@LiamG said:

Hi @jancsika

I just downloaded 2.1.1 for Windows and I can't seem to get [initbang] working. Should I file a bug report?

Does it instantiate?

Yes, it instantiates, but it doesn't seem to send a bang as expected.

@LiamG Fixed for the next release. You can also grab it off the CI builder when it finishes:

https://git.purrdata.net/jwilkes/purr-data/pipelines/872

Just click the relevant OS and then download the "artifact" on the page that pops up.

Don't have Windows in the CI yet, but I'm working on it...

@jancsika I have just upgraded to 2.2.3 win 10 but had a major problem where pd.exe and unins000.dat and unins000.exe were not replaced (install error: could not replace existing file,,,,) and of course the program would not run (pd,exe not found) and no admin powershell or command prompt would allow access to delete.
I reverted to shift restart (safe mode) and command prompt to finally delete the files and complete the install.

I updated because I thought it may have fixed my current problem of $0 in struct names and scalar names not allowing messages to change colours etc. I think you have already addressed this before, but I couldn't find the post. I will move to Purr Data when I can produce a usable $0 abstraction, but so far it has eluded me. Am I doing something wrong?

purr-test.pd
purr-test-2.pd

@Balwyn Should be fixed in the next release. When the CI finishes you can try a binary from it:

https://git.purrdata.net/jwilkes/purr-data/pipelines/896

@jancsika Thanks for the new 2.2.3 version! I don't see anymore the Win64 version, tho.
best!

@alfonso.santimone You should be able to run the 32-bit version on 64-bit Windows.

The 64-bit version with previous releases was a 64-bit GUI but still a 32-bit Pd engine.

Some more bug fixes with 2.2.4 (see above)

Hey man, Just wanted to say thanks and great work on the latest Purr-data. I am really stoked that development is being actively maintained on a new version of PD. I have noticed a bug/glitch that has been persistent throughout every version of purr-data so far, but cannot seem to find any mention of it anywhere, so I was wondering if this is a known bug or if I have just configured something in my system improperly. The behavior I am talking about is in many of the UI objects being unresponsive. For instance, [nbx] will send floats when clicked and dragged, but the updated value will not be reflected in the UI until the entire window is closed and reopened or resized. Pretty much every UI object does this, nbx, radios, sliders, bang... they work but the UI doesn't budge. I have worked around it by clever use of feeding the output into a send object which labels the object with its current value, or move canvases and structs around to revlect the position of a slider etc, but I admit that gets a bit tiresome. This behavior seems to be especially present in larger projects with multiple levels of GOP subpatches. Any info or insight on what is causing this would be greatly appreciated. Thanks again!

@th8a What's your OS, and does the bug still happen in 2.2.4?

https://git.purrdata.net/jwilkes/purr-data-binaries

If so, can you send me a demo patch that uses the absolute minimum number of objects required to trigger the bug?

By absolute minimum, I mean less than ten objects-- then I'll probably be able to find the cause of the bug in about 5 minutes. I can also try to track it down in a more complex patch but it will take me orders of magnitude longer.

If you come up with a minimal patch to demo the bug post it here or on the issue tracker:

https://git.purrdata.net/jwilkes/purr-data/issues

@jancsika Im using Ubuntu zesty 17.04. I built the latest git version to see if the bug had been fixxed but it was still doing it. It tends to be working OK when I start a new patch that has under 10 - 20 objects or so. The problem is consistent however in any more complex patch that I open from previous versions of pd-l2ork. I'll be staying at my folks house for the next couple of days with not too much to do, so I'll take a crack at recreating the error in a small patch. My instinct is that it deals with layers and overlaps of GUI with GOP properties, and also likely influenced by larger numbers of gui elements. I'll see if I can recreate it in a consistent manner and get back to you here. stay tuned and many thanks.

Still looking for a consistent reproduction of the bug, but I can confirm that it is present in OSX as well as Ubuntu 16, 17 and fedora. I'll zero in on it tonight and continue to narrow it down as much as possible.

and here you go

https://git.purrdata.net/jwilkes/purr-data/issues/365

Found the smallest patch that I could that causes the bug. Hope that is helpful.

Cheers!

@th8a Thanks!

I replied on the issue.

I am having difficulty with loading the libraries which come with the OSX L2ork dist.
Am I missing something obvious? I've added search paths to prefs, nothing loads on restart.
Tried path to extras folder within the application, as well as copied them out to a simpler path, to no success.

Any help would be appreciated.
(Thank you so very much for the OSX dist of L2ork by the way, and your work with Purr Data.)

@mumbleman Should be fixed in 2.2.4:

https://github.com/jonwwilkes/purr-data/releases

@jancsika ,thank you for fast response. It has indeed been fixed.

So I have been keeping tabs on your guy's work in developing, and I saw that a new commit had been made to fix the bug that I mentioned here previously... So I built the latest from the repo, and low and behold - the problem appears to be fixed and no longer occurring! I'll still run it through the full gamut of tests to be thorough, but from what I see here, it's completely resolved. This has me super stoked. Thanks again man. Good work, as always!

Ok, so I have ran some further testing. The situation is most definitely improved, however not completely resolved. If you would like, I could put together some more detailed notes etc. Up to you - say the word and I'll get to it.

@th8a Yes, go ahead.