Rosetta/bzr (was: What is an Upstream?)

[Tim Morley]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 8 Nov 2006, at 21:11, kiko at async.com.br wrote:

>>> Christian:
>>> So from this paragraph, for a checklist of features that you'd  
>>> like to
>>> be able to better manage translations:
>>>
>>>    - Who contributed an approved translation (implemented)
>> Tim:
>> Yes, definitely; I believe this information is already recorded, but
>> I need to be able to use it too.
>
> Does the current display at
> https://launchpad.net/distros/ubuntu/dapper/+source/kdebase/+pots/ 
> kdesktop/eu/10/+translate
> serve your purposes? I'm talking about the text saying
>      Translated by:          marcos on 2006-04-05 10:06:10 UTC
> for the record.

Right, sorry, my mistake on that one; I haven't actually used Rosetta  
for a little while, and you're right, this information is displayed.  
I had a hazy memory of a problem about this, but which I expressed  
wrongly above. It's true that for any particular string, we can see  
who contributed which translations, but I'd like^W^W I need to be  
able to do a "view by contributor" and potentially "throw everything  
by this contributor away".

>>>    - Who approved a contributed translations (to be rolled out
>>> next week)
>
> We'll have this online shortly, and when we do you'll have similar  
> UI to
> tell this.

That's a step in the right direction, certainly.  :o)

> Should the whiteboard be editable by anyone? There are pros and  
> cons to
> that approach, and I'm not sure what the right solution is.

Depends to a large extent on the team structure. If we have a certain  
amount of granularity in team membership (owner, trusted member,  
normal member, new/moderated member, etc.) then I'd give full access  
to maybe the first two, and add-only access to the others. If we've  
only got "team members" and "non team members", then... I dunno.

Like CVS though, all changes would hopefully be trackable.

> Alternatively we could have a set of individual comments, but they  
> could
> only really be rendered in the zoomed view unless we did some fancy JS
> expansion. I guess that's true of the whiteboard too, though.

I have to confess I haven't given the UI much thought at all, and  
wouldn't really be able to help much anyway; it's not my speciality  
at all. I just want it to work.  :o)

> So this whiteboard, it would be added to the pofile export as  
> hashed-out
> comments? Is that how you see it working? Would individual comments
> added to the file be a better match?

As I said before, this functionality that you've called whiteboard  
was served quite adequately by us just adding comments to the .po  
files, which then show up in poEdit/KBabel and in fact in Rosetta  
too. The difference with the standalone editors is that translators  
can *add* to them, which is critical. (Again, not sure how the UI  
would work, but that's Someone Else's Problem).

That feature alone would make Rosetta an order of magnitude more  
useful to our team, and would probably be enough to persuade me to  
start using it again. The other stuff would be good too though.  ;o)

> Well, comments in the .po file and also in the web UI, right?

Yes, of course. The comments need to be visible to a translator  
working on the file (as I think they are now in Rosetta) but also  
*editable*, specifically add-to-able.


Tim

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (Darwin)

iD8DBQFFUnHb4qQfaYPSZYwRAl3PAKC5BmE5lTuxihfPqWoOPjLCJcVz+wCgsC9E
IE2xXG7xTfErBUaOONmhXAs=
=mhM5
-----END PGP SIGNATURE-----