[avahi] Running client-test in Avahi 0.6.10 on Solaris
lennart at poettering.net
Sun May 14 16:45:41 PDT 2006
On Wed, 10.05.06 09:20, Padraig O'Briain (Padraig.Obriain at Sun.COM) wrote:
> I made the changes in the first attachment to get client-test to run on
> Solaris and got the results in the second attachment.
> Most of the changes I made were because printf crashes when a NULL
> argument is provided for %s on Solaris.
I guess Solaris' printf sucks, then. ;-)
The patch you posted is reversed? And it contains some superfluous
changes (like printing the value of AVAHI_BROWSER_CACHE_EXHAUSTED)? If
you clean it up I will definitely merge it. In fact I already merged
the avahi_address_parse() issue, since that's a real bug even on
Linux. (see http://avahi.org/changeset/1217)
I don't think client-test is a good test program for your purposes. We
probably should remove it entirely from the distribution. Instead use
the the better tested and supported end-user utilities we ship:
avahi-browse, avahi-publish-* and so on.
> I noticed that the name of a service is NULL in
> avahi_service_browser_callback. Is this what should happen? I was
> expecting to get "Lathiat's Site".
I assume that you're talking of these two lines?
> SERVICE-BROWSER: Callback on 28c58, interface (-1), protocol (-1), event (2), name (NULL), type (_http._tcp), domain (NULL), data (omghai3u) 2
> SERVICE-BROWSER: Callback on 28c58, interface (-1), protocol (-1), event (3), name (NULL), type (_http._tcp), domain (NULL), data (omghai3u) 2
These are events 2 and 3. Those are AVAHI_BROWSER_CACHE_EXHAUSTED and
AVAHI_BROWSER_ALL_FOR_NOW. These are "meta" events that inform you
that all entries from the cache have been read or that no more
services are expected to be read in the near feature. They do not pass
actual data. For more information see:
In short: yes this is the intended behaviour.
Lennart Poettering; lennart [at] poettering [dot] net
ICQ# 11060553; GPG 0x1A015CC4; http://0pointer.net/lennart/
More information about the avahi