Re: Moving from SVN to git - finally!

On 09.01.2021 at 15:41, Andreas Heigl wrote:

> Hey.
>
> Am 09.01.21 um 15:10 schrieb Andreas Heigl:
>> Hey Folks.
>>
>> Am 06.01.21 um 16:09 schrieb Christoph M. Becker:
>>> On 06.01.2021 at 14:52, Davide Pastore wrote:
>>>
>>>>    2. As of now IT documentation build is broken but it's strange because
>>>>    the error reported there ("Entity 'language.types.pseudo-types' not
>>>>    defined") doesn't come out when I try to launch the same command on my
>>>>    machine, since it works fine. Could you please help me in this sense?>
>>> Did you pull doc-en?
>>
>> I've just had a look into the history and it looks as the move from
>> building the docs from SVN to building them from git has also broken the
>> build for the italian translation.
>>
>> I'Ve jsut verified that it was *only* the build for the italian
>> documentation that weas affected. All other documentations were either
>> building without issues or with the same issues as before. Only in the
>> italian one seems to have been something awkward happening during the
>> transition.
>>
>> I'm currently investigating it.
>
> I have found the culprit.
>
> Since commit 206fedd1bce098ac3b6ee12de93d9f506cb21c26[1] from 2015 the
> italian translation is not part of the ACTIVE_ONLINE_LANGUAGES any more.
> Due to that the git-repo of the translation is not updated any more on
> the machine building the docs every 4 hours. And that brought up the
> issue as the update-job is iterating over the content of the
> ACTIVE_ONLINE_LANGUAGES but the build job is iterating over the result
> of an `ls` command. So the italian language gets not updated any more
> but it still gets build.
>
> The change that happened during the migration to git was that with SVN
> there was *one* svn update that updated *all* languages (even the not
> active ones). This needed technically to be modified to update all
> relevant languages. As relevant I took all those that are currently
> marked as Active. But as there were actually some commits coming in in
> the italian translation that one was also transfered to git despite it
> being internally marked as inactive.

Thanks, Andreas, for looking into this!

> So now the question is: Do we continue to keep the italian language as
> inactive? Or do we activate it?

The point of marking translations as inactive is to avoid presenting
translations which are grossly out-dated, since there is no info about
that easily available to readers, who may then read out-dated
information without being aware of that.  So, basically, we need to
review the rev-check info to decide which languages should be marked
inactive (I think DE currently should be).

In the long run, it would be great if PhD could inject warnings about
outdated pages/sections, but that's not easily accomplished.

Cheers,
Christoph

> Cheers
>
> Andreas
>
> [1]
> http://git.php.net/?p=web/php.git;a=commitdiff;h=206fedd1bce098ac3b6ee12de93d9f506cb21c26
>>
>> Cheers
>>
>> Andreas
>>
>