🚀 go-pugleaf

I guess this route to c.s.a. works then.

A quick note to say that a number of people have commented on 
adding menu templates to the Glass format.  Someone even posted 
a suggested format (thanks!).  However, I'm going to stick to my 
decision not to include more than windows and icon names in the 
first version.  The reason is that although the proposed menu 
extension was good, it didn't go as far as it could.  I've seen some 
ideas from TMOTA re: configurable menus, configurable hot-keys 
for menu items etc, and I'd rather do it in one clean sweep, than 
add a half-hearted attempt and bodge it later on when more time 
is at hand.

To sum up, hopefully a menu chunk isn't too far away, but it won't 
be in the first version.  Of course, once the chunk has been 
defined, someone's got to write the SWIs to handle it... <shuffles 
feet and looks around hopefully>


As to the Glass relocatable module (I'm liking this idea more and 
more) I should point out that the module will be PD, i.e. anyone 
can do what the hell they like with it. More importantly,  the 
*complete* source for the module will be Freeware, i.e. freely 
distributable but if you want to use the code in your own 
programs (as opposed to just being a SWI client) you have to get 
permission of the author (via the moderator i.e. me).   


Hence I suggest anyone not happy with letting their code go out 
under these conditions does not bother starting to write any code 
for the Glass module.  The reason the source will be available is 
because it's very useful when tracking down bugs or determining 
just *exactly* what a function does when the documentation is 
less than clear (as I found with DeskLib - this is supposed to be a 
compliment, Jason!).


Initial versions of the Glass module will be in C, but hopefully 
speed critical parts of it will be translated into ARM by me, or 
whoever wants to do it.  BASS seem to like wading waist-deep in 
registers and condition codes so maybe they'll do it :-)

Also, I think I'll back down on icon-decoding at run-time - the 
overhead of having the feature *available* in the Glass module 
will probably be so small as to make deliberately not including it a 
bit dumb.  Of course, it's up to the programmer whether they use 
it or use constants...

Apologies for formatting; I'm using a proportional font and have to 
guess.

Tim