Re: [Tails-dev] [review'n'merge:1.1] feature/torbrowser-24.6…

Delete this message

Reply to this message
Autor: anonym
Data:  
A: The Tails public development discussion list
Assumpte: Re: [Tails-dev] [review'n'merge:1.1] feature/torbrowser-24.6.0esr-0+tails1
08/06/14 22:24, intrigeri wrote:
> Hi,
>
> anonym wrote (08 Jun 2014 18:17:59 GMT) :
>> Ticket:    None, but I can open one if it helps.

>
> Yes, please (in general, for anything I'm likely to do, a ticket does
> help a lot).


https://labs.riseup.net/code/issues/7379

>> This branch and APT suite upgrades the web browser to
>> 24.6.0esr-0+tails1, and the browser profile is synced with TBB's. Please
>> merge into devel and testing.
>
> I've had a look (to the extent that one can "review" this without
> re-doing all the work, which I won't do), and did not spot anything
> obviously wrong, *but* the missing .mozconfig and corresponding change
> to .gitignore.
>
> Looking at `git log --stat .gitignore .mozconfig', and the Git history
> just before 8f2d3a8f that introduced it, should help understanding why
> it's nicer to have it around.


Ah, I did that mistake when I was preparing 24.4.0esr-1+tails1 but at
least spotted it and fixed it that time.

> I noticed that's my fault, since I failed to add all the needed
> documentation back when I revamped a bit how we handle mozconfig's.


Well, shame on me for not noticing a mistake I had done before already.

> In short, if you had been importing a released Iceweasel, I think it
> should be fine, as the "Bring our changes back:" part, in "If Debian's
> iceweasel was not pushed to Git yet", is supposed to re-introduce
> these changes after it was removed by the new upstream release import.
> *But* these bits are only in the "4. New Iceweasel release" doc, while
> "3. New Firefox release" has nothing to bring back our changes to
> .gitignore, and thus the .mozconfig's are lost.


Right, but it seems we rarely will have the benefit of importing
Iceweasel at release time, when it's most important that nothing goes
wrong, sadly.

> This probably explains why things are the way they are, and
> apparently, that's been the case in the last release or two, too.
>
> I've reintroduced the .gitignore change in (6ab1d59,
> iceweasel:tails/master), for no particular reason but testing my
> tentative doc fix (b8bfc14, tails:master).
>
> I'll give the .deb's a try, and merge if happy... probably tomorrow.


There's no hurry for this as far as I'm concerned.

> May I assume that this passes at least our unsafe browser and torified
> browsing automated tests?


Before uploading the squeeze backport I tested it inside Tails for
around half an hour of real usage (including watching funny cat videos
on youtube :)) and all seemed fine. Now I've also run those two
automated tests on what will become Tails 1.0.1 and they passed.

As for the wheezy backport I've only tested it by installing it in a
running Tails 1.1~beta1 session and tested basic functionality. I now
realize that I didn't import the Tor Browser profile.

Cheers!