On Wed, Jul 08, 2009 at 05:24:01PM -0700, Danek Duvall wrote:
> On Wed, Jul 08, 2009 at 05:19:48PM -0700, [email protected] wrote:
> 
> > On Wed, Jul 08, 2009 at 05:13:06PM -0700, Danek Duvall wrote:
> > > On Wed, Jul 08, 2009 at 05:05:06PM -0700, [email protected] wrote:
> > > 
> > > >         http://cr.opensolaris.org/~johansen/webrev-9929/
> > > 
> > > I don't understand why you split captive_portal_test() out into a wrapper
> > > and private backend.
> > 
> > I can re-coalesce.
> 
> Unless you actually need it, putting it back makes more sense.  I didn't
> know if you'd needed to split it out for some reason that went away as you
> finished up the changes.

Ok, will do.

> > > Does get_datastream() need a check, too?
> > 
> > Not sure that it does.  The get_datatstream is a generic call and it
> > isn't used by very many transport operations right now.  I can add one
> > here, but since the caller has to catch all of the other exceptions
> > coughed up by the transport, I expected that this is only going to be
> > used in special cases, and when it is used, the caller will know whether
> > he/she needs to run this test (or not).
> 
> Okay, either way.  I wasn't sure whether it was a "peer" of the other
> methods that had to do the call.

It's mostly a leftover for methods that needed to make a direct call to
the server to get an object.  I tried to eliminiate it's use, but I
think some weird method in the API still needed to do this to cat the
contents of a manifest remotely.

I'll provide an updated webrev in a moment.

-j
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss

Reply via email to