[noble/shotwell] Package Phasing Stopped

[Nathan Teodosio]

I think more users will be affected by the newer bug (crash when cropping) than
the older bug (OOM after hours of running slide-show) so I think we could wait,
but of course that's just a wild guess.

I already put the follow-up 2084777 in SRU format and attached the debdiffs; As
soon as Pluck is fix released I'll subscribe ubuntu-sponsors again.

Am 22/10/2024 um 08:47 schrieb steve.langasek at ubuntu.com:
> On Thu, Oct 17, 2024 at 10:29:44AM +0200, Julian Andres Klode wrote:
> > On Wed, Oct 16, 2024 at 03:47:42PM GMT, Steve Langasek wrote:  
> > > Hi Nathan,  
> 
> > > On Mon, Oct 14, 2024 at 10:30:26AM +0200, Nathan Teodosio wrote:  
> > > > Hi, this error is no regression as per the table of the bug:  
> 
> > > > 23.04   23.10   24.04
> > > > 336	234	5  
> 
> > > > So if nobody disagrees I'll file a merge request to except it.  
> 
> > > This shows that the error in question was seen in both 23.04 and 23.10. 
> > > However, looking at the error page, it appears that 0.32.6-1ubuntu2 is the
> > > first version of this package where the error was reported *in noble*. 
> > > Since this is an SRU, obviously this is not the first version of the
> > > package in the noble release; and what we care about in an SRU is avoiding
> > > regressions for users who have already installed that stable release.  
> 
> > > So no, the above table does not demonstrate that it's not a regression,
> > > and I would nack a merge request for this with only the above as
> > > rationale.  
> 
> > This all seems sort of expected looking at the commits. The original
> > commit fixed the crash but introduced the memory leak. So it was
> > reverted for causing a regression (and not believing it fixed a crash),
> > but of course now the crash is back.  
> 
> Thanks for this analysis.
> 
> Based on this, I'm divided on the question of whether we should override the
> phasing; which is worse, the memory leak and an eventual OOM crash, or a
> crash that happens right away?
> 
> But https://bugs.launchpad.net/ubuntu/+source/shotwell/+bug/2084777 was
> opened for this particular crash and has already been uploaded to plucky,
> with intent that this also be SRUed to noble and oracular; so maybe we
> should just push ahead with that and side step having to decide between two
> bugs.
>