I'm trying to send data to and from an existing Windows app (VB/NET) that I cannot at this time change. The app apparently is written to expect the ";" at the end of data characters being transmitted. This was written to work in PD, which it does, using netsend/netreceive. But i want to re-work this to use Max. I have not been able to simply replace netsend/netreceive with anything i can find (and I have looked) in Max. So far as i can see, using for example net.tcp.send (and tcpClient too) in Max, i never get the ";". It is stripped off if I try to add it in the string to be sent (of course - that would be too easy!). I verified this using a packet sniffer. I think this is why the app I'm trying to connect to keeps crashing. ANy suggestions would be GREAT. I'm in over my head... again!

OK that makes sense.

I just realized that the normal receive~ does accept a "set xxx" to allow dynamic setting. I did not realize that, as i didn't see it in the documentation. Am i missing something? Why do we even need receive13~ ?

Yes, I can also use send~ and receive~ fine but not the "13" version. I am on Win7 64. i just now recompiled on my machine using mingw/gcc and cannot get the object to load, even with no errors or even warnings from gcc

Nothing changed that i know of. I am 98% sure it was working before. i wrote 43.3 but meant 43.4, so is is current. i could try an older version. Regular send/receive works fine. i am trying to compile the receive13~ source on my pc, but I am not very good at it. using Mingw...

I tried various combinations, and yes i used send13~ (the help file for example). I tried deleting and recreating the receiv13~ obj too. Nothin works!

Receive13~ object creates fine. i do get error messages about "receive13~ no matching send" when i create it. But I cannot get any data from the send to the receive. Like i said, the help file does not work, i.e. no data is received when data is sent. i would like to programmatically create patches with individual receive names.

DSP is on and I have cycled it on and off.

I thought i had this wroking but now, receive13~ does not work. I mean, just opening the help file for receve13~, no other patches open, it doesn't work. I figure if the help file doesn't work, it may be broken for Win7, 43.3 extended. The normal send/receive~ works fine. Anyone have this problem?

Thanks, that sounds like a worthy addition to my debugger.

OK, I removed the list2symbo and makefilename:

[openpanel]
|
[list append sl sr]
|
[set read $1 $2 $3 $4(
|
[msg box(
|
[soundfiler]

This works. Is there some way to "see" whether a message is one atom or 5 atoms? I used my "debugger", a route object with bang, float, symbol, anything, and udefined, outputs and in both cases it came out undefined. i could not really distinguish the two cases. Thanks.

It definitely seems to be smething that is not apparent from just looking at the message that is sent to soundfiler. I even tried sending the message to print, and the working version and non working version look identical.

[openpanel]
|
[list append sl sr]
|
[prepend read -resize]
|
[list2symbol]
|
[set $1(
|
[msg box(
|
[soundfiler]

Not sure if I notated that correctly. As I said, editing the message box by adding and removing a space makes it work.

This is really wierd: i have a patch where I'm trying to load up a file into soundfiler using openpanel. I figured out how to make the symbol, including the filename, to make it work.

"read -resize D:/dirxyz/filenamXYZ.wav tableL tableR" (Quotes added)

I could never get this to work, getting "no method for read -resize D:/dirxyz/filenamXYZ.wav tableL tableR" error message in pd. I have noticed though, that if I edit the message box containing the read statement, just add a space and then remove it, all the sudden it works! It is the exact same message i'm sending as far as I can see. Can anyone explain it, and HOW do I make it so i dont have to do that every time?

Hoping someone will reply.

I'm trying to get some realtime partitioned convolution going in PD extended. I have longish impulse responses, so FIR~ is not an option. I at one point HAD [bsaylor/partconv~] working (26 individual realtime convolutions of 32768 length) but today it won't run, just crashes pd every time. I have tried changing partition size, buffer size (PD audio settings), and a host of other things, but it always just crashes after around 5 seconds. no way to find out what happened. I had it running with pd block size set to 2048, partition size in the partconv~ object set to 512, block size in the subpatch where partconv is set to using switch~ to 128 , and latency in audio settings to 50 ms. These settings had worked before but not now. I'm using only inlet/outlet between subpatches with different block size.

Should this work, or was I just hallucinating when it worked for 10 minutes before? Is there any way to debug this? Would recompiling have any chance to help? The external is a few years old and I am on a 64 bit Windows 7 platform.

Only thing I found that works is using 0.43.4. That works for me for >2 channels, no problem. RME Fireface UC, WIndows 7.

I started out just trying to run partconv~. Well I was able to get sound but it never sounded right. Then I started getting crashes when i would try to change to block size under ASIO settings in PD. PD would basically just stop responding (not exactly freezing but close enough). Now, PD won't run at all. I mean, it runs, i can see it in the task manager (PD 43.4, Windows 7, Lenovo 550 laptop, RME Fireface UC sound card), but no PD GUI visible. Tried rebooting still same problem. How do I get my GUI back?

I had this problem, and turning on my sound card made the mouse start responding. I not even kidding!

I am trying to use partconv~. I just tried the help patch, loaded a couple of hundred ms impulse and started up some audio signal on soundcard inputs (RME Fireface UCX, running Windows 7). Sound is coming out but sounds horrible, like it is slowed WAY down and/or scratchy/distorted. I checked the usual stuff, sample rate etc. and tried changing the block size. Still no go. I'm wondering if there is anything else I could try? Wondering is I need to load fftw3, since I think partconv uses it. Then again, perhaps fftw3 is already included in the partconv dll, since it DOES run (sorta). ANy help appreciated...

Tried to run a patch originally developed and working in 42.5 extended. I use openpanel object to load files. I used to sometimes get a dissapearing GUI (gripd) in 42.5 when i would cancle out of an open file diaglog. This time, in 43.4, the gui again dissapeared, and all of my abstractions went blank, and had a really hard time to quit out of pd. Kept getting errors about "can't talk to pd.exe" or somesuch. I may be going back to 42.5.

.. a few minutes later.
Froze up on me again, unresponsive and can't exit. This version does not seem stable at all. Also, operations that were fairly quick on 42.5 takes very long here, such as loading files, and creating tables. I don't get it.

Does anyone know what this does? I can't find much documentation other than that it "automatically connects objects as they are created" or somesuch. i created some new objects with it turned on and nothing in particular happened.

I don't see the save button for sound settings (under media) in 0.43.4. That was a handy feature, anyone know why it is gone?