I have fixed up Top/threads.c to compile on Linux. I was able to build and
run Csound and CsoundVST in csound 5 on Fedora Core 1. Everything still
compiles and runs on Windows as well.

The portaudio DLL for Windows in CVS is now built for ASIO. Thanks to the
PortAudio people for getting the ASIO thunks working for MinGW, even if they
have not done anything about blocking read and write yet.

You PortAudio people, note that there is code in Csound5/Top/pa_blocking.c
that uses mutexes to create a blocking read and write from asynchronous
ASIO.

Csound people, please note that it is now necessary to upgrade SWIG to
version 1.3.21 to build CsoundVST on any platform.

Breaking a performance on Windows with control-C can leave PortAudio hanging
with an eventual blue screen of death. Suggestions for how to trap control-C
on Windows (works fine on Linux) would be appreciated.

----- Original Message ----- 
From: "John ffitch" 
To: "Csound Developers Discussion List" 
Sent: Tuesday, August 10, 2004 9:35 AM
Subject: [CSOUND-DEV:5175] Re: Preliminary implementation of low latency
audio for Windows


> ...but it does mean that Linux Csound5 idoes not build
> ==John ff

Should be the same, CTRL-C is just SIGINT. I use this entirely conventionally to 
stop playtback in my "paplay" soundfile play program using portaudio.

In essence:

static int file_playing;

void playhandler(int sig)
{
	if(sig == SIGINT)
		file_playing = 0;
}

int main()
{
   signal(SIGINT,playhandler);

  /* start:*/
  file_playing = 1;
   /* playback loop*/
   while(fileplaying){
     do_portaudio_stuff(...);
   }
   close_portaudio_stuff();
   return 0;
}

This works unchanged on both Windows and OS X. SO it should jolly well work on 
Linux too! Probably need to do more for a re-entrant multi-threaded monster, but 
hopefully the principle is the same. Often the trick is to trap CTRL-C and do 
nothing, so the user can't stop a program that way.


Richard Dobson

Michael Gogins wrote:
> I have fixed up Top/threads.c to compile on Linux. I was able to build and
> run Csound and CsoundVST in csound 5 on Fedora Core 1. Everything still
> compiles and runs on Windows as well.
> 
> The portaudio DLL for Windows in CVS is now built for ASIO. Thanks to the
> PortAudio people for getting the ASIO thunks working for MinGW, even if they
> have not done anything about blocking read and write yet.
> 
> You PortAudio people, note that there is code in Csound5/Top/pa_blocking.c
> that uses mutexes to create a blocking read and write from asynchronous
> ASIO.
> 
> Csound people, please note that it is now necessary to upgrade SWIG to
> version 1.3.21 to build CsoundVST on any platform.
> 
> Breaking a performance on Windows with control-C can leave PortAudio hanging
> with an eventual blue screen of death. Suggestions for how to trap control-C
> on Windows (works fine on Linux) would be appreciated.
> 
> ----- Original Message ----- 
> From: "John ffitch" 
> To: "Csound Developers Discussion List" 
> Sent: Tuesday, August 10, 2004 9:35 AM
> Subject: [CSOUND-DEV:5175] Re: Preliminary implementation of low latency
> audio for Windows
> 
> 
> 
>>...but it does mean that Linux Csound5 idoes not build
>>==John ff
>>
> 
> 
> 
> _______________________________________________
> Portaudio mailing list
> Portaudio@techweb.rfa.org
> http://techweb.rfa.org/mailman/listinfo/portaudio
> 
> 
> 

Does this mean that csound for Linux is working again?  I have not had
any messages from Michael since he said he "I will deal with this tonight"
==John ffitch

I tried compiling tonight and was able to make it to the end using the
latest threads.c Michael checked in.  I tried running it on a few csd's
and it seemed to be running as it did before.

steven

On Wed, 2004-08-11 at 02:25, jpff@codemist.co.uk wrote:
> Does this mean that csound for Linux is working again?  I have not had
> any messages from Michael since he said he "I will deal with this tonight"
> ==John ffitch
> 
> 

Actually the answer is no it does not compile on Linux.  I did an update 
this morning and the thing fails to compile in the identical way as 
yesterday morning.  So either I wait or I have to try to understand why it 
is compiling this code at all.

==John ff

On Wed, 11 Aug 2004, stevenyi wrote:

> I tried compiling tonight and was able to make it to the end using the
> latest threads.c Michael checked in.  I tried running it on a few csd's
> and it seemed to be running as it did before.
> 
> steven
> 
> On Wed, 2004-08-11 at 02:25, jpff@codemist.co.uk wrote:
> > Does this mean that csound for Linux is working again?  I have not had
> > any messages from Michael since he said he "I will deal with this tonight"
> > ==John ffitch
> > 
> > 
> 
> 
> 
>