On 10 February 2014 17:33, Olemis Lang <[email protected]> wrote:
> On Mon, Feb 10, 2014 at 11:02 AM, Saint Germain <[email protected]> wrote:
>
>> On 10 February 2014 16:48, Olemis Lang <[email protected]> wrote:
>> >> > In general I think the patches are in a good shape but the fact is
>> >> > that the current code for translations will also break the English
>> >> > version of the widget headers under certain circumstances ...
>> >> >
>> >> Do you mean the problem which is blocking is the "alert text rendered
>> >> outside container" ?
>> >
>> > No. I mean that the translation of widget titles (previosly explained
>> with
>> > fr+es) breaks the English site as well.
>> >
>>
>> Hum in that case I am a little lost: which problem are you referring
>> to specifically ?
>>
>
> see
> https://issues.apache.org/bloodhound/attachment/ticket/694/bh_widgets_x_150_i18n_titles.png
>
> I mean Spanish + French text , which also leads to translated + English
> text even if lang=en
>

Ok I understand what you mean now.

It's a mix of 3 separate problems:
1) Incorrect language switching (some part remains in the previous language)
2) Mainnav headers (i.e. force/Attempt a translation of the label
given by the user in base.ini)
3) "alert text rendered outside container"

So to answer:
1) Seems quite difficult, I don't think I can fix it in the near future.
2) I can fix this today
3) I have to be able to reproduce the problem first

>> (I may have misunderstood one of the previous problem, because I don't
>> see which one can break the english site if no translation is
>> available)
>>
>>
> no ... this obviously happens If translations are compiled into the
> corresponding catalog , but if translations are offered and they break the
> English web site then I'd consider that a regression and a blocker to
> release . Others might think in a different way , it's just IMO .
>

We can put a warning stating that compiling the catalog can break the
website ;-)
IMO it is safe to commit as it is (given point 2&3 above are corrected)

If we don't commit it, it will be quite painful to maintain (I've been
bitten by committing back and forth, as sometimes some util.pyc
remains and are taken instead of the directory util) and difficult for
people to participate as it is currently under bitbucket management (I
personally enjoy working with it, but as the official way to
contribute is through patch posted as diff in ticket, I have to
regularly post the patch/diff to the ticket).

Another opinion on the subject ?
If there is no other opinion, I will follow Olemis advice as he has
quite more experience on the subject, and postpone the commit after
the 0.8 release.

Reply via email to