[or-cvs] r11434: Complete implementation of proposal 104: remove bandwidth hi (in tor/trunk: . doc src/or)

Nick Mathewson nickm at freehaven.net
Fri Sep 14 19:24:49 UTC 2007

On Fri, Sep 14, 2007 at 07:16:39PM +0100, Robert Hogan wrote:
> On Wednesday 12 September 2007 18:33:23 nickm at seul.org wrote:
> > Log:
> >  r14392 at Kushana:  nickm | 2007-09-12 11:04:20 -0400
> >  Complete implementation of proposal 104: remove bandwidth history lines
> > from router descriptors.
> >
> >

[Lines re-wrapped.]

> I recently completed a feature in tork that utilises the bandwidth
> history. I dug a bit more into how i would use the 'extra-info'
> option to retrieve the bandwidth history in future and have a couple
> of notes/questions:
> There's currently no way of getting extra-info for just one
> descriptor from the control interface. Tor either downloads
> extra-info for all descriptors or none at all. What would be the
> best way of making extra-info available on demand from the control
> interface?

Actually, it's probably best to not fetch the extra-info stuff for
clients at all.  Fetching on-demand is a risky feature, since it tells
the directories which servers you're interested in.  Are you sure this
is really something you need?

> Should tor download extra-info for a router if it receives a getinfo
> server/desc/fd/xxx for that router, in anticipation of a possible
> extra-info request?

Probably not; see above.

> Or would a setevents extra-info interface be better? If enabled,
> this would provide an asychronous extra-info message body for every
> server/desc/fd request received.
> Also, there are a couple of typos in control.c, one of which
> prevents getinfo extra-info/digest from working.

Will apply, thanks.

Nick Mathewson
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 652 bytes
Desc: not available
URL: <http://lists.torproject.org/pipermail/tor-dev/attachments/20070914/37da02c0/attachment.pgp>

More information about the tor-dev mailing list