Am 19. Juli 2024 19:19:56 MESZ schrieb Alexandre Torres Porres porres@gmail.com:

What I'd really hope to see in this update is this

https://github.com/pure-data/pure-data/pull/2380

My understanding is, that the sub-releases (eg 0.55-1, 0.55-2), are *bugfix* releases. The mentioned PR is not a bugfix, but clearly a new feature.

I also think it needs more testing from various people on various platforms with different keyboard layouts.

Also, I would like it to be accompanied with https://github.com/pure-data/pure-data/pull/2383, which certainly needs more testing (AFAIK nobody but me tested that yet).

Since the two PRs conflicts, my plan is to have both tested independently, and then merge them manually into a single PR.

Obviously (having spent time implementing both) I would like to see them in Pdf, I just don't think that a bugfix release is the correct place to introduce them.

mfg.sfg.jfd IOhannes

Hi,

I can post an issue on the github repo if this is something that needs to wait until later...

All bug reports should go to GitHub, regardless of their urgency.

Christof

On 19.07.2024 20:17, William Brent wrote:

Hi Miller and all,

If it's not too late to investigate for this update, I noticed that Pd 0.55-0 suddenly stopped supporting JACK with 8+ output channels. When I set output channels to 8 or more, Pd freezes and I have to issue a "kill -9" for the process. This configuration still works for me using Pd 0.54-1, where I regularly use 20 output channels. Here's the software/hardware situation where I'm experiencing the problem in case anyone else can reproduce it:

• macOS Sonoma 14.5
• Pd 0.55-0 (binary downloaded from https://msp.ucsd.edu/software.html)
• jackdmp version 1.9.22
• MacBook Pro M1 64GB RAM

I can post an issue on the github repo if this is something that needs to wait until later... William

On Fri, Jul 19, 2024 at 12:40 PM Miller Puckette mpuckette@cloud.ucsd.edu wrote:

To Pd Dev -

What do you think, is it time to put 0.55-1 out?

If I don't hear any objections I'll stick out a "test" one
Saturday, and
maybe the real one in a week or so.

cheers

M




_______________________________________________
Pd-dev mailing list
Pd-dev@lists.iem.at
https://lists.puredata.info/listinfo/pd-dev

-- William Brent

“Great minds flock together” Conflations: conversational idiom for the 21st century

www.conflations.com http://www.conflations.com

---

Pd-dev mailing list Pd-dev@lists.iem.at https://lists.puredata.info/listinfo/pd-dev

Oh, I forgot about these questions of mine. Please, someone, help me, so I can also update this ;)

Em ter., 23 de jul. de 2024 às 19:11, Alexandre Torres Porres < porres@gmail.com> escreveu:

Hi, I'm kinda done with that 'audio examples' revision I mentioned and working on other doc related stuff to make it into the next update.

Still regarding the 'audio examples', I saw a bit of text that could maybe be improved and updated. I actually already updated it, but I wanted to discuss it with you here on the dev list.

Originally - in example A04.line.pd - it did say "*On most machines, you will hear an interruption in the sound one second after you click on the first or third box. This is because the graphical updates are likely to eat more CPU time than your audio buffer has pre-buffered for. You can avoid this if you keep your graphs in sub-windows and open them only when you need them. In some future version of Pd this behavior will be improved. Until then, you'll have to avoid having arrays getting re-drawn during music performances.*"

I think it's safe to say that this doesn't happen anymore "*On Most Machines*", especially on this kind of example. I can see that dragging an array with a huge sample does choke the audio, or plotting it as well, maybe, I don't know, but not something simple like this.

The text also say that "*future version of Pd this behavior will be improved.*" so I ask, hasn't this been improved already? :)

What do you think I should say here, if anything?

Thanks

Em sex., 19 de jul. de 2024 às 16:59, Christof Ressi < info@christofressi.com> escreveu:

...

Hi,

I can post an issue on the github repo if this is something that needs to wait until later...

All bug reports should go to GitHub, regardless of their urgency.

Christof On 19.07.2024 20:17, William Brent wrote:

Hi Miller and all,

If it's not too late to investigate for this update, I noticed that Pd 0.55-0 suddenly stopped supporting JACK with 8+ output channels. When I set output channels to 8 or more, Pd freezes and I have to issue a "kill -9" for the process. This configuration still works for me using Pd 0.54-1, where I regularly use 20 output channels. Here's the software/hardware situation where I'm experiencing the problem in case anyone else can reproduce it:

• macOS Sonoma 14.5
• Pd 0.55-0 (binary downloaded from https://msp.ucsd.edu/software.html

)

• jackdmp version 1.9.22
• MacBook Pro M1 64GB RAM

I can post an issue on the github repo if this is something that needs to wait until later... William

On Fri, Jul 19, 2024 at 12:40 PM Miller Puckette < mpuckette@cloud.ucsd.edu> wrote:

...

To Pd Dev -

What do you think, is it time to put 0.55-1 out?

