Hi :-)

Google applications: well these are a workaround for me, since we are working on a collaborative system to collect and share lexical data. We have already been testing it with the Agrovoc (FAO) data. It just doesn't make sense to start coding on some intermediate application until Ambaradan works well. We want to take a demo version with us to Randa, so that's why we are actually down-under with too many things on this end.

Latest news: I know that we have multimedia working (video and audio) but it still needs plenty of work on that end. There is one UTF-8 problem which is driving us crazy ... I can't explain details since I am not a programmer :-)

I am simply not publishing where we exactly stand, because it would involve many questions and discussions where I cannot answer and Bčrto, who is coding does not have time.

Who has time to read up can do this on our website:
http://www.voxhumanitatis.org/content/ambaradan-owm2-storage-engine
(here you will still find MySQL mentioned - the system was migrated to Postgres)
The first page is actually the most relevant one.
If you wish to have all the text on one page, without having to click through the various chapters:
http://www.voxhumanitatis.org/book/export/html/11

This is also the reason why, right now, I am not pressing to get a converter for the tables I have - the python scripts and bash commands work fine for what is needed right now even if I have to make looooong stupid one command after the other .sh scripts etc.

My own spreadsheets became too large to be digested by google (which initially was quite comfortable) and that's the reason why I am now working in OOo base and basically do mainly copy/paste work from data I get in manually. If anyone has some magic sql command so that I can read in contents from other tables in my table ... well ... it would be a nice feature - if not: I just go ahead doing things one by one.

The problem is "getting the data" from people and that is where we have to make "input" easy for them. Right now, for many particular languages a google form is the easiest way for me and them to handle.

All for now :-)

Cheers, Sabine

On Mon, Apr 26, 2010 at 10:09 AM, Frederik Gladhorn <gladhorn@kde.org> wrote:
Hi,
I'm currently away and totally out of the loop, so I thought, I'd just throw
in how I think vocabulary should be collected:
I would like to see it happen as openly as possible, if we don't give people
the possibility to work on it they won't... So I like the wiki approach. But I
think wikis are way to unstructured. It is hard (Sabine had someone who did
that, and it is overly complex) to get data out of wiktionary for example.

I do not like the idea of using the google applications for several reasons. I
don't think the interface really fits the task - vocabulary is more complex
that just two word mappings. I think it lacks any of the great advanced stuff
that applications could make use of. And it lacks a way of easily structuring
the data so that it can be split into lessons without major trouble. This is
prone to breaking. Finally I don't like that it's proprietary software. But
even if there was a free spreadsheet alternative that wouldn't make the cut
for me.

In my opinion what is needed is a web app (maybe later accessible from within
parley for example because I do not enjoy using web apps most of the time).
I think the data should be stored in a read database and there should be a
specific front end that makes it possible to edit vocabulary in a meaningful
way - including grammar things such as word type, conjugations etc.

One more note about books: it is sad, but experience shows that editorials
that produce school books tend to insist on some silly copyrights for word
lists because of their meaningful pedagogical arrangement... this seems to be
a gray zone of copyrights... the individual translations are luckily not
copyrightable (how insane would that be). I only looked into this a bit in
Europe.