Re: Icons of program

[Toshio Kuratomi <badger prtr-13 ucsc edu>]

I think you're doing a pretty good job of explaining this now.  What do you
think of this:

SimpleVFS has two components: the meta-info component and the Simplified File
Space component.  I think the meta-info component could be generalized into a
parallel tree architecture which could then be used for any available
filesystem.  SimpleVFS would then only be a Simplified FileSpace which VFS
aware apps (like gmc) could hook into.

SimpleVFS would have an advantage over regular fs's in the meta-info area
because it can only be accessed via VFS (since the VFS takes care of meta-info
in the parallel tree, SimpleVFS meta-info should usually be up to date with
respect to what's actually in the tree.)  However, the meta-info model would
be the same whether the file was in SimpleVFS or not.

If this sounds okay to you, we could think about how the two components shoudl
be designed to work in more depth.  (Should real files be allowed in
SimpleVFS?  Should SImpleVFS be a special directory under the user's home dir
by default, or the user's home dir?  How much overhead is there in a parallel
tree?  How much in a database?  Could apps needing efficiency sync a db with
the parallel tree?  How would we keep the files up to date? and so on...)

-Toshio

-- 
badger  \"The Difference between today and yesterday is not so much what has
@prtr-13 \ changed between then and now as what I hope to change by tomorrow."
.ucsc.edu  \~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~

PGP signature