>Hallo,
>Chuckk Hubbard hat gesagt: // Chuckk Hubbard wrote:

>> I've been trying for a few weeks to get my preset memory set up just right.
>> I got it right just tonight, and then it disappeared.

>Unfortunatly it is really hard to follow your patch for an outsider
>like me.

Yeah, I figured, I just wondered why Pd wasn't showing half of what was in the file.  I built it "organically", i.e. using numerous workarounds, and it's grown pretty huge, so its function isn't very clear.  But suddenly there were no connections showing on the main page, and several of the objects just weren't there at all- though I figured out why last night.

>Without knowing how it is supposed to work it's near impossible to
>tell, where it starts breaking. So here are some random remarks:

>You use things like [pd abc $1 $2]. This will not do what you may
>expect: Subpatches made with [pd] don't accept their own arguments,
>they use the arguments of the surrounding patch. You can delete the
>"$1 $2" from [pd abc $1 $2], it's totally ignored by Pd.

I figured this out too, I had just wanted to make sure the arguments from the patch were passed on to the subpatch.  I never bothered to delete them when I noticed it didn't matter.  I also haven't taken the time to delete all the extraneous GUI objects, since it wasn't quite ready for use.

>[until] is a very dangerous object, as it might not receive the
>stopping bang, if something unexpected happens. You use [until]
>several times to drive a "from-to-counter". It is much more secure to
>not send a bang into [until]'s first inlet and hope for the bang to
>the right inlet to arrive, instead you can send a float value to the
>until, which will make [until] bang as often as the float specifies,
>after that the until will stop automatically.

Thanks, I thought I had gotten rid of all of the ones without floats.  I had to reboot a few times early on and learned my lesson.
At any rate, I thought saving my work regularly was a failsafe, but, like I said, this disappearance of objects took place on all saved versions, when they were working fine before closing.
It appears to be just too much stuff!  The connections are all at the end, so when the file reached a certain size it just didn't get to them.  Same with the 8 "JIvoice" objects, they were added later, and it stopped reading before them.  So I simply deleted all of the data from my pd-presetmemory log (in a text editor), which I had forgotten to do: every time I was testing the preset function, Pd was adding every preset I used for testing to the log and saving it.  I added a "Clear" button and problem solved.
Is there a maximum file size that Pd will read?
Interestingly, I could still open the patch in a text editor and it was all there, but if I added anything to the truncated version of the patch and saved it in Pd, it actually deleted all the truncated stuff.  It was only a little over half a megabyte when it started truncating, which doesn't sound right...

Anyway, I fixed everything last night, and also proceeded to work out the score-based tempo control, so things are looking pretty good.  When I get video playback set up I'll be ready to score some student animations.
This was great because it forced me to start deciphering patches in text format.  Neat stuff.

Thanks, Frank.
-Chuckk

>Ciao
--
 Frank Barknecht                 _ ______footils.org_ __goto10.org__