Modules ? (Re: Squeak book !)

Karl Ramberg squeak-dev at lists.squeakfoundation.org
Mon Sep 16 21:30:00 UTC 2002


danielv at netvision.net.il wrote:
> 
> Karl Ramberg <karl.ramberg at chello.se> wrote:
> > I'm not at all up to date with the modules...
> > Could you give a brief explanation of the problems with the 3.3a stuff?
> 
> * It's a big jump.
> * No part of it isn't completely finished yet.
> * If anyone is making it ready, except for Andreas, I don't know about
> it.
> 
> > > Of course any work we do on this will be compatible with (and even more
> > > useful in the context of) the future module system modus operandi.
> > >
> > > If people think this is a good way to go, I have some ideas on how we
> > > can make progress in this direction.
> > A brief explanation here, too.
> 
> 3.3a tries to give the following -
> * Namespaces
> * Automatic loading of packages from web repositories
> * A code model that supports class extensions (specifying that
> Object>>beep is logically not part of Object's module, but of another).
> 
> The problem is that that's alot to give, and it'll take concentrated
> effort for that to happen.
> 
> In order to benefit from a modularized image, for example, to have
> shrink scripts that work (or module that unload), we first have refactor
> every package in the image so that it can be unloaded.
> 
> This is the most important thing - nothing bad will happen if we don't
> have namespaces for a while more, and we can always add automatic
> handling of loading, *but the reason we began this whole effort in the
> first place is to cut things up clean*. This requires only a code model
> that supports class extensions.
> 
> It just so happens we already have one in the image, already mostly
> integrated into the tools, and it's mostly a matter of agreeing to use
> it.


Thanks,
Karl



More information about the Squeak-dev mailing list