[physfs] 2.0 wishlist...

Gaetan de Menten ged at bugfactory.org
Tue Sep 21 11:14:41 EDT 2004


On Tue, 2004-09-21 at 06:31, Ryan C. Gordon wrote:

[...]

> - Find some way to relax or remove the security model so apps can have
> roughly the same codepath for the game itself and external tools. I
> think this probably needs a little more discussion.

This is just an idea but I think a nice way of handling that would be to
have some kind of "configurable function" (I mean a pointer to a
function) which would take a path as parameter and would return whether
the user is allowed to "go there" or not. A default function would have
to be provided to match the current (restrictive) behavior. The tool
programmer could then provide another function to have a relaxed
security. Alternatively PhysicsFS could also provide another function
for relaxed security that the tool programmer could use if he wants to.

I don't really have any other comment about the other items on your
wishlist since the other "mailing-listers" have already raised all I had
to say.

By the way, when I wanted to check if I had already suggested something
or not (I had indeed ), I looked for a mailing list archive but couldn't
find any on the PhysicsFS page... I eventually found it by googling for
it (there is a link to the archive on freshmeat).

Another "problem" about the webpage is that the link to the web
interface of the CVS repository 
(http://cvs.icculus.org/horde/chora/cvs.php?rt=physfs) seems to be
broken. Gives a nice error 404.

See you and keep up the excellent work on this library (and many of your
other projects),
Gaetan.




More information about the physfs mailing list