[Mono-list] Native Mono UI widgets

Jacobo Polavieja jacobopolavieja at gmail.com
Mon Jul 16 17:33:52 EDT 2007


Adrien Dessemond escribió:
>> My concerns are about:
> 
> 
>> 2. GTK+ is a nice library and Gtk# is a good binding, but again, we
>> are tied to the computer infrastructure.
> 
> Whatever you use, you are tied to the computer architecture somewhere :-)
> See this as a strength, not a weakness.
> 
>> Having a Mono specific UI library, maybe built on top of the current
> 
> For what reason ? I don't see one. Things are heavy enough, trying to make
> them more is not a winning solution and worse, brings additional confusion
> to your users as they have to accommodate to another UI "philosophy".
> 
> Stacking layers may be nice approach but this has a cost unfortunately and
> in a managed world it can be huge.
> 
>> Having a Gtk# implementation built on top of System.Drawing could also
>> be valuable in the same way.
> 
> Beyond code duplication considerations, this would lead to a precise
> defect of Swing : too slow, too heavy. :-/
> 
> 
> Kinds regards,
> 
> Adrien
> 
> 
> _______________________________________________
> Mono-list maillist  -  Mono-list at lists.ximian.com
> http://lists.ximian.com/mailman/listinfo/mono-list
> 

Isn't something similar what would need to be done if we wanted to
support WPF?
I don't know if the mono team plans to implement that (I thing there's
no mid-term plan for start hacking on it), but is mostly like having to
write another toolkit from scratch, am I wrong?
Maybe that's the reason we're never gonna have it. It would be nice to
have it, anyway.
Maybe the implementation of Moonlight could lead to WPF in the end?

>From my humble point of view... it seems like GTK+ is getting too hard
to update and add new things due to it being written in C or other
issues. It seems like Qt is much more better designed, and so could be
WPF, but this can be completely false as I'm not an expert in this topic.
I don't say to have it written in C# or other .Net language as it would
be too heavy for a toolkit, but maybe C++ and a good planification from
the beginning.
Again, I don't know if this is actually possible or worth the effort.

Cheers!


More information about the Mono-list mailing list