Current state of Pyo and next steps?

https://www.youtube.com/channel/UC_utjGYbSizWE0dNyr0Vdmg

--
You received this message because you are subscribed to the Google Groups "pyo-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pyo-discuss...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/pyo-discuss/2c43dea7-c047-467f-b223-664cd2d6da10%40gmail.com.

Gabriele Battaglia

unread,

Dec 15, 2025, 11:55:31 AM12/15/25

to pyo-d...@googlegroups.com

Ciao Aaron,

Based on my experience, nothing higher than Python 3.11 works.

Best,
Gabriele

Inviato dalla nave Gabryphone17ProMax.

To view this discussion visit https://groups.google.com/d/msgid/pyo-discuss/CAFKYMmwescRTZx7_apRoFgqD8C5nwM9CdzPkbF9qnrxc3gLY%2BQ%40mail.gmail.com.

Chris Csikszentmihalyi

unread,

Dec 15, 2025, 12:33:36 PM12/15/25

to pyo-d...@googlegroups.com

This would be very sad. One of the things I love about PYO is that it integrates well into a real programming language without having to do anything special. I am using it for a project with ROS2 (robotics) and nicegui, for example. Sure, one can eventually get anything to talk to anything, but with PYO I can get a system together in a day with all of python's libraries, and quick scripting at hand.

I have been hyping PYO for a decade but sadly it doesn't seem to catch on widely. I think, frankly, it needed a few influencers who could have shown what was possible, especially by leaning on python.

If Olivier is just taking a pause, I would love to hear from him! Meanwhile I thank him for powering several of my projects. The API he created is the most elegant and sane one I have seen.

C.

Chris Csíkszentmihályi

rob...@gmail.com | edgyproduct.org

"Art means… to resist the course of a world that unceasingly holds a gun to mankind's chest."
--Theodore Adorno

To view this discussion visit https://groups.google.com/d/msgid/pyo-discuss/CAMMNeJLYtmiR4qKvZ1X5YMebivQtKVeUG9GQN%3DT0hx66dn%2BsFw%40mail.gmail.com.

barmin

unread,

Dec 15, 2025, 12:44:35 PM12/15/25

to pyo-d...@googlegroups.com

Hello,

I thought this list was dead, but obviously it works again! :-)

I had a contact with Olivier regarding these questions a couple of months ago.

I can't speak for Olivier, but AFAICT he is willing to go on with the development of pyo, although he is currently very busy and cannot always keep up.

Note that he pre-published pyo 1.0.6, compatible with python 3.9 to 3.13, for testing (https://github.com/belangeo/pyo/discussions/293). He told me he didn't publish it officially yet because he didn't have enough feedback from the community regarding this test version. Presumably he's not the only one to be too busy!

(I'm pleading guilty here: although I did ask for an update compatible with latest python versions, I did not test 1.0.6 yet.)

BUT the good news is: at Les Chemins de Traverse (https://www.lescheminsdetraverse.net/), we are currently planning a project about the "Charming Snake" framework, based on pyo (https://bitbucket.org/MatthieuAmiguet/charmingsnake/src/master/). We submitted an application for a grant by the Swiss Arts Council. Olivier is part of the project, and if we get the money, we could fund some development work for pyo too. So hopefully in early 2026, we should see some news on the pyo front.

Don't give up on pyo too soon, as it's one of the best tools out there!

And, fingers crossed! Let's hope we get some funding to boost pyo development!

Cheers,

Matthieu

Le 15.12.25 à 13:30, Alexandros Drymonitis a écrit :

> --
> You received this message because you are subscribed to the Google Groups "pyo-discuss" group.

