Re: [fuse-devel] HTTP-FUSE Xenoppix(Experimental version) Release

[Miklos S. <mi...@sz...>]

> The post to plan9-ML causes interesting discussion.
> FUSE is compared to v9fs.
>    http://lists.cse.psu.edu/archives/9fans/2006-January/thread.html

Thanks.

One of the more interesting observations about FUSE is that it lacks
support for interrupting operations.

Not the abort style 'kill -9' interrupt that already works, but a
graceful interrupt which notifies the filesystem about the need to
cancel the operation.

How do people feel about this?  Are there situations where this would
be useful?

One problem related to this is that some applications assume that
filesystem operations are uninterruptible, which is generally true for
disk filesystems, but not necessarily for network filesystems.

Miklos