RE: CF: Suggestion and comments

["Kristofer M. Bosland" <krisb@seattleu.edu>]

On Mon, 8 Apr 1996, Ken Woodruff wrote:

> 
> Brian Thomas writes:
> > A Windows client would probably achieve the stated goal of 
> > expanding the popularity of CF more than any one other 
> > software objective I know of. Ken Woodruff (you still 
> > out there Ken??) once stated that he was interested in
> > programing the client. 
> 
> Still out here, still interested, haven't started.  I'd guess it's still
> possible I might want to take up this project later this year (summer
> looks to be relatively slow).  I'm open to suggestions as to approach,
> for example do you think it's worthwhile to try to develop a common
> API for the user interface and network routines that will allow us to share
> most of the client code across UNIX/X and Windows versions or should
> we take the client/server communication protocol as a starting point and
> develop a client for Windows from scratch using more standard (from a
> Windows perspective) programming and user interface techniques?   
> 
> --Ken

	I'd recommend a strong definition of the client/server
communication protocol as a start.  If you make one that is more text than
binary (i.e. move object7 x7 y30\n instead of \0x13\0x07\0x07\0x1e) I'd be
willing to start on a Tcl/Tk client.  Tcl/Tk has been ported to Windows
and the Mac, so with a little work, a Tcl/Tk client should be very general.



-Kris Bosland