> To unsubscribe from this group and stop receiving emails from it, send an email to pyo-discuss...@googlegroups.com <mailto:pyo-discuss...@googlegroups.com>.
> To view this discussion visit https://groups.google.com/d/msgid/pyo-discuss/9101757b-6511-4c7f-95a9-d9841d46d2afn%40googlegroups.com <https://groups.google.com/d/msgid/pyo-discuss/9101757b-6511-4c7f-95a9-d9841d46d2afn%40googlegroups.com?utm_medium=email&utm_source=footer>.

barmin

unread,

Dec 15, 2025, 1:00:19 PM12/15/25

to pyo-d...@googlegroups.com

> I think, frankly, it needed a few influencers who could have shown what was possible, especially by leaning on python.

I'm afraid I'm not an influencer, but I truly did my best: talk in EuroPython 2025 in Praha, talk in PyConDE 2024 in Berlin, talk and gig in EuroPython 2019 in Basel, talk at the Swiss Python Summit 2016, plus several wider-audience talks (the videos and slides of all the talks are available here if you're interested: https://matthieuamiguet.ch/a-propos/)

I also invited Olivier for a talk about pyo at RFLA'16 in Neuchâtel: http://www.augmented-instruments.net/rfla2016

And wrote 15+ posts related to pyo on my blog since 2013: https://matthieuamiguet.ch/tags/pyo/

I did "convert" a few people along the way, but I'm afraid pyo definitely stays a niche within a niche...

Cheers,

Matthieu

Alexandros Drymonitis

unread,

Dec 15, 2025, 1:09:35 PM12/15/25

to pyo-d...@googlegroups.com

My opinion is that Pyo would greatly benefit from a wider developer
community than Olivier alone. I know it's his creation and I respect
that he possibly wants it in a certain way so he might want to have full
control over it. Having a developer community around though doesn't
necessarily mean loosing control. Take Pure Data for example, Miller
Puckette is probably the "Benevolent dictator for life"
(https://en.wikipedia.org/wiki/Benevolent_dictator_for_life) while a
number of developers contribute code and help with compiling for various
systems.

My coding skills are rather limited, but I would definitely be in to
become a member of Pyo's developer community where I would help in any
way I can.

As for influencers, I have written a book on Pyo
(https://www.routledge.com/The-Python-Audio-Cookbook-Recipes-for-Audio-Scripting-with-Python/Drymonitis/p/book/9781032480114),
but I'm not sure how much this has contributed in spreading the word. I
also try to introduce it to academic institutions where I occasionally
teach, but that's that.

It's great that Matthieu was in contact with Olivier only two months
ago. We could also organize an online meeting with anyone who is
interested in contributing in development or maintenance and see what
steps we can take. Of course, it would be best if Olivier could attend.

I know I might sound like I want to take over, but that's not the case.
What I really want is that Pyo stays alive!

Aaron Krister Johnson

unread,

Dec 15, 2025, 1:39:21 PM12/15/25

to pyo-d...@googlegroups.com

Absolutely this. Before I heard anything about Olivier's signs-of-life, I was ready at some point to fork the repo and attempt to hack it into current-ness.

Glad that we possibly don't have to go this route.

In any event, count me in on discussion regarding widening the development pool and giving Olivier help.

One idea I had would be to create a framework via Faust (which would really mostly entail a PR to Faust itself, but would perhaps require changes in Pyo to make it "pluggable") to extend Pyo. I know there's docs for that in Pyo, but a way to make this "fluid" and "semi-automated" would be nice. Dreams for the future, anyhow...

Aaron Krister Johnson
Music, etc.:
https://soundcloud.com/aaron-krister-johnson

https://www.youtube.com/channel/UC_utjGYbSizWE0dNyr0Vdmg

https://github.com/seanwayland

To unsubscribe from this group and stop receiving emails from it, send an email to pyo-discuss...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/pyo-discuss/e0530e35-7f03-4d86-8858-ad0644bda448%40gmail.com.

Alexandros Drymonitis

unread,

Dec 16, 2025, 3:13:32 AM12/16/25

to pyo-d...@googlegroups.com

Matthieu, is it possible to get hold of Olivier and pass our concerns on to him? I cannot know what he thinks of all this, but both Aaron and I have been thinking of forking Pyo's repo to keep it alive, so I think having a meeting wouldn't be that bad (unless he really has no time at all). Still, I think it's nice that people are so interested in his software, so I only see good in all this.

