You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Per convo with @joemfb, unix.c is not guaranteed to prevent disastrous outcomes from malformed %ergo effects coming from Clay, such as attempts to write to .urb or .run.
Desired case is that there is, at least, a hard-coded and unified list of paths that Vere will not write to as the result of a Clay effect.
Such a unified list of paths should also include a .host directory so that providers have a safe place to put pier-specific files that won't be touched by the runtime.
The text was updated successfully, but these errors were encountered:
Rather than an explicit whitelist, ensuring that clay mountpoint names (ie, desks) conform to (sane %tas) is probably the simplest way to accomplish this. That would need to be done in each of the desk-relevant effect handlers in unix.c: u3_unix_ef_ergo(), u3_unix_ef_ogre(), u3_unix_ef_*() ...
Per convo with @joemfb,
unix.c
is not guaranteed to prevent disastrous outcomes from malformed%ergo
effects coming from Clay, such as attempts to write to.urb
or.run
.Desired case is that there is, at least, a hard-coded and unified list of paths that Vere will not write to as the result of a Clay effect.
Such a unified list of paths should also include a
.host
directory so that providers have a safe place to put pier-specific files that won't be touched by the runtime.The text was updated successfully, but these errors were encountered: