On Tue, May 03, 2011 at 10:51:24PM +0200, Julien Claassen wrote:
Just want to explain a few things. The idea of osc_query was
to write a program that prints out all possible osc paths.
In a second step I could add a #type-signature, or #documentation
part to all those paths, issue a query with that address, and
print the replies of all those as well.
This is how I do it know:
- make a new lo_address for the specified port
- set up the server (now the server will also listen on that address)
- add a method for any replies
- send one query for the root path
Now when I get the reply in that method, I send out
new queries from within that method for all the subdirectories
I received. The same method will receive all subsequent replies
as well, and send out new queries for the subdirectories of the
subdirectories. I will do this until it receives replies which
indicate terminal nodes, and then prints them out.
What you do now in osc_query, is send out one query for //#documentation,
that is, for the documentation of the root node.
If you want to send one /path/#documentation query to test if
it works with the server, it would be a lot easier to use oscsend
with the patch I send. Oscsend only sends one query, and with the
patch it will print out one reply to that query.
If that works with your server, I could modify osc_query to print
out _all_ type-signatures and documentation that your server provides
Hope this clears things up a bit,
Linux-audio-user mailing list