To view this discussion visit https://groups.google.com/d/msgid/pyo-discuss/CAFKYMmx94PLLnCDCzX1qha0NH8CqAnTTEv7Fd-0K%2B5kCGe1JLg%40mail.gmail.com.

Alexandros Drymonitis

unread,

Jan 5, 2026, 3:44:41 AMJan 5

to pyo-discuss

Hello and happy new year!

With the new year, I'm hoping we can get Pyo to an active state again! Anyone has any news on this? Anyone else besides Aaron and me is interested in having an online meeting so we can discuss on what can be done?

Matthieu, since you are probably the only person that is (was?) in contact with Olivier, have you reached him?

Cheers,
Alexandros

p.s...@outlook.com

unread,

Jan 5, 2026, 1:00:08 PMJan 5

to pyo-discuss

Hi all,

Just piping in to say thanks for keeping up the conversation about this. I don't have the skills to do development work, but I use a few basic elements of pyo in my live music performance code still (taking after Matthieu, whose writings introduced me to it). I managed to build 1.0.6 from sources recently. I'll link to my use of pyo from the Github discussion area. It's unfortunate Olivier is not able to keep it up, but I fully understand it and am grateful for whatever work continues to be done by him and others.

I understood this Google Group to have been completely abandoned in favor of the Github discussions, due to bot spam. I hope somebody at least made a few dollars off of destroying our active forum. (sarcasm).

Paul

Alexandros Drymonitis

unread,

Jan 5, 2026, 3:49:05 PMJan 5

to pyo-d...@googlegroups.com

On Mon, 2026-01-05 at 10:00 -0800, p.s...@outlook.com wrote:
> Hi all,
>
> Just piping in to say thanks for keeping up the conversation about
> this. I don't have the skills to do development work, but I use a few
> basic elements of pyo in my live music performance code still (taking
> after Matthieu, whose writings introduced me to it). I managed to
> build 1.0.6 from sources recently. I'll link to my use of pyo from
> the Github discussion area. It's unfortunate Olivier is not able to
> keep it up, but I fully understand it and am grateful for whatever
> work continues to be done by him and others.

To be clear on this, Olivier has never stated that he has abandoned
Pyo. He has gone AWOL sort of, but Pyo being abandoned is not official.
I took the initiative to start this conversation as an attempt to get
Pyo up and running again.

p.s...@outlook.com

unread,

Jan 5, 2026, 5:47:06 PMJan 5

to pyo-discuss

Apologies for the misstatement.

barmin

unread,

Jan 15, 2026, 9:48:37 AMJan 15

to pyo-d...@googlegroups.com

Hi there,

Happy new year! :-D

Le 05.01.26 à 09:44, Alexandros Drymonitis a écrit :

> Matthieu, since you are probably the only person that is (was?) in contact with Olivier, have you reached him?

I did not have any contact with Olivier recently, but I expect it to happen in the next few weeks. I'll make sure to relay to him what's being discussed here in case he missed it and I'll certainly report here any news I might have.

Cheers,

Matthieu

Sean Wayland

unread,

Jan 15, 2026, 8:21:32 PMJan 15

to pyo-d...@googlegroups.com

Hi all,

I am a python dev at work and write a bit of C++ for fun have released a few plugins.

I downloaded a zip of the repo and showed "claude" the repo and asked "her" to take a look.

Supporting all the different OS versions may be tricky for me . I am on ventura 13.2.1 and have a windows computer somewhere who knows what OS that is.

It would take a community of developers with access to different operating systems to attempt to get some sort of stable release going.

We could create some sort of unoffical fork and contribute branches and work through some of the issues and try and keep it alive.

Perhaps with the AI tools it might not take too much time and hopefully we can avoid filling it with slop.

