Re: [gtk-list] Re: announce: yet another gtk+ C++ wrapper (no caps as Owen suggest)

[Kenneth Albanowski <kjahds kjahds com>]

On Wed, 9 Sep 1998, Karl Nelson wrote:

> 
> I personally side with Laurent on this one.  
> 
> > >- it's confusing to the users
> > 
> > i do not think that gtk+ users will be confused, they are not are babies or
> > stupid and can judge what is good and what is garbage :-)
> 
> But there will be some confusion especially when it comes to end
> users.  Why should there be three shared libraries on system, two of
> which are providing wrappers for the same thing?  The question
> becomes how different are they. (not just in writing philosophy,
> but also in end use.)  I had enough trouble trying to sort through
> all the widget sets much or less figure out which wrapper I wanted.
> 
> (By the way, which is garbage and why?  Gtk-- seems to work
> well for me and you seem to feel VDK is good, so we
> aren't talking about something as simple as one being garbage.)

Let me just point out, as a comparison of sorts, that Windows itself has
about four coding levels commonly available (raw graphics, the widgets and
windows built into Windows, and the MFC and VCL C++ libraries from
Microsoft & Borland, respectively.) Note that VCL and MFC are both
surviving quite well, and are quite different in approach. Both are, at
least in part, just "wrappers" over the plain Windows widgets.

Borland thought that MFC was icky enough that they should write their own
library, and they did (actually, they've written several. VCL is the
survivor.) The result has survived very nicely, and is quite pleasent to
work with (from a Windows standpoint. It's embarrasing some things that
even an immature Gtk does that VCL cannot -- and vice versa, of course.) 

I'm not about to say that Gtk-- is icky, but if someone wants to write
their own wrapper, please give them the benefit of the doubt, and the
opportunity to demonstrate their ability.

> On the other hand, my limited understanding of VDK is that it
> has largely different design goals.  Therefore, we should
> certainly not dismiss it.  However, some discussion of its
> scope may be needed.  If one or the other projects
> expands to eliminate the need of the other, people are
> going to be unhappy.  
> 
> I saw VDK appear on the Gtk web pages before its announcement here.
> I was quite surprised, as there had been any talk of a competing
> C++ wrapper before.  And as I help with Gtk--, I would have liked time
> to comment earlier.  (Maybe I missed it.)	

Let's try "additional", instead of "competing". There's no need to go
zero-sum.:-)

> Having more than one choice is only good if the choice is real and
> distiguishable.  If the two are largely indistiguishable, it will 
> quickly decay to personal preferences, marketing, and non-technical 
> reasons.  Eventually one will die and likely the developers will simply
> disappear.  I would perfer to avoid competing packages if possible.
> So the arguement that two is better would be moot, because in the
> end there would be just one.

Agreed. If all else fails, someone can just sit down and combine the two. 
(For some odd reason, I find that sort of thing fun.) 

-- 
Kenneth Albanowski (kjahds@kjahds.com, CIS: 70705,126)