brad at danga.com
Sat Aug 11 17:40:57 UTC 2007
Looks good. Just this part scares me a bit in the POD docs:
- $mogc->new_file($key, $class, $opts_hashref)
That's no longer an allow calling convention? I'd hope so.
Fix that (the code or the docs) and feel free to commit.
On Thu, 9 Aug 2007, John Berthels wrote:
> Thanks very much for the review.
> > -- the whitespace isn't consistent, either internally
> Sorry about that. Hopefully fixed.
> > -- document, wherever users are likely to see it,
> Some more notes in the POD in MogileFS::Client.
> > -- document also the requirements for the webserver?
> > webdav, PUT, etc? does your new test pass with
> > default mogstored+perlbal? or only your (which?)
> > webserver?
> The client tests now poke the store and skip (with appropriate
> message) if it doesn't support DAV MOVE and partial PUT. The POD now
> mentions the requirements too.
> > -- I don't like this internal method:
> > + ($list, $fid, $dmid, $key) = $self->_get_paths($args);
> Fair enough. The patch now doesn't touch cmd_get_paths, and copies out
> what it needs from there into cmd_edit_file.
> > -- document also above cmd_edit_file that this is experimental.
> > Clean those up, post another version here on the list, and mail me
> > separately about commit (http://danga.com/commit.bml) and when it's
> > good, you can just commit this stuff yourself.
> Thanks, will do. New version of patch attached. Should apply cleanly
> to r1109 with 'patch -p1'. I've also updated the client and server
> CHANGES in line with your commit guidelines above.
More information about the mogilefs