basic information
about the (UPnP) device to clients. Please refer to UPnP Device
Architecture document for more details. However, what the XML files
are for is totally irrelevant here, whats relevant is that XML parsers
used by some of the devices in the market are broken and we need to
provide all XML files in a particular format for such devices to be
able to even discover rygel.

> What are
>> the risks involved in approving this patch?

  None that I can imagine for the changes in 8c0391d..14071c0.

>  - there's at least one unrelated fix ("media-export: Add null checks to
>   standard XDG folders")

  There was a bug about this on debian bugzilla. Now that you mention
it, this shouldn't matter to a normal GNOME desktop but rygel is used
by people on headless machines where this issue will realize.

>  - there are some commits that don't matter at all (update git ignore
>   list)

  Ah yes, for some reason i thought those patches were about
POTFILES.in and POTFILES.skip. So never mind these ones.

> My understanding is that the first three commits go together. How
> critical are those changes?

  The first one (206f019) fixes a leak introduced in previous release.
8c0391d..14071c0 are the fixes for IOP.

> How many devices will fail to work with
> rygel?

  AFAICT 3 different commercial products, however I can only be 100%
certain about one of them since the other two bug reporters haven't
managed to verify the bugs yet.

  So I modify my request to only include the following commits:

206f019..f6bfed4, 42a0f7b and f0b0127.

-- 
Regards,

Zeeshan Ali (Khattak)
FSF member#5124