If I don't hear any objections I'll stick out a "test" one Saturday, and maybe the real one in a week or so.

cheers

M

---

Pd-dev mailing list Pd-dev@lists.iem.at https://lists.puredata.info/listinfo/pd-dev

-- William Brent

“Great minds flock together” Conflations: conversational idiom for the 21st century

www.conflations.com

---

Pd-dev mailing listPd-dev@lists.iem.athttps://lists.puredata.info/listinfo/pd-dev

---

Pd-dev mailing list Pd-dev@lists.iem.at https://lists.puredata.info/listinfo/pd-dev

Let me forward here to this list my communication with Shahrokh. I thought I'd write him directly about these issues and he quickly replied to say he's gonna work on it, so I wanted to let people know things are moving and that we're on it.

---------- Forwarded message --------- De: Alexandre Torres Porres porres@gmail.com Date: ter., 20 de ago. de 2024 às 21:11 Subject: Re: [expr] updates for Pd 0.55-1 To: Shahrokh Yadegari sdy@ucsd.edu

Em ter., 20 de ago. de 2024 às 19:54, Shahrokh Yadegari sdy@ucsd.edu escreveu:

Dear Alexandre,

Thank you for your email and reporting the bugs that you have found! I will address them as soon as I can (hopefully in the next few days).

Yay!

Some of them are kind of features like the spaces before the symbol itself in output of symboln(), but if that is not clear to you, perhaps I should change the API.

hmm, ok, I think I get it now, but, well, yeah, I think it makes better sense to not put the extra missing characters as spaces in the beginning.

I worked on the array[$v#] implementation, but many interpolation issues came up (some related to backward compatibility) and I decided to leave that for the next release. I wanted to get the symbol processing changes out asap.

I see. Even if we leave it for later, I think we can discuss what we expect for the signal input and interpolation. I was discussing that on the PR I opened.

Please stay tuned for a more complete email and please keep the bugs coming!

Ok, let's kinda focus on bugs and let me filter them out a bit and describe them more. I will leave the signal support for index aside for now.

This is one we're already discussing https://github.com/pure-data/pure-data/issues/2399

This is a simple and classic bug for heap corruption / memory leaks that you can have a look at https://github.com/pure-data/pure-data/issues/531

And here's another clear bug https://github.com/pure-data/pure-data/issues/402 but it seems this is more about how Pd deals with string handling, which escapes some of the special characters, so maybe a job for IOhannes in the develop branch.

For last, this feels to me like a bug as well https://github.com/pure-data/pure-data/issues/2398

The rest is more like feature requests, and I'll focus on this one that aims to allow symbol inputs to be treated as variables, seems like a good feature to be part of the whole symbol support update that we're focusing on https://github.com/pure-data/pure-data/issues/2400

And another related feature request would be expanding [value] to also take symbols so we can make use of the inside [expr], but more about [value] than [expr], I guess that a PR and a commit for this one could be done once we're through with symbol support and other bugs.

Cheers

Cheers, Shahrokh

On Tue, Aug 20, 2024 at 2:25 PM Alexandre Torres Porres porres@gmail.com wrote:

...

Hi again, I've been creating new issues, new PRs and writing comments in 'expr' related issues and PRs. To organize things, I created a new 'expr' label for issues and PRs. Here are them all https://github.com/pure-data/pure-data/labels/expr https://urldefense.com/v3/__https://github.com/pure-data/pure-data/labels/expr__;!!Mih3wA!Em8VTuW7HdeNM2CSOJFad7RmuHN-g07GrXnZ7z5i0FhjjUS3-Jr_oOF7qYJF1FQw78Y2zdxE-JXfhC4$

There are more expr related issues and PRs but these are the ones I thought it was worth tagging for now, being more important and relevant while we're making changes to 'expr'. Not that I also think ALL of them should be taken care of right now...

cheers

Em ter., 20 de ago. de 2024 às 14:26, Alexandre Torres Porres < porres@gmail.com> escreveu:

...

Hi Sharok Yadegari, how are you?

I see Miller just merged an update of the [expr] family of objects and it's really exciting to have support for symbols in [expr] now.

I was writing about this in the pd-dev list and I'm not if you are there or following it, so I thought I'd just write to you directly.

Since the expr 0.58 version is not official out yet, we can maybe work out some bugs to make it into this update.

I just came across this bug for instance https://github.com/pure-data/pure-data/issues/2398 https://urldefense.com/v3/__https://github.com/pure-data/pure-data/issues/2398__;!!Mih3wA!Em8VTuW7HdeNM2CSOJFad7RmuHN-g07GrXnZ7z5i0FhjjUS3-Jr_oOF7qYJF1FQw78Y2zdxEzRXfvCE$ I also see that the update notes within the expr code mentions *" fixed the expr~ array[0] "bug"* and I wonder what bug that was... when trying to test it is when I came across this bug I reported...

