Why am I having outdated command, yet it's not in user.gmic or update405.gmic?

Software G'MIC
1

I echoed these two commands, and the first one is not in disk.

C:\Windows\System32>gmic e $$rep_rde_compsel_comb
[gmic]./ Start G'MIC interpreter (v.4.0.5).
skip ${3=x} k,t:=sort([${1-2}]) check isint($k,2) N={perm($k-1,$t-1,0)} if !isint(_'$3',_'x',_'y') u $N return fi if $k==$t 1,1,1,$k,c return fi axis={_'$3'-_'x'} {v=vector(#3,1);v[$axis]=$N;v},$1 +rep_mt_f32v_map {$N>>[0,12]},0 eval. :"begin( const t=$t; const k=$k; const dk=k-1; const dt=t-1; const N=whd#-2; gosper_hack()=( g0=cn&-cn; g1=cn+g0; cn=int((xor(cn,g1)>>2)/g0)|g1; ); step()=( ik=dk;n=cn; p=t; while(n, r[--p]=n&1?(ik--):ik; n>>=1; ); gosper_hack(); I[#-2,x]=r; ); ); j=J(-1); x=j[0]<<24|j[1]; i=i0<<24|i1; r=vector(#t); x?( index=N-1-x; p=ix=0; ik=dk; while(++p<t, if(ik, block=perm(ik-1,dt-p,0); index<block?( ++ix; --ik; ):( index-=block; ); ); r[p]=ix; ); p=cn=0; while(++p<t,if(r[p]>r[p-1],cn|=1<<(dt-p););); gosper_hack(); I[#-2,x]=r; ):( cn=(1<<dk)-1; step(); ); while(++x<i,step();); _0; " rm.
[gmic]./ End G'MIC interpreter.

C:\Windows\System32>gmic e $$+rep_rde_compsel_comb
[gmic]./ Start G'MIC interpreter (v.4.0.5).
skip ${3=x}  k,t:=sort([${1-2}])  check isint($k,2)  N={perm($k-1,$t-1,0)}  if !isint(_'$3',_'x',_'y')  u $N  return  fi  if $k==$t  1,1,1,$k,c  return  fi  axis={_'$3'-_'x'}  {v=vector(#3,1);v[$axis]=$N;v},$1  +rep_mt_f32v_map {$N>>[0,12]},0  eval. :"begin(  const t=$t;  const k=$k;  const dt=t-1;  const K=k>2;  r=vector(#t);  K?(  const dk=k-1;  const lp=dk-1;  const N=$N;  pos=vector(#dk);  orig_r=expr('x',lp)+2;  orig_pos=expr('x',dk)+t-k+1;  );  );  O=J(-1);  P=O[0]<<24|O[1];  Q=i0<<24|i1;  K?(  P?(  idx=N-1-P;  p=q=id=0;  ik=dk;  while(++p<t,  if(ik,  block=perm(ik-1,dt-p,0);  idx<block?(  ++id;  --ik;  pos[q++]=p;  ):(  idx-=block;  );  );  r[p]=id;  );  ):(  pos=orig_pos;  repeat(dk,p,  r[pos[p]]=p+1;  );  ind=lp;  );  I[#-2,P]=r;  while(++P<Q,  m=lp;  while(m&&pos[m]-pos[m-1]==1,--m;);  M=m+1;  --pos[m];  m==lp?(  r[pos[m]]=M;  ):(  copy(r[pos[m]],[M],orig_pos[M]-pos[m],1,0);  );  m<lp?(  rem=lp-m;  copy(pos[M],orig_pos[M],rem);  copy(r[orig_pos[M]],orig_r[m],rem);  );  I[#-2,P]=r;  );  ):(  copy(r[dt-P],[1],P+1,1,0);  p=dt-P;  I[#-2,P]=r;  while(++P<Q,  r[--p]=1;  I[#-2,P]=r;  );  );  _0;  "  rm.
[gmic]./ End G'MIC interpreter.

The + one is newer. What happened?

EDIT: I just deleted it and I’m still getting the old command code even though it’s not there in update405.gmic. And none exist in user.gmic. The old one use gosper hack trick.

2

Are you somehow loading the old one? Common mistake of mine when I am doing bazillion things at once. Other times, I just reboot or restart the console.

3

I don’t know, it should not be there.

4

  • If your old command was defined before the release of latest G’MIC release, then the gmic binary has a copy of it directly in its binary executable. Each new released binary embeds the full content of the stdlib + community updates.

  • Then, if you had an old command defined as my_command : (some code), then decided to “redefine” it with +my_command: (some different code), then the initial version of my_command is not replaced by +my_command, because that’s actually two different command definitions that trigger commands in different contexts (whether the command is called with a + prefix or not). So if you call my_command in your script somewhere, that’s the first version that is still used (unless you call it explictely with +my_command of course).

That may explain your issue ?

5

Yes, that makes sense, so basically all I need to do is wait for 4.06. Alright.

6

You can also decide to redefine your old command (without the +), so that the old one become superseeded

my_command : +my_command $"*"
1 Like