Re: totem string freeze breakages

[Christian Rose <menthos gnome org>]

2005/10/24, Bastien Nocera <hadess hadess net>:
> On Sun, 2005-10-23 at 23:40 +0200, Christian Rose wrote:
> > 2005/10/23, Bastien Nocera <hadess hadess net>:
> > > Sorry, it's going to be 3). I will not revert those changes, but you
> > > were kind enough to warn the translators before I had the chance to.
> > > The string changes (2 new strings, and 1 string change) were necessary
> > > to fix bugs. I could gratuitiously bump version numbers, but I still
> > > consider this version of Totem to be stable, and fixes known bugs.
> >
> > Sorry, "I felt for committing some random bug fixes" is not a good
> > enough answer for gratiously breaking string freeze after several
> > months of freeze.
> >
> > Some questions to answer:
> > 1) What exact bug reports do you refer to? Please include links.
> > 2) For each of those bugs, what impact on translateable strings do
> > they have, and why do you feel that they need to go into the stable
> > branch instead of just in HEAD? (keep in mind that "I happen to
> > consider it a bug fix, that's why" isn't an acceptable answer)
>
> The strings addition were to add filters to Totem:
> http://bugzilla.gnome.org/show_bug.cgi?id=318350
> and
> http://bugzilla.gnome.org/show_bug.cgi?id=170226
>
> The string change (although it hasn't been mentioned directly in a bug)
> means that user now know that the location can't be played, rather than
> the type of file (this only affects the xine-lib backend which isn't in
> the Desktop release, per se).

It's five messages that are affected if I'm correct. This is very late
in the process, but then again, it's fairly simple messages, and it
seems that they actually fix bugs.
The actual string changes are fine with me, although obviously I am in
no way happy about how they happened.

Danilo might be of another opinion, though.


> > > As long as no major features are added to the current stable branch, I
> > > don't want to rob users from bug fixes because of a simple string
> > > change.
> >
> > We have allowed well motivated bug fixes that affected strings to go
> > in post freeze in the past (but obviously only after being given the
> > chance to review them first), so I really do not understand your
> > attitude towards the freeze and translators.
>
> You might want to double-check the original message you sent. Even
> though none of it was false, the wording leaves a bit to be desired, and
> makes me (who spend all day yesterday fixing bugs) feel like a guilty
> person. It's not helpful for either of us.

When a thing like this unfortunately happens, it's usually not at all
possible to be sure about whether it was intentional or an accident.
That's why I try to stick to the facts.
Granted, I honestly wanted to beleive that it was an accident in this
case -- your truly amazing number of commits on the affected day made
me beleive that you simply forgot to branch before starting new
development.  Too bad it wasn't that simple and innocent, but a real
and deliberate attempt to bluntly ignore a freeze with no respect for
the process, and, in the end, the work of translators.

Christian