Re: Proposal to change l10n system (WAS: DDTSS is down)
On Thu, Jan 27, 2011 at 1:33 PM, Martin Bagge / brother
<brother@bsnet.se> wrote:
> With regards to DDTSS/DDTP I think there must be something done in the
> foreseeable future. As far as I can tell there are unacceptable
> maintenance tasks that rely on some individuals. And adding that the
> system in some cases breaks so bad that some translation efforts are
> lost (or did I get that impression wrong?) is devastating for the moral
> of translators.
The DDTSS isn't great. It was written by me as an alternative for the
mail interface. It wasn't really designed for this sort of usage. As
for whether efforts were lost, it's hard to say. What generally
happens is that a new key is inserted into the database and because
the index is corrupt it can't be found again. When I rebuild the DB I
ignore the index so those lost entries should reappear. OTOH, maybe
it's possible to really lose stuff, I don't know.
(For reference, it's a Berkeley DB v1 database, btree)
Back in the time the DDTSS was integrated into the DDTP it was
expected that Pootle would eventually cover this case. I gather later
it was decided/discovered that DDTP wouldn't work in Pootle and AFAIK
there's not been a plan B.
> I haven't really worked with this (partly because when I did read some
> of the documents it was a too steep learning curve. Might have been
> solved since it was years ago) but doing work here is something I have
> seen as a must. If I just had the time to volunteer... but with help?
I think it would be useful to work out what the next step should be.
If that is that DDTP will forever need its own system, then we need to
spec out how that should look. If people think the interface is OK but
the implementation sucks, we can work on that. But I agree something
needs to be done.
Have a nice day,
--
Martijn van Oosterhout <kleptog@gmail.com> http://svana.org/kleptog/
Reply to: