Re: VCL.NET revisited...

I.P. Nichols wrote:

IIRC most of your prior comments (thoughts) relate more to Delphi
users rather than to trying to attract new non-Delphi users and I
don't dispute your point of view about advantages for existing Delphi
users.

That's probably because I *am* an existing Delphi developer (among
other things), and I can only speak reliably from my own experience.
Still, I think the things that are good for existing developers will
also be good for anyone using Delphi.


I have always readly achnowledge this difference in UI rendering.

Me too. :)


I have never been exactly happy with the slower UI rendering of
WinForms but I'm hard put to blame all the sluggish response of .NET
apps soley to slow rendering, often poor programming is as much or
more to blame.

Really? A lot of the non-UI C# and Delphi for .Net code I have to deal
with runs well for me. It's the UI painting itself that seems to be
the bottleneck. Someone with a bunch of time on their hands should do
some benchmarking.

Speaking of which, I sometimes think it would be interesting to add
..Net functions to the FastCode B&Vs, at least where they make sense.
This would give a much clearer picture of how they measure up against
Delphi's RTL and the best that FastCode has to offer and put a lot of
speculation to rest. I have a sneaking suspicion that some of the .Net
routines will hold up pretty well. Crap. There goes a bunch of my
free time. :(


With Vista and it's new Desktop Window Manager (DWM) there will be
rendering performance improvements for all apps especially in way
WM_PAINTs are handled.

That would be cool. Have you noticed that WinForms applications paint
better running on Vista compared to running on XP? Do VCL applications
seem to paint any faster or slower?


That's not exactly true, with .NET 2.0 WinForms have acheved a level
of maturity where extensive further development probably isn't
warranted but WinForms are being enhance and an example is "Crossbow"
in .NET 3.0 that makes it simple to have hybrid WinForms with both
WinForm controls and WPF rich graphic content on them.

Adding Avalon support isn't what I had in mind for new WinForms
development. No hardware acceleration or feature work. Any other
examples? And I'd be a little suspicious of *any* development company
claiming that their framework is mature enough that it didn't need any
more work.


Of course DevCo is committed to continue VCL development same as
Microsoft is committed to further develop the .NET framework but what
continued VCL.NET development do you expect will result in improved
VclForms? I assume the much desired unicode support will be one
improvement.

The .Net framework, yes, but not WinForms.

I'm not sure what DevCo has planned, but I expect ongoing evolutionary
and revolutionary improvements going forward. A lot like what we've
seen so far, except possibly at a faster pace.

Robust Unicode support in the native VCL is definitely on my wish list.
Better support has been added incrementally over the past several
versions, but I think it needs to be much more complete. And quickly.
This isn't an issue with VCL.Net, which is as Unicode capable as the
rest of the .Net framework.

The .NET framework as a well th
ought out and extremely rich class
library ain't exactly what you call chopped liver and IMO the
decision to retain .NET 2.0 as-is and build .NET 3.0 on it is a sign
of it's growing maturity.

You're right. I really didn't mean to imply that the .Net framework as
a whole was chopped liver. I should have said that they would have
access to *another* framework.

As for continuing to use the C# 2.0 compiler for .Net 3.0, I can only
hope it's at least partly because someone at Microsoft is trying to
reduce breakage of user code between versions. You might have noticed
that this is a pet peeve of mine.


I'm aware of and don't disagree with your position on these
advantages for Delphi users, but my POV is how well will the VCL.NET
features serve to attract new non-Delphi users?

Ultimately, I think the advantages of Delphi as a whole is what will
draw new users. The VCL/VCL.Net is just one part of this. It also
contributes to some of Delphi's other compelling (to me) advantages
like code compatibility (across versions and platforms). I have to
think that things like protecting the substantial investment in
someone's code base over multiple versions would be important, even if
they are only interested in .Net development. And I think this would
ring especially true to VB6 developers, even if Borland was very late
to that particular party. Another example is the one way migration
from 1.1 Winforms applications to 2.0. Simultaneously supporting some
customers who have upgraded to 2.0 others still running 1.1 just became
more... "interesting".

DevCo just has to let people know what options are available.
Enthusiastically, if possible.

--
Regards,
Bruce McGee
Glooscap Software
.