This is what claude says about the code base :)

Best Sean

```
Read project configuration and source files
Let me check the C extension code and see if there are any Python version compatibility issues:

Searched for regex patterns and read files
Now let me check for any CI/CD or test configuration files:

Searched for configuration files and read YAML
Now let me look at the actual C code to understand dependencies on deprecated Python APIs:

Searched for regex PyObject_NewVar|PyMem|Py_TPFLAGS|tp_new|tp_dealloc|Py_ssize_t|PyVarObject|PyObject_New (**/*.c), 30 results
Now let me check the dependencies (portaudio, libsndfile, etc.) to see if they support Python 3.14:

python3 --version
Based on my analysis of the codebase, here's a comprehensive overview:

Summary: Python 3.14 Support for Pyo
Current Status: Pyo supports Python 3.9-3.13 (already updated for 3.12 & 3.13 in the codebase).

What Would Be Required for Python 3.14 Support
1. Minimal C Extension Changes (~2-4 hours effort)
The C code appears well-structured and uses standard Python C API that's generally backward-compatible:

Uses PY_SSIZE_T_CLEAN macro (good practice)
Standard PyVarObject_HEAD_INIT pattern
Uses PyMem_* allocation functions correctly
No deprecated APIs like PyInt_* or old-style type objects
You'd mainly need to:

Test compilation against Python 3.14 headers
Verify no new deprecation warnings appear
Potentially address any new compiler warnings
2. Configuration Updates (~30 minutes)
pyproject.toml: Add "Programming Language :: Python :: 3.14" classifier and ensure requires-python = ">=3.9, <4" (already correct)

CI/CD Workflow: Update build_manylinux_wheels.yml to include cp314-cp314 in the Python versions list

3. Dependency Verification (~2-4 hours)
The critical dependencies you'd need to verify support Python 3.14 stable ABI:

libsndfile - Core audio file I/O (mature, likely OK)
portaudio - Audio driver abstraction (mature, likely OK)
portmidi - MIDI support (needs checking)
liblo - OSC support (needs checking)
JACK - Professional audio server (needs checking)
4. Testing (~4-8 hours)
Run unit test suite against Python 3.14
Test example scripts
Test each audio backend (PortAudio, JACK, CoreAudio, etc.)
Maintenance Effort Assessment:
Low to Moderate (~40-60 hours of work per new Python release):

Per New Python Release:

✅ Automated - CI/CD changes (1-2 hours)
✅ Automated - Dependency rebuilds via manylinux (mostly automated via GitHub Actions)
✅ Simple - Configuration updates (0.5-1 hour)
⏱️ Manual - Testing on multiple platforms (8-12 hours)
⏱️ Variable - Potential C API deprecation fixes (0-20 hours, depending on Python changes)
Ongoing Maintenance:

Monitor Python 3.14 releases and deprecation notices
Update CI matrix with each new minor version
Test quarterly or with each release
Track dependency updates (portaudio, portmidi, etc.)
Why It's Relatively Low Effort:
Stable C API: Pyo doesn't use PY_UNSTABLE_API features; it uses the stable ABI
Mature Codebase: No cutting-edge Python features; sticks to proven patterns
Mature Dependencies: All external libs (portaudio, libsndfile) are established and well-maintained
Already Multi-Version: Built to handle 3.9→3.13 range; Python 3.14 wouldn't introduce major new changes
Potential Risks:
Deprecation of tp_print (already removed in 3.11) - appears not used
GIL changes in Python 3.13+ - shouldn't affect this audio library
PEP 703 free-threading - Pyo uses threading but not in ways that would be affected
Bottom line: Adding Python 3.14 support would be straightforward; the challenge is ongoing testing across platforms (Linux, macOS ARM64, macOS x86_64, Windows) rather than code changes.
```

--
You received this message because you are subscribed to the Google Groups "pyo-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pyo-discuss...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/pyo-discuss/bbf27faf-5c99-497a-b870-6061ff5f19d3%40bluewin.ch.

