I think this is a very good approach and if it is done correctly it will give 
TERP a boost in the way it is seen/experienced by new users.

Some more thoughts..

Location of Changes: Fabien, are you planning to change this in code and in the 
Basic XML files or are you using an localisation/l10n file for lets say en_UK 
language?
The former would mean that some translations get lost, with the later approach 
you have for english the same advantage as for the other language packs. Simple 
real time editing and changing - together with the users maybe. This could help 
to get it modified to a customers usage/like for some installations. The 
original wordings inside the code are called pidgin en for example and are 
somehow like msg-IDs.
 
Help translations: Make help text translateable and editable in the views. This 
way also a translation of the mouse -over help is available to the user.

Further advantage is that later one can dynamically create a manual from the 
detailed field descriptions maintained in the help text, it would also work as 
a kind of data repository.

Central translation database: Sorry, this is maybe overkill - but I dream of 
it.  Instead of keeping the different localised versions in seperate CSV files 
- which is messy and error prone why not hold all in one huge table. The 
structure would be the same as inside the TERP database, maybe with some 
extensions to tack change history.

>From here automatically extracts can be done (into CSV or JSON) and some 
>interactive access - either via TERP or via a seperate web frontend (a small 
>Django App?) - is given. 

In the interactive access there is a kind of list view - showing the original 
(I did call it pidgin) value together with the help text explaining the meaning 
in its context. 

I imagine it could look like 

pidgin value (Source)  - original help (eg. en_FR)
lang A value - lang A Help (eg. en_UK) 
lang B value - lang B help (fr_FR)
.....
...
lang M value - lang M Help (eg. en_US)

which is similar to the view in TERP - just blown up

As the english localisation is also only a additional record, it is shown, and 
here the value and the help content can be modified once better wording pops up 
/ evolves. A discussion board (like a very simple comment blog) would ease 
collaboration on difficile words and help to track the changes.

Of course there will be access issues and spam and stupid changes - but it 
might be a way to get a very powerful repository - from where simple manuals 
could be generated semiautomatically - just give for each field the name and 
its detailed description.

You could even think about autotranslation of standard expressions - once 
agreement is that for a given pidgin word 'X' the translation int lang A is 'Y' 
and into lang B is 'Z' - then next time the same pidgin is used  all ocuurences 
for A and B are are filled automatically.

Some queries enabling special views like the one in 
http://tiny.be/download/menus.txt could be later added - this way giving always 
convenient and up-todate reports of the structures.

Hope this makes sense and doesn't sound to confusing,

rgds Günter





_______________________________________________
Tinyerp-users mailing list
http://tiny.be/mailman/listinfo/tinyerp-users

Reply via email to