While it was 15/1/04 12:14 am throughout the UK, Paul Runde sprinkled little black dots on a white screen, and they fell thus:
> Walter wrote:
> 
>> 
>> What I'd like to have happen, rather than design a new GUI from scratch,
>> instead take the best one out there and port it to D.
>>
> Any suggestions as for which GUI library is the best one?  wxWindows?

Depends on what you want out of a library.

There are those that try to completely mask the OS API.  These have the advantage that they can be made cross-platform, but tend to sacrifice such native programming concepts as resource files for dialogs, menus and the like.

And then there are those that are geared to the API of a specific OS.  A prime example is Borland's ObjectWindows Library - presumably Microsoft Foundation Classes is the same sort of thing.  This has the programmer working with Windows API entities, and many of the OWL classes are simply wrappers around these entities.  In effect, a programmer-friendly interface to the API is provided that retains access to near-enough the whole API.

But either kind of library would tend to work in the same basic way - having a class for each GUI object (window, dialog, control, whatever else) and having the programmer define a method for each event (or dialog control) to be caught.

I am working on a GUI library in the style of OWL.  At the moment it's only really capable of producing a program that does nothing, but that's soon going to change.

Stewart.

-- 
My e-mail is valid but not my primary mailbox, aside from its being the unfortunate victim of intensive mail-bombing at the moment.  Please keep replies on the 'group where everyone may benefit.

Walter wrote:
> "Paul Runde" <prunde@consolidated.net> wrote in message
> news:bu4lvc$2qo1$1@digitaldaemon.com...
> 
>>Walter wrote:
>>
>>>What I'd like to have happen, rather than design a new GUI from scratch,
>>>instead take the best one out there and port it to D.
>>>
>>>
>>
>>Any suggestions as for which GUI library is the best one?  wxWindows?
> 
> 
> wxWindows is certainly a contender.
> 

wxWindows is the best choice in my opinion.
1. API is well designed easy to remember.
2. It supports a lot of GUI widgets and a lot of utility functions.
3. Cross-platforms. It has supported many platforms.
4. Its license that allows to make a commercial applications.
5. Borland has supported some works.
6. It's possible to build wxWindows with your great DMC so why not D :P
...

I really would like to see wxWindows as a part of D GUI toolkit.
Just my 2cents.

Regards,
Parinya

P.S. Sorry for my bad English.

A general way of applying any infix operator between each pair of adjacent items in the array, would be desirable.  In Haskell and other functional languages this is usually called foldl or foldr I believe.

something like:

(+)(a[]*a[]) // squared length of vector a

Should be able to get it to work with min and max also.  There's already
sum(), min(), and max() functions that take arrays, overloaded for many
common types.  But that's not a general solution.

Sean

"Stewart Gordon" <smjg_1998@yahoo.com> wrote in message news:bu1cgg$cnm$1@digitaldaemon.com...
> While it was 1/1/04 12:03 am throughout the UK, Mark T sprinkled little black dots on a white screen, and they fell thus:
> > I really only have 1
> >
> > 1) That the D language version 1.0 is finalized and released in 2004.
The window
> > of opportunity closes a little each year for getting a new language off
the
> > ground and adopted by a decent sized developer community so it doesn't
die a
> > premature death.  I can't use it at work unless it becomes popular.
>
> Good idea I reckon.  Especially:
> (a) implementation of array arithmetic
> (b) getting std.file.read and std.file.write working
>
> Re (a): which are actually supposed to be supported?  DMD 0.77 reports a
> mixture of coding errors and not implemented errors when I try them.
> - array op scalar
> - array op array
> - scalar op array
> - array op= scalar
> - array op= array
> - ...
>
> on which operations?  Should each work on all of +, -, *, / and whichever others?
>
> 2) Array aggregate properties like sum, product, maximum and minimum. This would have its potential for parallelisation, and also lead to a nice, concise notation for the dot product of vectors:
>
> (v1 * v2).sum
>
> Of course putting maximum and minimum in raises the issue of avoiding confusion with the .max and .min properties on data types.
>
> 3) A Mac OS X implementation, I guess.
>
> OK, so maybe the chances of 1, 2 and 3 being completed before I finish my PhD are fairly slim....
>
> Stewart.