Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
[e4-dev] Re: Some comments on XWT

Kevin McGuire wrote:
> e4 is an incubator and as such there can be many overlapping technologies 
> being explored (this is good!).  To that aim, I believe Hallvard was going 
> to contribute his work which AFAIK is being put through the IP process as 
> we speak (Hallvard is this correct?).  Plus we should not forget UFaceKit 
> which pre-existed e4 but is relevant nonetheless (just happens to be in a 
> different project).
> 
> So no there's been no decision, and in fact there won't be until a given 
> technology "graduates" to the main development stream (e.g. 4.0, 4.1, 
> possibly even 3.x).  This is what we've said all along but likely hasn't 
> been clear.  Once we ship 0.9 in August then there will be increasing 
> pressure to define what the 4.0 platform will be, presumably based on 
> maturation of code base, support, and interest from the community.

Thanks for clarifying this, Kevin. I didn't realize that there are more
than one declarative UI technologies being explored - my fault, I missed
the discussion on the mailing list, so I'm sorry for the noise.

>> This brings up the namespace topic again: do we really need namespaces?
>> If yes, then we should probably support them in CSS too. I know Angelo,
>> you want to have them anyway ;-), I'm still hesitating - I'd prefer to
>> keep the CSS simple...
> 
> I'm missing something... how does existance (or lack) of namespaces in XWT 
> and CSS related to each other?  Do you envision them referring to each 
> other?  Or are just saying that the same reasons that drive the need for 
> namespace in XWT probably also argue for them in CSS?

I just imagine that if a custom widget is being referred to using a
namespace in XWT and using a (maybe prefixed) name in CSS, this would be
a bad design. When there are conflicts between widget names, we need a
method of distinguishing between say, SWT Label and my.custom.Label. And
then, shouldn't this method should be the same in XWT and CSS?

Anyway, I see that namespaces cannot be avoided in XWT, so we can also
defer this issue in CSS until it happens to be a real problem and/or we
see any alternatives.

Ralf

> I'd like to avoid namespaces in CSS for as long as we can get away with it 
> since it then gets us on the hook for CSS3 which will require a different 
> parser.  If people come forward to do a new parser then fantastic but it's 
> a significant undertaking which I don't want to make a *requirement* for 
> the e4 CSS work.



Back to the top