RE: i18n

(I sent this two days ago but somehow didn't receive it back - so excuse me
if you get it twice...)

Hi Mark,

I am actually surprised that you are so dismissive of the idea of
server-side translation completely, given netWindows mission to avoid
expensive server traffic. I agree with you that exclusive client-side
translation (or parsing in general) is nonsense. But here are my arguments
for doing the job at least partially on the client.

It takes a while for nW to start up (caching or not) and it cannot know
which language the user wants an application to be in - locale and all (the
user can be sitting in a Bombay internet café). While nW could still reload
all ist open windows in the desired language, I think that is a waste of
time if it can be done differently.

More importantly, if I dislike templating  and want to write a true
client-side application in Javascript (which I do), then I need such a
translation service, since I want to be able to write messages to the user
without sending the content first to the server and have it translated....

So I still believe that post-load language switching makes sense, and if it
can be implemented without lots of code, then why not loading the same file
that the server scripting language uses for translation into the browser
(they are usually not that large). I might be wrong as in the case of
XML-RPC, but I think I will go ahead and let you know if and how it
works.... Thanks for the input anyways!

C.

> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]] On Behalf Of Mark Anderson
> Sent: Friday, September 05, 2003 3:31 PM
> To: netWindows development discussion list
> Subject: Re: [NW-Devel] i18n
> 
> 
> i guess what i'm missing is why a client-side solution is
> preferable or otherwise attractive.
> 
> server-side templating solutions generally support (or at
> least have to confront) dynamic content based on user 
> preferences, client device, site customization, and a variety 
> of other dimensions. language is just one such customization.
> 
> what would be next? javascript-based switching of currency
> and temperature units?
> 
> furthermore, http itself supplies a preference for language
> in the standard.
> 
> i feel the same about pure client-side lang switching as i
> do about pure client-side css switching -- a cool hack perhaps, but 
> not particularly necessary AFAIK.
> 
> in any event, one approach would be to use the standard
> xml:lang attribute, and on the client exclude any elements 
> (and their children) which possess an xml:lang attribute and 
> its value differs from the desired one.
> 
> -mda
> 
> 
> _______________________________________________
> The netWindows developers list: [email protected]
> http://netwindows.org/mailman/listinfo/devel_netwindows.org
> 
> 


_______________________________________________
The netWindows developers list: [email protected]
http://netwindows.org/mailman/listinfo/devel_netwindows.org