[Xfce-i18n] xfce4-time-out-plugin soon to be released

[Florian Rivoal]

On Fri, 10 Dec 2010 18:28:31 +0900, Jari Rahkonen  
<jari.rahkonen at pp1.inet.fi> wrote:

> 10.12.2010 00:00, Алекс wrote:
>> В Пн., 06/12/2010 в 19:01 +0900, Florian Rivoal пишет:
>>
>>> I'll soon make a release of xfce4-time-out-plugin. Due to recent  
>>> feature
>>> additions, most languages aren't fully translated. Thanks in advance to
>>> anybody who finds time to improve on this.
>>>
>> Please fix the awful message style in that app. The constructions like
>> _("%d minutes") cannot be properly localized. The proper way is using
>> ngettext("1 minute","%d minutes",n) as it handles plurals correctly. At
>> least use ngettext("DUMMY%d","%d minutes",n) if you want "1 minute" to
>> be a special case.
>>
>
> Please see Florian's mail[1] (and the associated patch/bug[2]) from last  
> month asking for input on the problem of having several plural gettext  
> translations in the same string. Basically it seems that the current  
> behaviour is considered the lesser of two evils, but a third option  
> might be preferable. I suppose one big problem with the current way is  
> that it doesn't really facilitate translation to languages with multiple  
> plural forms.



Yes, this is the issue I sent a mail a while about. Now body answered, so  
I assumed people didn't care either way, in which case it was easier to  
leave things as is, to avoid having to retranslate the languages which  
were already complete.

This mail is a clear sign that it is indeed an issue, so I'll fix it.

After giving a second look to the patch in that bug, I now see that it  
causes fewer issues that I originally though, so I'll use it as is.

> Oh wait, there's always the option of dodging the problem by dropping  
> this "verbose" form of the message altogether and only providing  
> something like '1:01:01 left' or '1:01:01 to go until next break'. It's  
> not like it is an essential feature or anything.

Yes, this is likely to have fewer issues, but I am not sure it would have  
no issue at all. So if I am going to change the code and force everybody  
who had already provided a translation to redo it, I want to be reasonably  
sure that what I pick instead will do the trick. As I said above, I now  
think the patch in bug #5611 is pretty much that silver bullet, so I'll  
apply it.

  - Florian