Ned Konz <ned at squeakland.org> wrote:
> On Wednesday 17 March 2004 12:52 pm, goran.krampe at bluefish.se wrote:
> > Yes, something is strange somewhere in there. There was some weird thing
> > happening earlier that I never really got a hang of - it seemed that
> > IIRC:
> > ¾¾¾¾¾¾¾¾SMSqueakMap class superclass subclasses includes: SMSqueakMap class
> > ..evaluated to false.
> >
> > Not sure if that has anything to do with this. Frankly - if anyone can
> > debug this and understand what goes wrong I would appreciate it.
> 
> I *think* that was a side-effect of a problems with class redefinition that (I 
> think) got fixed.

Oh, good. :) I must admit I kinda ignored it when I couldn't figure it
out.

> But I could be wrong on that. However, I think that if you are trying to 
> remove and then re-compile the same class within the same compiled block of 
> code, you're going to be referring to the old one, right?
> 
> So you probably need to break up such chunks with '!' in the change sets so 
> that they get compiled separately.
> 
> Or at least never refer to the class itself; do
> 
> (Smalltalk at: #SMSqueakMap) ...
> 
> everywhere.

I think I do that in these scripts. There is a "brutal hack" in the one
that removes SM1 - where I forcefully set the name of the class
SMSqueakMap to #SMSqueakMap - have no idea why that was needed. Perhaps
KCP related. 

regards, Göran

PS. Ned - I know at least you noticed that the current SM includes tons
of old methods. Ehrm. Cough, cough. Yep. I did the mistake of loading
SM2 into an image which had SM1 in it before continuing work. Duh. I am
cleaning it up all over again in the upcoming fix release for 3.7. Well,
what you don't have in your brain you just gotta have in your...
fingers.

PSS. Also mentioned co-maintainership of SM or the loader in the other
thread on the list. Holler if you are interested.