We are just testing some solutions (like http://qooxdoo.org/)
We will not use a template engine as we think it must be a real web
application.
But it is not the main priority, as you can see with the demo server
(http://www.tryton.org/demo.html) the GTK client works well through the
web. And we have a single exe for windows.
> I've seen some plans of using genshi templates in the past somewhere
> but can't find it anymore...
>
I don't think that genshi will provide the needs.
--
Cédric Krier
B2CK SPRL
Rue de Rotterdam, 4
4000 Liège
Belgium
Tel: +32 472 54 46 59
Email: cedric...@b2ck.com
Jabber: cedric...@b2ck.com
Website: http://www.b2ck.com/
I worked on it to try to provide all the feature to convert GTK
application into javascript. But there is a lot of work to do and others
from the pyjamas community don't seem to be interested with the GTK
part.
But I think if we write a client in javascript, his structure will be
very closed to the GTK client, to be easy to maintain the both.
We have a windows client that doesn't need to be installed. You can put
it on a USB key, it is only 8.5M (or even download it each time).
And I think we will provide a Linux and MacOS version also.
> Like the idea to have a native client for that all but imo its
> neccessary to have some, maybe restricted, access to monitor things.
Why not making a full featured client?
> Also I don't see a reason that you need too adapt the whole user
> interface to web ui the hard way like using qooxdoo.
> What was wrong with the turbogear variant of oerp?
It is too slow. The main issue is that the server must simulate for
every user the client behavior. So the load on the server will increase,
I find it is smarter to put those stuff on the client host.
And we have a minimal way to discribe a form (the xml view), it is
faster to send just this information to the web client instead of an all
page.
By the way, qooxdoo (or an other toolkit) will not be much harder than GTK :-)