Play - Implement J(x) P(y) - Protected Jump

[SciLor]

The protected jump J(x) P(y) doesn't seem to be implemented, so you can't test script using it.
I am sorry that I cannot test the code on the tiptoi itself. I will test it when I have it here, to know whats the "right" behaviour.

Code

  haus_bild:
  - $playing==1?
  - $playing:=1 J(_haus_bild) P(nop)
  _haus_bild:
  - J(_resetPlaying) P(b) P(wait_1000ms) P(wie_in) P(wait_0200ms) P(bild)

  _resetPlaying:
  - $playing:=0 J(_random) P(nop)
  
  _random:
  - T($r,65535) $random+=$r $random*=25173 $random+=13849 $nextId2:=$random $nextId2%=2 $nextId3:=$random $nextId3%=3

Reaction

Executing:  $2:=1 J(3948) P(60)
Executing:  J(3998) P(4) P(101) P(105) P(99) P(6)
Executing:  $2:=0 J(3997) P(60)
Executing:  T($3,65535) $4+=$3 $4*=25173 $4+=13849 $0:=$4 $0%=2 $1:=$4 $1%=3
State now: $0=1 $1=2 $4=31223

[nomeata]

I actually don't fully understand the “protected jump”. Maybe someone should write a proper documentation in the book first :-)

[SciLor]

If I have some time to play with it, I will post some infos for you here

[SciLor]

I have played around a bit and found following (using the newer tiptoi)

-Only one J(ump) per line is allowed. If multiple Js are found, only the last one is taken.
-You need to place at least one valid P(lay) after a jump or it waits about 2 seconds
-A wav with the length of 1ms is too short, I needed at least 5ms (for my nop)

So the J(ump)-Command just seem to set a register, where to jump after the line.

  a:
  - J(_haus_apfel) P(nop)
  _haus_apfel:
  - P(a) P(wait_1000ms) P(wie_in) P(wait_0200ms) P(apfel) P(wait_1000ms) J(_haus_apfel2) P(nop)
  _haus_apfel2:
  - J(_anything) P(biss_kauen) P(wait_0500ms) P(koestlich) P(wait_1000ms) P(huh_bin_ich_satt)

What the tiptoi plays:
a, wie in, apfel, biss_kauen, koestlich, huh_bin_ich_satt

What your simulation plays:
a, wie in, apfel, (aborted, because you jump immediatelly)

[exolab]

-You need to place at least one valid P(lay) after a jump or it waits about 2 seconds
-A wav with the length of 1ms is too short, I needed at least 5ms (for my nop)

That fixed all my problems with annoying delays.

If you don't happen to have an ogg lying around, you can generate it like this:

ffmpeg -f lavfi -i anullsrc -t 5ms -ar 22050 -c:a libvorbis audio/5ms.ogg

[nomeata]

Now also tracked at #231