Hi all--

	Tim writes:

 > > ...it's a new way of doing the whole socket/serial/midi/file thing
 > > and so involves people spending the effort to learn it and dump the
 > > old code and habits.

	Colin responds:

 > I think "dump the old code" is the crux of the issue here. The
 > problem isn't so much that it's a new way of doing things as that it
 > can't co-exist with the old way of doing things. This means that
 > switching to Flow can't be done gradually, and so any package
 > maintainer that moves his package to use Flow instantly makes himself
 > incompatible with all other packages that use files, sockets, streams
 > etc.

	With all due respect, that is simply not true. Have you looked at Flow?

	The Flow support exists quite happily with the old support in the same 
running system. I released a system which demonstrates this in August 
2002 (Flow 2a1, http://netjam.org/flow/releases/current ). I did rename 
some of the old classes, but that is a matter of personal taste (one 
could rename the Flow classes instead). In the forthcoming Spoon 
version, class name conflicts are irrelevant. (I grant that I haven't 
released that version yet, but I think it's relevant to this discussion.)

	A community could indeed stage a gradual transition from one to the 
other. My own preference would be for something on the rapid side 
(having seen the rather jagged attempt at gradual change with Michael 
Rueger's network rewrite), but it's not a requirement.

	Stef responds to Colin:

 > Ok I was always a bit concerned that "we ignored" the work of craig
 > and this is good that you mentioned that. I remember now that this was
 > the key point.

	If it was, that's a shame, because the rationale wasn't valid.


	thanks,

-C

-- 
Craig Latta
improvisational musical informaticist
www.netjam.org
Smalltalkers do: [:it | All with: Class, (And love: it)]