Hi list Sems that the trouble comes from the "name.local" way of addressing on Macs which is to avoid, and this for a reason I ignore being by noway a network expert. Works fine (means immediate response in case of absent / unexisting host) with either names without the .local extension (as in the netsend help patch) or with numerical IPs as suggested by Martin. Thanks JM
Le 17 oct. 2012 à 14:35, Martin Peach a écrit :
Are you connecting to a named address or a numeric one? Connecting to a hostname causes a lookup to be performed, which blocks until a name server replies. If you already know the numeric IP address of the remote machine then it won't need to be searched for, and, under UDP, any messages will be sent even though the remote host isn't there.
Martin
On 2012-10-17 05:31, Jean-Marie Adrien wrote:
Hi list Sorry to come up on the subject again, but there must be someone having solved this :
when trying to maintain a network of computers alive, it can happen that a machine tries periodically to connect to a distant host which is offline (because it is dead). In that case sending the message [connect> with a non available host to [netsend 0 or 1] results in a freezing of PD for 15 long seconds, which stops any interaction in realtime, which is kind of serious ! happens on OsX 10.6 /10.7 Pd 42.5 Is there a way to configure something to go around this ? Thanks JmA
Le 7 oct. 2012 à 16:33, Martin Peach a écrit :
On 2012-10-07 05:06, Jean-Marie Adrien wrote:
Hi list When given a connect message with a host which is absent / offline, the netsend object takes one minute or so to respond "Bad Host" on Mac OsX. During that time, Pd is not responding to edit, close open and other GUI commands, Pd seems dead... and wakes up again only when the Bad Host message appears. Is there a way to go around this ?
Maybe use UDP instead of TCP: [netsend 1]
Martin
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list