matrix resizing problem


    Jan 30 2008 | 12:11 am
    hi all, I am building an application in Jitter, and have come across this possibly very simple, yet baffling problem. My matrices keep switching planecount on me?!?
    Now, this is absolutely user error at the root, but surely @adapt 0 should prevent this from ever happening? I need a jit.matrix 1 char 16 16, and it keeps screwing up for me. I have attached a small picture of the infected area for your perusal in lieue of a readable patch to give you.
    Cheers for any insights... this is driving me nuts, to be honest.

    • Jan 30 2008 | 12:18 am
      On 30 janv. 08, at 01:11, Andreas Wetterberg wrote:
      > hi all, > I am building an application in Jitter, and have come across this > possibly very simple, yet baffling problem. My matrices keep > switching planecount on me?!? > > Now, this is absolutely user error at the root, but surely @adapt 0 > should prevent this from ever happening? I need a jit.matrix 1 char > 16 16, and it keeps screwing up for me. I have attached a small > picture of the infected area for your perusal in lieue of a readable > patch to give you. > > Cheers for any insights... this is driving me nuts, to be honest.
      You're right, @adapt 0 force the matrix to not adapt to the characteristic of the input matrix, but in [jit.rubix] support only 4 planes (and @type char). So no matter what you do, [jit.rubix] will output a 4 planes matrix. You probably need a [jit.unpack] if you want to extract only one plane.
      HTH, ej
    • Jan 30 2008 | 12:36 am
      On Jan 29, 2008, at 4:11 PM, Andreas Wetterberg wrote:
      > hi all, > I am building an application in Jitter, and have come across this > possibly very simple, yet baffling problem. My matrices keep > switching planecount on me?!? > > Now, this is absolutely user error at the root, but surely @adapt 0 > should prevent this from ever happening? I need a jit.matrix 1 char > 16 16, and it keeps screwing up for me. I have attached a small > picture of the infected area for your perusal in lieue of a > readable patch to give you. > > Cheers for any insights... this is driving me nuts, to be > honest.
      Look at jit.rubix.help and/or the html reference page. This object only accepts 4 plane char matrices. So your attempt to change this is being overridden by the behavior of the object.
      You'll have to find either another solution or work with 4 plane matrices and ignore the other planes.
      -Joshua
    • Jan 30 2008 | 1:40 am
      Yeah, the problem did also appear to occur with regular jit.matrix'es, though... I figured that as long as I had adapt 0 on both the problem shouldn't be permeating the sytem like it did... hm. Thanks for your help, Joshua and EJ.
      Andreas.
      Joshua Kit Clayton skrev: > > On Jan 29, 2008, at 4:11 PM, Andreas Wetterberg wrote: > >> hi all, >> I am building an application in Jitter, and have come across this >> possibly very simple, yet baffling problem. My matrices keep >> switching planecount on me?!? >> >> Now, this is absolutely user error at the root, but surely @adapt 0 >> should prevent this from ever happening? I need a jit.matrix 1 char >> 16 16, and it keeps screwing up for me. I have attached a small >> picture of the infected area for your perusal in lieue of a readable >> patch to give you. >> >> Cheers for any insights... this is driving me nuts, to be >> honest. > > > Look at jit.rubix.help and/or the html reference page. This object > only accepts 4 plane char matrices. So your attempt to change this is > being overridden by the behavior of the object. > > You'll have to find either another solution or work with 4 plane > matrices and ignore the other planes. > > -Joshua > > . >
    • Jan 30 2008 | 11:41 am
      On 30 janv. 08, at 02:40, Andreas Wetterberg wrote:
      > Yeah, the problem did also appear to occur with regular > jit.matrix'es, though...
      Really? I can't reproduce that.
      ej