klearnnotes2-devel

[kln2-devel] Installation problems solved

[Mo <Mo...@no...>]

Hello everybody,

Thanks for the foward Marek. It really did help :)


> 
> Hi e-body,
> Carlo had the same problems with compiling and here is his solution.
> 
> Mo, I hope this works for you too.
> ~Marek

Yes, the following changes did help and the program now compiles :)

> 
> 
> here the diff from file klearnnotes2/src/voice_recognition.cpp
> 2564c2564
> <   int           prefetch_N    = 5;
> ---
> >   const int           prefetch_N    = 5;
> 2612c2612
> <       memcpy((prefetch+prefetch_pos*(FRAG_SIZE)), buffer_raw,
> FRAG_SIZE);
> ---
> >       memcpy(((unsigned char*)prefetch+prefetch_pos*(FRAG_SIZE)),
> buffer_raw, FRAG_SIZE);
> 2637c2637
> <       memcpy(data->buffer, prefetch+(i%prefetch_N)*FRAG_SIZE,
> FRAG_SIZE);
> ---
> >       memcpy(data->buffer, (unsigned
> char*)prefetch+(i%prefetch_N)*FRAG_SIZE, FRAG_SIZE);
> 

I have updated the modified file to the CVS tree.

> ** still a problem: I cannot manage to have MIDI playback running


I am having this problem too :(
   "DEBUG: midi_turnOn(): couldn't open midi device!"

Although I have ALSA working on my computer. What else am I missing?

Thanks!
-Mo

[kln2-devel] midi output

[Marek W. <Mar...@fu...>]

Hi e-body,

On Mon, 23 Feb 2004, Mo wrote:
>   "DEBUG: midi_turnOn(): couldn't open midi device!"
>Although I have ALSA working on my computer. What else am I missing?

Mo, it's not you, it's kln2 that is missing something. :( It uses
OSS/Free, because this is what I have, and this is what I could test.

Remarks below do not end with anything constructive, so if you are not
interested just ignore them. This is just to let you imagine where I am
with thinking about the problem.

==============================================

Yes, there are different ways of generating (midi) sounds and users should
have a choice.

A good source for implementing different solutions is for example
'solfege'. They have there:
* OSS (/dev/sequencer , /dev/sequencer2 == /dev/music)
        sorted out better than what I have written based on very raw code
* external programs:
        timidity -idqq %s
        drvmidi -i d %s
        playmidi -f %s
(although the last two would work only with OSS, I believe). Plus, they
use some properties of python which look almost like overriding virtual
C++ members, so maybe even part of the structure of the (quite clear and
extensible :) code could be preserved.

Timidity is very popular and works for virtually everyone, so this is
probably the solution we need fast (Javier voted for this quite some time
ago). And it should be very easy to implement (based on e.g. solfege). So,
this may happen soon.

ALSA? How? (maybe see how old 'rosegarden' uses it)

Another solution would be to use some library which would take care of
different hardware and sound libs.

        libkmid
is the most obvious choice (but I wasn't able to implement it; on the
other hand I spent on this very little time; Mo: does kmid program play
midi files on your computer?). And making use of aRts (instead of
suspending it, as it is done now) would be nice. BTW: is there a gnome
equivalent?

        allegro
?

        jackd
? (see how new 'rosegarden' uses it)

BTW, right now the only SF kln2 programming "help wanted" post is about
this problem. So maybe someone experienced with the subject will still
volunteer.

================================================

Conclusion: one day (let's hope soon) someone will have to look at this.
For now, sorry, midi will not work for you. From kln2 point of view it is
not that important. Please, try to ignore the problem. :(

~Marek

-- 

     \/   /|\  Marek Wieckowski
  #####  | | |
 = . . =  \|/  Institute of Theoretical Physics
    U      |   Warsaw University
  / ~ \___ |
 <|   |    |
  > . <    |   http://www.fuw.edu.pl/~wiecko
 <<___>>   |   http://klearnnotes2.sourceforge.net