[Calendarserver-maintainers] Bug#841247: Bug#841247: calypso: UnicodeDecodeError when importing an .ics

[Mathias Behrle]

* Guido Günther: " Re: [Calendarserver-maintainers] Bug#841247: calypso:
  UnicodeDecodeError when importing an .ics" (Thu, 17 Nov 2016 20:22:46 +0100):

> On Thu, Nov 17, 2016 at 01:20:59PM +0100, Mathias Behrle wrote:
> > * Guido Günther: " Re: [Calendarserver-maintainers] Bug#841247: calypso:
> >   UnicodeDecodeError when importing an .ics" (Wed, 2 Nov 2016 21:30:21
> > +0100):
> > 
> > Hi Guido,
> >   
> > > Hi Mathias,
> > > 
> > > On Wed, Nov 02, 2016 at 01:14:02PM +0100, Mathias Behrle wrote:  
> > > > control: affects -1 + tryton-modules-calendar
> > > > tryton-modules-calendar-classification
> > > > tryton-modules-calendar-scheduling tryton-modules-calendar-todo
> > > > tryton-modules-party-vcarddav tryton-meta   
> > > > > From: Guido Günther <agx at sigxcpu.org>
> > > > > To: Jens Reyer <jre.winesim at gmail.com>, 841247 at bugs.debian.org
> > > > > Subject: Re: Bug#841247: calypso: UnicodeDecodeError when importing
> > > > > an .ics Date: Sun, 23 Oct 2016 12:28:37 +0200    
> > > >  
> > > > 
> > > > Hi Guido,
> > > >     
> > > > > The unicode handling for python 2.7 got broken in upstream commit
> > > > > 
> > > > >     https://github.com/eventable/vobject/commit/b3f9bbcf4cf222f0dda3ac29f96364c5d7ab5f16
> > > > > 
> > > > > Let's see if upstream cares at all. If not we should rather drop the
> > > > > python2.7 version for stretch.    
> > > > 
> > > > There are currently 6 Tryton modules affected as rdepends of
> > > > python-vobject. There are still some bits in the Tryton framework
> > > > lacking Python3 support, so it is currently not possible to switch the
> > > > whole Tryton stuff to Python3 (and I suppose this won't be the case for
> > > > stretch). Dropping the Python2 version of python-vobject would
> > > > seriously hurt those Tryton modules. So please let's find a way to keep
> > > > the Python2 version in the archive (for stretch).    
> > > 
> > > It breaks calyso as well which I'd rather see in the archive than
> > > removed.
> > > 
> > > The bug is fixable. Someone needs to sit down and cook a patch. I have
> > > it on the TODO list but work is piling up at the moment - but I hope to
> > > get it done til the end of the year.  
> > 
> > While I am currently not able to assist in making a patch I had a quite
> > intensive test with a complete Tryton calendar setup using vobject 0.9.3
> > under Python2. It is working like a charm. So it seems the Tryton modules
> > are not *directly* affected by this bug.
> > 
> > I would want to elaborate the different measures to take now (time is
> > running away and the autoremoval is scheduled in 27 days).
> > 
> > For me the severity of this bug is definitely important, not grave. I think
> > we should downgrade the severity immediately. Would you agree with this?

You didn't answer this question. Severity grave was set explicitely by you, not
by the original bug reporter, hence I am asking you to re-consider the
situation with the new information (see above). According to my tests this
severity is wrong, but severity important is just correct:

A bug which has a major effect on the usability of a package, without rendering
it completely unusable to everyone.

What's your opinion on this one?

> > As another measure to avoid the autoremoval of basically unaffected
> > Tryton modules it could be argued, that calypso is not ready for current
> > vobject and reassign the bug to calypso.  
> 
> This is not related to calypso. The ABI has changed significantly.

There was a jump from 0.8 to 0.9 in python-vobject, which in many softwares
means, that ABI breakage must be expected. 
 
> We're miles away from an autoremoval so keep calm ;)

It is not my way to keep calm, if problems can be solved. And you said,
that you only have time for a fix by the end of the year, which is way
after the autoremoval date.

Sorry if stepping on your toes, it is for sure not my intention to create more
work load for you. But this issue can be solved right now with respect to the
tryton-modules, why I am asking you again to agree on a correct bug severity.

Cheers,

Mathias


-- 

    Mathias Behrle
    PGP/GnuPG key availabable from any keyserver, ID: 0xD6D09BE48405BBF6
    AC29 7E5C 46B9 D0B6 1C71  7681 D6D0 9BE4 8405 BBF6