[wx-discuss] wxTNG: compatibility with wx2 API

[Vadim Zeitlin]

On Fri, 17 Feb 2006 07:39:35 +0100 Stefan Csomor <csomor at advancedconcepts.ch> wrote:

SC> I was referring to the step by step in the following sense :
SC> 
SC> - when I implement a new wx3-dc on mac, I have to write a wx2-wx3 bridge
SC> in order to test it, but if I have bridged colour, brush, dc then I can
SC> test it already, I can now add an implementation for msw, use the same
SC> bridge test that again etc ..

 So you'd write a new wx::DC implementation but the only way to use it, for
foreseeable future, would be via existing wx2 wxDC. I don't find this very
interesting because, again, I'm mostly interested in using new API directly
and this is not going to be possible before wx3 is finished.

SC> >  OTOH I'm still thinking of
SC> > 
SC> > 1. better API
SC> > 2. improved code quality
SC> > and surely in these aspects wx3 won't be affected by wx2 
SC> > codebase at all.
SC>  
SC> but the code quality in wx3 does not help if string manipulations in wx2
SC> are faulty.

 I don't think they're faulty (or if they are, we have to fix them anyhow).
It's just that the API could be improved.

SC> >  As I said, I do expect wx2 to stay alive for quite some time 
SC> > (I'm speaking about years here, maybe 5, more likely 10). But 
SC> > this doesn't mean wx3 will still be using it in a couple of 
SC> > years -- e.g. I fully expect/hope to have a native .NET 
SC> > implementation of wx3 relatively soon.
SC> 
SC> who will be able to benefit ? IIUC only people writing against the wx3
SC> API - or also wx2 users ?

 I hope both. But for this it should be possible to use both wx2 and wx3 in
the same program, which is why I put it as one of important goals in the
wiki page.

SC> of course having a clean API with pluggable implementations for .NET,
SC> wx2, qt etc. could help acceptance by boost, but this would be an
SC> entirely new framework and wx2 users would never be able to benefit
SC> from it

 wx2 would be interoperable with wx3 version using wx2 backend. Of course,
the existing wx2 code would need to be ported to wx3 to profit from the new
features and new backends, but this is not only unavoidable IMO but also
good as this will motivate people to move to wx3. But without forcing them
to do it all at once.

SC> I don't think making everyone happy is key,

 Well, IMNSHO it is. We speak a lot about wx users but we, wx developers,
should also be motivated to develop wx3. If we're not, I don't believe this
project can succeed.

 E.g., personally, I don't want to embark on a project which is not going
to bring any fruits at all before several years... For me it is important
to have something working relatively soon is important. If we can achieve
this -- in whatever way -- I'll be happy. I.e. if you can propose a
development plan which would allow us to write applications using wx3 API
soon, I'd be glad to accept it. Do you have such plan?


SC> IMHO consistency is,

 Of course, but nobody puts this in doubt AFAICS.

SC> but IIRC both approaches still have one area in common 
SC> 
SC> - craft a new API
SC> 
SC> this is not touched - or should not be touched by the way it is
SC> implemented. 

 Yes, this is important. But you can't craft an API in thin air without
thinking about implementing it. E.g. the idea of supporting multiple
backends probably imposes quite a few constraints. This is why I'd like to
have a _broad_ agreement about how it will be implemented from the
beginning.

SC> I don't know how much agreement there is on the fact that this API
SC> should stand where possible on the shoulders of boost etc, instead of
SC> corresponding wx2 classes.

 This is a separate -- and important -- topic. To stay with this one, let
me summarize the state of discussions so far:

1. we all agree that we should provide
 a) new API not encumbered by wx2 compatibility
 b) allow the existing wx2 applications to use wx3 in some way

2. we don't agree about how to do 1b in practice

Maybe we can still work together on 1a even though, but I think it is still
necessary to come to an agreement about 1b too. Again, any (concrete)
proposals are very welcome.

 Regards,
VZ


---------------------------------------------------------------------
To unsubscribe, e-mail: wx-discuss-unsubscribe at lists.wxwidgets.org
For additional commands, e-mail: wx-discuss-help at lists.wxwidgets.org