[V3dot10] Rio status
keith_hodges at yahoo.co.uk
Fri Mar 9 15:57:07 UTC 2007
>> I don't try this, but if we going to change what are doing now, I go for
>> FileMan-mu.35.mcz or any newer.
Edgar, I am not suggesting changing what you are doing now. This is
looking into the future.
Fileman is simply a layer (another 30k of source) on top of
FileDirectory, as such there is no simplification provided by adopting
Fileman. Adding Fileman makes the problem of simplification harder,
because there is yet more functionality included, and lots more code to
> Stef, if I don't was wrong you said something about this a time ago ?
> Flow is .... flowing
According to Craig a loadable Flow is under development by a third party.
Rio is providing references for manipulating to local and remote file
objects, it doesn't do the streams itself, so as such I think that it is
orthogonal to Flow.
I am posting here because this is the place where I first received
encouragement to have a go at it, and this is the place where people who
are interested in modularizing and shrinking Squeak are gathered.
There are other areas of the Kernel where it has been suggested that a
kernel, and rest of, division could be made. Kernel-Chronology being
one. Rio may also serve as a test case of this principle.
So, if I am designing Rio with the sightest potential of being adopted
as a piece of kernel, for the image shrinkers, I figured that this is
where that particular activity would be best reviewed/discussed. If not
then off to squeak-dev, but I would wait until I feel that rio is
actually usable as a module for third parties, (a few more tests).
Squeak-dev which is not always a sympathetic audience.
All new Yahoo! Mail "The new Interface is stunning in its simplicity and ease of use." - PC Magazine
More information about the V3dot10