Since I moved and renamed files, and migrated from PortAudio v18 to v19, I
can only surmise that you did not do a fresh install, i.e. rename your
existing csound 5 directory and do a fresh checkout from CVS. Last night I
did such a fresh checkout for Cygwin and everything built and ran. I will
do the same for Fedora tonight, and let you know what happens.

It seems clear that it will take at least a few days before the build
system is stable and reliable on Cygwin and Linux. You could either try to
help get it going, or use the no-make-am branch, or email me any changes
you want in CVS and I will commit them for you.




Original Message:
-----------------
From:  jpff@codemist.co.uk
Date: Tue, 24 Feb 2004 17:23:08 +0000
To: csound-dev@eartha.mills.edu
Subject: [CSOUND-DEV:4137] Csound5 building


I currently fail in make with errors in both Engine/entry1.c and
OOps/fprint.c.  diskin.c is a total mess (looks like missing the last
commit i made)
  Actually many other errors....

/home/jpff/Sourceforge/csound5/Opcodes/bowedbar.c:332: multiple definition
of `opcode_init'
.libs/physmod_la-physmod.o:/home/jpff/Sourceforge/csound5/Opcodes/physmod.c:
962: first defined here

Opcodes/pitch.c:30:19: pitch.h: No such file or directory
Opcodes/pitch.c:31:19: uggab.h: No such file or directory
make: *** [pitch_la-pitch.lo] Error 1

Opcodes/sfont.c: In function `SoundFontLoad':
Opcodes/sfont.c:101: syntax error before `->'
Opcodes/sfont.c:123: `p' undeclared (first use in this function)
Opcodes/sfont.c:123: (Each undeclared identifier is reported only once
Opcodes/sfont.c:123: for each function it appears in.)
make: *** [sfont_la-sfont.lo] Error 1

 gcc -DHAVE_CONFIG_H -I. -I. -I./H -DBETA -IH -I/usr/local/include
-I/usr/local/boost -DBETA -IH -I/usr/local/include -I/usr/local/boost
-falign-loops=4 -falign-jumps=4 -Wall -falign-loops=4 -falign-jumps=4
-ffast-math -fomit-frame-pointer -finline-functions -funroll-loops
-DWITHx87 -O2 -g -Wall -O2 -g -Wall -MT spat3d_la-spat3d.lo -MD -MP -MF
.deps/spat3d_la-spat3d.Tpo -c Opcodes/spat3d.c  -fPIC -DPIC -o
.libs/spat3d_la-spat3d.o
In file included from Opcodes/spat3d.c:32:
H/ftgen.h:28: parse error before `->'
make: *** [spat3d_la-spat3d.lo] Error 1

and so forth.

Before doing this I did a cvs update which i assumed updated any of my
files which were out of date.  So whence to I go from here?  The code
for sndfile does not seem to be present, other random things are
broken, and a make csound complains

So for the foreseeable future I am out of the CVS, as both csound and
csound5 have become way out of step with my changes.  I have fixed
dumpk locally, but I am trusting that someone else will fix the CVS.

==John ffitch


--------------------------------------------------------------------
mail2web - Check your email from the web at
http://mail2web.com/ .

But I did commit diskin.c and libsnd.c and the versions on the CVS are 
not they.  So why shoudl I attempt to commit anything else?

The version of entry1.c I have declares itablew1 but uses itablew and that 
has nothing to do with anything I have done.  A complete checkout is a lot 
of work which frankly I do not have time for at present.

==John ff

Hi John,

I'm looking at the CVS resource history in my editor and it shows the 
last version checked in for libsnd.c by you was 2/16/04, and two more 
commit after that by Michael on 2/21/04 and on 2/23/04.  The comment 
from Michael's 2/21/04 checkin is:

Many changes to enable a single Makefile.am, to build the ustub library, 
to get plugin opcodes automatically loading and running.

The 2/23/04 check has no commit message.  The diff between the 2/21/04 
and 2/23/04 has only to changes from testing for HAVE_ALSA to 
HAVE_LIBASOUND. 

If there was another commit after 2/16/04 that you did, then it did not 
get in.

For diskin.c, the last check-in by you I see is also 2/16/04.  There one 
checkin after that by Michael on 2/21/04.  The commit message is:

Many changes to enable a single Makefile.am, to build the ustub library, 
to get plugin opcodes automatically loading and running.

and the only diff I got was the addition of:

#if defined(HAVE_CONFIG_H)
#include "config.h"
#endif

When doing checkouts, are you using ext or pserver?  Sourceforge has 
been known to delay mirroring between the ext server (the ones for 
developer access) and anonymous pserver access, and maybe that could be 
a factor?

steven


John ffitch wrote:

>But I did commit diskin.c and libsnd.c and the versions on the CVS are 
>not they.  So why shoudl I attempt to commit anything else?
>
>The version of entry1.c I have declares itablew1 but uses itablew and that 
>has nothing to do with anything I have done.  A complete checkout is a lot 
>of work which frankly I do not have time for at present.
>
>==John ff
>
>
>
>  
>

On Tue, 24 Feb 2004, steven yi wrote:

> Hi John,
> 
> I'm looking at the CVS resource history in my editor and it shows the 
> last version checked in for libsnd.c by you was 2/16/04, and two more 
> commit after that by Michael on 2/21/04 and on 2/23/04.  The comment 
> from Michael's 2/21/04 checkin is:

Yes, but the version I recovered fron CVS was not the same file as the one 
I committed.   So I am rather loath to join in what seems liek a lottery

> For diskin.c, the last check-in by you I see is also 2/16/04.  There one 
> checkin after that by Michael on 2/21/04.  The commit message is:

I changed diskin on 17 February at 06:49 Universal time


> When doing checkouts, are you using ext or pserver?  Sourceforge has 
> been known to delay mirroring between the ext server (the ones for 
> developer access) and anonymous pserver access, and maybe that could be 
> a factor?

ext is the usual system as far as I understand this stuff.  Never used 
pserver

The version I have ok entry1.c is plain broken.  I did not change it and 
certainly would not have changed function names in that bizarre fashion.


I wait until someone fixes this; I am very anoyed that the time I spent on 
fixing audio has apparently vanished.  I hope someone for whom this system 
works is changing dumpk (an error which I suspect may be my fault for 
which I apologise)

==John ff