Re: [linux-audio-dev] Laaga "default" audio I/O hooks

New Message Reply About this list Date view Thread view Subject view Author view Other groups

Subject: Re: [linux-audio-dev] Laaga "default" audio I/O hooks
From: Jarno Seppanen (jams_AT_cs.tut.fi)
Date: Mon Jul 23 2001 - 17:06:35 EEST


Paul Davis <pbd_AT_Op.Net> writes:

> Not quite sure about those names, but the point is well made. However,

Better names can be substituted, I didn't think very long when choosing them.

> I don't think this will quite work in the way you think. If the
> patchbay client has tied "out 1" to "output to driver 1", then "out1"
> is an input port (other clients send data there) and "output to driver
> 1" is an output port (other clients read data from it). hence, no
> other client can connect their own output streams to this 2nd port.

The intention is that clients change the graph that is downstream from the 2nd
port, using the standard laaga api. The patchbay functions as the "well-known"
port names in this sense.

Two options:
(a) for a parallel configuration, a third client can find the driver by
following wires from "out to driver 1", and then wire itself between "out to
driver 1" and the driver input port;
(b) for a cascade configuration, a third client can unwire the previous client
from "out to driver 1", and then wire its input to "out to driver 1" and
output to the client that was previously wired to "out to driver 1".

> 1) is this supposed to be a plugin or an external process?

A "plugin" in the same way that a driver is, i.e., constructed and destructed
by the engine and run in the engine stream. At startup, the engine should
construct a patchbay right after constructing the driver, and then wire the
driver to the patchbay.

> 2) you don't need to cleanup ports or anything else before
> calling laaga_client_close (). it takes care of all
> that for you.

Ok, that can then be removed.

-- 
-Jarno


New Message Reply About this list Date view Thread view Subject view Author view Other groups

This archive was generated by hypermail 2b28 : Mon Jul 23 2001 - 17:08:16 EEST