[CM] Pointers into snd-help source?

[Bill Schottstaedt]

>> GUILE_LIBS=-lfth
>> I guess there's no FTH_LIBS...
> I think Guile was first.  You can read GUILE_* as XEN_* names.

I'll change the names to XEN_CFLAGS and XEN_LIBS.


>> Have there been any discussions/thoughts
>> about how to approach providing the "H_" strings defined in the code now
>> that Snd is multi-lingual?

yes, and some things are handled correctly -- constant names,
example code snippets, etc:


Scheme:
:(snd-help "make-fft-window")
"(make-fft-window type size :optional (beta 0.0) (alpha 0.0)): -> fft data window
(a vct). type is one of the sndlib fft window identifiers such as kaiser-window,
beta is the window family parameter, if any:
  (make-fft-window hamming-window 256)"

Ruby:
>snd_help("make_fft_window")
make_fft_window(type, size, :optional, (beta 0.0), (alpha 0.0)): -> fft data window (a vct).
type is one of the sndlib fft window identifiers such as kaiser_window, beta is the window
family parameter, if any:
  make_fft_window(Hamming_window, 256)

Forth:
>"make-fft-window" snd-help
(make-fft-window type size :optional (beta 0.0) (alpha 0.0)): -> fft data window (a vct). type
is one of the sndlib fft window identifiers such as kaiser-window, beta is the window family param\
eter,
if any:
  hamming-window 256 make-fft-window


But, as you can see, there are still infelicities.  One real improvement would be to
get gtk/motif/opengl documentation into the xg/xm/gl.c help strings.  And, as always,
we need more examples.


>> 4) Would there be a reason why compiling xg.c takes an order of
>> magnitude longer with FTH than with Guile?

This is probably a matter of system memory -- xg.c has thousands of functions and whatnot,
so if you happen to hit your memory limit while running the compiler, you'll be brought
to a standstill as it swaps.  I don't notice any real difference on a machine with 4 GBytes RAM:

Guile:   0:43.05
Gauche:  1:09.41
Ruby:    0:52.58
Forth:   0:42.39