Message written at 23 Apr 1999 13:21:13 +0100
--- Copy of mail to pipe@algonet.se ---

>>>>> "Anders" == Anders Andersson  writes:

 Anders> On 20-Apr-99, you wrote:

 >> and if you want to do xyin opcode you need the following two; or at
 >> least dummies liek these
 >> 
 >> void MakeXYin(XYINDAT *wdptr, float x, float y)    /* initial proportions
 Anders> */
 >> void ReadXYin(XYINDAT *wdptr)

 Anders> _*Question:*_
 Anders> What do you mean with /'initial proportions?'/
 Anders> Is it the aspect ratio of the window, or is it the absolute window size?

You are passed the range of values which are being requested from teh
xyin opcode.  So it is up to the implementor what (s)he does with
them.  You can create your own size or whatever.


 Anders> 1. What's happened with the function "oload()" from "oload.c"
 Anders>    used in "musmon.c" since V3.40?  It seems to do nothing for
 Anders>    about 5-10 seconds now. In V3.40 I didn't even notice the
 Anders>    time!
 Anders>    Ofcourse the long wait depends on my *slow* computer (68030/50mhz),
 Anders>    and that the Amiga is generally slow on memory access.
 Anders>    but the speed difference between the old and new version is about
 Anders>    2000%, so something must have happened.
 Anders>    Could it be that PMAX has increased *A LOT*?
 Anders>    What did it used to be in V3.40?

You could (and maybe we should) precompute the cpspch table.

 Anders> 2. When displaying a graph, should the program ALWAYS wait for the
 Anders>    user to click on the window before starting to draw?
 Anders>    If so, why? I think it's very annoying, but that could be IMHO.

This is rather open.  I instituted an environment parameter (CSNOSTOP
I think) to control this.  On X11 it does not stop at there are
multiple windows.  On DOS where there is only ever one it stops so you
might be able to see it.  Reeally it is up teh the platform
implementor

 Anders> 3. Has anyone compiled csound in GCC using "-ffast-math"?
 Anders>    I have used that for my test versions, and it has worked flawlessly
 Anders>    on every score i've synthesized.

Never heard of it.....

 Anders> 4. Is there something wrong with the makefile for "anal/adsyn"?
 Anders>    I can't get it to compile! I get the error message
 Anders>    "makef:13: *** multiple target patterns.  Stop.".
 Anders>    I bet this has something to do with my "make" program, but
 Anders>    if I'm lucky, it's an error in the makefile! =)

yes there is a silly error which I have corrected but not distributed.

 Anders> 5. Are there any documentation for using ".csd" files, or is it
 Anders>    just Me that's blind?

I think it was documented somewhere....probebly in my head.
The format is that it starts with
 or 
and ends with  or 
although these are optional!

The sections all start with a tag and end with a /tag

Sections, in any order are
   and 

The middle two of these shoudl be obvious.  In CsOptions any command
line thing can be given, spread over as many lines as wanted, except
that if an argumen is required (such as -o) then it must be on the
same line.

In  a binary file is presented, after a line
Size = %d
which says how long teh MIDIfile is.  I think there is a bug here as
it is mixing character and binary files, which is not necessarily
correct in ANSI C.

Does that help?

==John ffitch