device that kills Live dead
Posted: Fri Oct 23, 2009 5:23 am
Hi, here is tonight's test case. This bug is more about the potential difficulty of recovery than it is about the actual configuration of the code.
This device will actually run and report index values. After that, however, it kills Live instantly when it is loaded into an audio track. I recognize that the coding scenario itself is pretty stupid. Nonetheless, I did something nearly this simplistic by mistake, and was unable to recover my work since the device would kill Live every time I loaded the device to try and remove the obviously suicidal communication. Of course, once I went into the pfft subpatch from within Max first, rather than editing from Live, I was able to disconnect the send~, and then undo the bozo bug. This would be the workaround. Is there a way to turn off the DAC/scheduler in Live? Configure the audio driver to None??
Here is the parent device. It calls out to a pfft subpatch, and has a receive~ object wired to a capture~ object. The capture~ object is part of the presentation.
<pre><code>
----------begin_max5_patcher----------
402.3ocyTtsTCBCDF9Z3oHStl1gvgZw6zWiNcbRg.MNPBCDTzNkmcIgCRUak
h1p2Dl7uK69mujY2oqA2vKI4PvsfU.Msc5ZZJIofV6dMXBtzOFmqRCxHOy27
HznIjfTJTx93TQQFopKPHmIX3DhJ3cYTbL3ddbPW3zLRNgIvBJm8PFwWzXAK
Gy4lF.jER9w0RsY4bSv51+iUjPYwDgxKnA8Jm9ppWHYA5ykWH5RtSMEK72RY
QmSWG51gMlFnNd07XlKTJsWWWtX7yXYsyHzmHUfPNej7b5bAcBtX6IQgmBHK
79DVFb7c5LRSkEujRZpALmFwvwPv5eO9jFFJp.su31TDMSJ.bL8V.b9K4EBc
i5AjZE4XcJfYeUAVbQTceplBbrNG3Xcb33rTgCjhJt1mhMVeGaLtXThx9G.I
6QvHzDYjpVvXJ6iS8UtQpeH3x4EY9ckt8gK3cGEPxETV+fwU8Wf.TeNaoAAD
1vYvIzfTNkIZsvQtDGqiPGzsu1QG55KsibFAibupNBMxaso5n5M60eieTTOW
-----------end_max5_patcher-----------
</code></pre>
Here is the pfft~ subdevice. It contains nothing more than a send~ object wired to the fft index.
<pre><code>
----------begin_max5_patcher----------
349.3ocuSsrbBCBEcM4qfg0VGIopwtq82vwoCFHJcRtjIfSs0Q+1KPRT5Cs9
nS2Pf6yy4buYSDhLWsVnI3GvSwHzlHDxaxY.09FQJYqyJXZeXDP7pZ9KjdMt
Lh0Fu47biZkYGl14IWAFfUJ7derVxJvOoJ3ctgUkRnPX7EMNHGs7ceNzA8Gb
HVas6BtyZEyjsTBKdtVjYZX.M0kDldu+yvlGo8Gfm0lij6gik.2ESbl1FE4N
5cyjWBVtiAk850n.zKQARNkBLxQ5wotyQSNk.reR0TYyaUhlZPzxE.qfz6Gt
gm82oZZAv2gyUpyTvppE1TLLiTAgTdRrmlwC8C8zuw4qWnOwpVB02N53wGoq
gnMrwASfjvUPehjBI70eG8M2Y+yJrVspNqah0NOwGZOWns6j669ztk9vXVJ4
bADRzRIuRIASKDNxz9RPD8LPD8eEQw+BhRtAMx9XazG.oa+O+.
-----------end_max5_patcher-----------
</code></pre>
I've also got an alp for this if you need it.
Live 8.1b6, Max 40387, OSX 10.5.8
This device will actually run and report index values. After that, however, it kills Live instantly when it is loaded into an audio track. I recognize that the coding scenario itself is pretty stupid. Nonetheless, I did something nearly this simplistic by mistake, and was unable to recover my work since the device would kill Live every time I loaded the device to try and remove the obviously suicidal communication. Of course, once I went into the pfft subpatch from within Max first, rather than editing from Live, I was able to disconnect the send~, and then undo the bozo bug. This would be the workaround. Is there a way to turn off the DAC/scheduler in Live? Configure the audio driver to None??
Here is the parent device. It calls out to a pfft subpatch, and has a receive~ object wired to a capture~ object. The capture~ object is part of the presentation.
<pre><code>
----------begin_max5_patcher----------
402.3ocyTtsTCBCDF9Z3oHStl1gvgZw6zWiNcbRg.MNPBCDTzNkmcIgCRUak
h1p2Dl7uK69mujY2oqA2vKI4PvsfU.Msc5ZZJIofV6dMXBtzOFmqRCxHOy27
HznIjfTJTx93TQQFopKPHmIX3DhJ3cYTbL3ddbPW3zLRNgIvBJm8PFwWzXAK
Gy4lF.jER9w0RsY4bSv51+iUjPYwDgxKnA8Jm9ppWHYA5ykWH5RtSMEK72RY
QmSWG51gMlFnNd07XlKTJsWWWtX7yXYsyHzmHUfPNej7b5bAcBtX6IQgmBHK
79DVFb7c5LRSkEujRZpALmFwvwPv5eO9jFFJp.su31TDMSJ.bL8V.b9K4EBc
i5AjZE4XcJfYeUAVbQTceplBbrNG3Xcb33rTgCjhJt1mhMVeGaLtXThx9G.I
6QvHzDYjpVvXJ6iS8UtQpeH3x4EY9ckt8gK3cGEPxETV+fwU8Wf.TeNaoAAD
1vYvIzfTNkIZsvQtDGqiPGzsu1QG55KsibFAibupNBMxaso5n5M60eieTTOW
-----------end_max5_patcher-----------
</code></pre>
Here is the pfft~ subdevice. It contains nothing more than a send~ object wired to the fft index.
<pre><code>
----------begin_max5_patcher----------
349.3ocuSsrbBCBEcM4qfg0VGIopwtq82vwoCFHJcRtjIfSs0Q+1KPRT5Cs9
nS2Pf6yy4buYSDhLWsVnI3GvSwHzlHDxaxY.09FQJYqyJXZeXDP7pZ9KjdMt
Lh0Fu47biZkYGl14IWAFfUJ7derVxJvOoJ3ctgUkRnPX7EMNHGs7ceNzA8Gb
HVas6BtyZEyjsTBKdtVjYZX.M0kDldu+yvlGo8Gfm0lij6gik.2ESbl1FE4N
5cyjWBVtiAk850n.zKQARNkBLxQ5wotyQSNk.reR0TYyaUhlZPzxE.qfz6Gt
gm82oZZAv2gyUpyTvppE1TLLiTAgTdRrmlwC8C8zuw4qWnOwpVB02N53wGoq
gnMrwASfjvUPehjBI70eG8M2Y+yJrVspNqah0NOwGZOWns6j669ztk9vXVJ4
bADRzRIuRIASKDNxz9RPD8LPD8eEQw+BhRtAMx9XazG.oa+O+.
-----------end_max5_patcher-----------
</code></pre>
I've also got an alp for this if you need it.
Live 8.1b6, Max 40387, OSX 10.5.8