Also, this may seem like a 'feature', but it seems to me is just how things should work intuitively, which is being able to set an array index with audio signals, see https://github.com/pure-data/pure-data/issues/719 https://urldefense.com/v3/__https://github.com/pure-data/pure-data/issues/719__;!!Mih3wA!Em8VTuW7HdeNM2CSOJFad7RmuHN-g07GrXnZ7z5i0FhjjUS3-Jr_oOF7qYJF1FQw78Y2zdxEmVkdiPQ$ and there's already a PR for that https://github.com/pure-data/pure-data/pull/853 https://urldefense.com/v3/__https://github.com/pure-data/pure-data/pull/853__;!!Mih3wA!Em8VTuW7HdeNM2CSOJFad7RmuHN-g07GrXnZ7z5i0FhjjUS3-Jr_oOF7qYJF1FQw78Y2zdxEMLHJl2w$ maybe you as the author can revise this and work on it yourself on top of the new code.

As for the new string functions, I opened this issue https://github.com/pure-data/pure-data/issues/2399 https://urldefense.com/v3/__https://github.com/pure-data/pure-data/issues/2399__;!!Mih3wA!Em8VTuW7HdeNM2CSOJFad7RmuHN-g07GrXnZ7z5i0FhjjUS3-Jr_oOF7qYJF1FQw78Y2zdxEeswcH9Y$

And I would also like to support value names in [expr] with symbols input, so I opened this https://github.com/pure-data/pure-data/issues/2400 https://urldefense.com/v3/__https://github.com/pure-data/pure-data/issues/2400__;!!Mih3wA!Em8VTuW7HdeNM2CSOJFad7RmuHN-g07GrXnZ7z5i0FhjjUS3-Jr_oOF7qYJF1FQw78Y2zdxEhtbo6MI$

What do you think about these issues and requests?

I could try and work on a PR that takes care of it all and send it to you for revision.

cheers

ps. I hear Miller say you were also planning on supporting multichannel connections in [expr~]/[fexpr~], that'd be great, but I assume this could take a while and be a reality in the version 0.59?

-- Shahrokh Yadegari Professor, Music Department Associate Director, Qualcomm Institute Director, Sonic Arts R&D and IDEAS, QI University of California, San Diego Email: sdy@ucsd.edu Web: http://yadegari.org

Em ter., 20 de ago. de 2024 às 18:22, Alexandre Torres Porres < porres@gmail.com> escreveu:

I have been creating new issues, new PRs and writing comments in 'expr' related issues and PRs. to organize things, I created a new 'expr' label for issues and PRs.

Here are them all https://github.com/pure-data/pure-data/labels/expr

there are more expr related issues and PRs but these are the ones I thought it was worth tagging for now.

Em ter., 20 de ago. de 2024 às 01:39, Alexandre Torres Porres < porres@gmail.com> escreveu:

...

Hi, I see you merged doc updates and develop, thanks! Seems that a test release might be coming up soon!

Now, I also see that we have a nice surprise with new features in [expr] for string manipulations! Awesome!

In this context, I think we should enjoy the opportunity and also work on some bugs in expr. I just came across this bug for instance https://github.com/pure-data/pure-data/issues/2398 I also see that the update notes within the expr code mentions *" fixed the expr~ array[0] "bug"* and I wonder what bug that was... when trying to test it is when I came across this bug I reported...

Also, this may seem like a 'feature', but it seems to me is just how things should work intuitively, which is being able to set an array index with audio signals, see https://github.com/pure-data/pure-data/issues/719 and there's already a PR for that https://github.com/pure-data/pure-data/pull/853

And string support in [expr] may also call for support for symbols in [value] and we also have PRs for that, see https://github.com/pure-data/pure-data/pull/817

cheers

Em sex., 19 de jul. de 2024 às 13:40, Miller Puckette < mpuckette@cloud.ucsd.edu> escreveu:

...

To Pd Dev -

What do you think, is it time to put 0.55-1 out?

If I don't hear any objections I'll stick out a "test" one Saturday, and maybe the real one in a week or so.

cheers

M

---

Pd-dev mailing list Pd-dev@lists.iem.at https://lists.puredata.info/listinfo/pd-dev

Seems like we have another serious regression in libpd (caused by yours truely): https://github.com/pure-data/pure-data/issues/2436

Hopefully I can fix this over the next few days.

On 18.09.2024 22:06, Alexandre wrote:

Hi, as I see it, we have serious regression bug in 55-1 which was already fixed in the stable development branch, see https://github.com/pure-data/pure-data/issues/2432

I also have some documentation updates I'm working on. I already fixed some regression bugs I introduced in 55-1 in the audio examples.

Shahrokh is out of town but heading back this friday, and we've been talking of yet more quick features and fixes for [expr] that he could work on in a couple of days.

I think that should grant a new bugfix release, what do you people think? Specially Miller, of course.

cheers

---

pd-dev@lists.iem.at - the Pd developers' mailinglist https://lists.iem.at/hyperkitty/list/pd-dev@lists.iem.at/message/5U6WUVMF75X...