Hi list,
I was just wondering what is the most efficient way of doing a HTTP query from PD. It should look like this:
GET /puredata/index.php?foo=bar HTTP/1.1 Host: www.sourcelibre.com
I tried with wget and curl and it's all right, but [shell] seems to make some processes last longer than it should be... and then I cannot make many queries per minute as I would like to. I kind of remember there was an external for this, but I would prefer use builtin objects.
Otherwise, I searched the list and the web, but you would agree that it's not easy to do a search about "HTTP" when all the URL on the web start with that... :-)
Thanks !
aalex
On Sun, 13 Nov 2005, Alexandre Quessy wrote:
I tried with wget and curl and it's all right, but [shell] seems to make some processes last longer than it should be... and then I cannot make many queries per minute as I would like to.
what do you mean they last longer? zombie processes?
does that [shell] thing allow you to start several processes at once?
you are using the "&" operator of the shell ?
I kind of remember there was an external for this, but I would prefer use builtin objects.
[shell] is not an internal (builtin) object class.
Mathieu Bouchard - tél:+1.514.383.3801 - http://artengine.ca/matju Freelance Digital Arts Engineer, Montréal QC Canada
Le 05-11-13, à 11:13, Mathieu Bouchard a écrit :
On Sun, 13 Nov 2005, Alexandre Quessy wrote:
I tried with wget and curl and it's all right, but [shell] seems to make some processes last longer than it should be... and then I cannot make many queries per minute as I would like to.
what do you mean they last longer? zombie processes?
I mean that, for some reason, [shell] tells me that an old process is still running. But, the thing is that this process should be done by now. Don't know why it stills lasts. Pd 0.39 OS X.
does that [shell] thing allow you to start several processes at once? you are using the "&" operator of the shell ?
Well, the following command is all right :
curl "http://www.sourcelibre.com/puredata/rpc.php?foo=bar&hello=world" &
I think it *is* better to use the & operator of the shell when using [shell].
Any better way to do this ?
aalex
hi
On 11/13/05, Alexandre Quessy listes@sourcelibre.com wrote:
I mean that, for some reason, [shell] tells me that an old process is still running. But, the thing is that this process should be done by now.
i admit i don't know the api here, so i may be completely off; but, have you tried sending an explit 'exit' to the shell?
-- \js [ http://or8.net/~johns/ ]
Alexandre Quessy wrote:
Le 05-11-13, à 11:13, Mathieu Bouchard a écrit :
On Sun, 13 Nov 2005, Alexandre Quessy wrote:
I tried with wget and curl and it's all right, but [shell] seems to make some processes last longer than it should be... and then I cannot make many queries per minute as I would like to.
what do you mean they last longer? zombie processes?
I mean that, for some reason, [shell] tells me that an old process is still running. But, the thing is that this process should be done by now. Don't know why it stills lasts. Pd 0.39 OS X.
am i completely mistaken that the very same problem ([shell] on OSX not finishing up correctly and leaving behind some zombie processes) has been discussed and resolved on 11th of november this year on this very list?
guys, this has been the day before yesterday!
please feel free to read this list even if the posts are neither originating from you nor are direct replies to your mails.
(if you have already deleted those mails, you might want to try the archives)
mfg.a.sdr IOhannes
Hi Alexandre,
I was just wondering what is the most efficient way of doing a HTTP query from PD. It should look like this:
GET /puredata/index.php?foo=bar HTTP/1.1 Host: www.sourcelibre.com
I tried with wget and curl and it's all right, but [shell] seems to make some processes last longer than it should be... and then I cannot make many queries per minute as I would like to. I kind of remember there was an external for this, but I would prefer use builtin objects.
Otherwise, I searched the list and the web, but you would agree that it's not easy to do a search about "HTTP" when all the URL on the web start with that... :-)
it might not be "most efficient" but how about using Python or ruby scripting?
gr~~~
-
\js -
Alexandre Quessy -
IOhannes m zmölnig -
Mathieu Bouchard -
Thomas Grill