--

http://seanwayland.com

Sean Wayland

US cell phone: +1 (347)5231455

Alexandros Drymonitis

unread,

Jan 16, 2026, 1:23:29 AMJan 16

to pyo-d...@googlegroups.com

Thanks a lot Matthieu, that would be great!

Alexandros Drymonitis

unread,

Jan 16, 2026, 1:29:40 AMJan 16

to pyo-d...@googlegroups.com

Hi Sean,
I've had a video call with Aaron about this and we decided to fork Pyo
and try to keep it at a stable state, plus try to compile it for
Python3.14. Matthieu replied yesterday that he'll be in touch with
Olivier in the next few weeks, so I suggest that we wait to see what he
has to say about all this.

I also believe that a project like Pyo needs a greater developer
community, but I would definitely like to have Olivier's opinion on
this. After all, it's his creation and we respect that.

On what you wrote, I've also used AI a bit, but would rather stay away
from it as much as possible. The way I see it, sometimes it takes too
many iterations to get something right out of it, that not using it
could take less time. On the other hand, I've used it when I wanted to
do stuff I didn't know how to approach it, and either it only gave me
an idea of how to deal with it, or (rarely) it gave me code that works.
This applies to integrating Python and Pyo to C++ in openFrameworks,
where I used ChatGPT to get code that would give me access to Python's
STDOUT and STDERR inside C++. After quite a few iterations, it worked.

I don't know what the rest think, but personally I'd keep the code as
much AI-free as possible.

p.s...@outlook.com

unread,

Jan 16, 2026, 8:46:12 AMJan 16

to pyo-discuss

Excellent! I look forward to hearing what Olivier suggests.

Re AI: given the years of deliberate attention that Olivier has put into this, I would really much better trust him to suggest these steps to other developers here and flag some likely issues. I'm also confused that the "Configuration updates" section seems to be suggesting a change of two lines, that it's written out for you, that would take 30 minutes (revised a tiny bit further down to 1 hour). I can understand use of AI as a launch point for a new project or new idea, as Alexandros points to, but I would not trust it to apply routine maintenance especially without the primary developer's input.

Paul

Sean Wayland

unread,

Jan 16, 2026, 10:56:07 AMJan 16

to pyo-d...@googlegroups.com

I don't have a lot of time between writing code for a living and playing music for fun and the new AI tools make managing all sorts of code so much easier. its a huge productivity at work and also at home having fun. you still have to read the code and understand what its doing. would be lovely to keep pyo going and up to date

--

You received this message because you are subscribed to the Google Groups "pyo-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pyo-discuss...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/pyo-discuss/059c9375-8325-49f0-b6a7-179b63c45ff9n%40googlegroups.com.

Alexandros Drymonitis

unread,

Feb 5, 2026, 11:08:43 AMFeb 5

to pyo-d...@googlegroups.com

Hi Matthieu,

Did you have any news from Olivier?

Aaron Krister Johnson

unread,

Feb 14, 2026, 11:06:10 AM (10 days ago) Feb 14

to pyo-d...@googlegroups.com

Hi all,

Before I go too far down a rabbit-hole with this idea of getting pyo "bleeding-edge" to python-3.14 (I am busy with many other projects ATM, and I need to know my time spent on this is actually used by others) I noticed Gabriele Battaglia said that pyo wasn't working for him on python > 3.11.

However, in my own experience, this is decidedly not true, and others have confirmed they have no problem with pyo, as advertised, going up to 3.13.

Perhaps, Mr. Battaglia, you need assistance getting to 3.13? Before we take on 3.14, which is the work of at least an afternoon, we want the entire community to be comfortable and working on 3.13. 3.13 isn't bleeding edge, but sufficiently new enough. I'm using 3.13 at work, for instance.

Aaron Krister Johnson
Music, etc.:
https://soundcloud.com/aaron-krister-johnson

https://www.youtube.com/channel/UC_utjGYbSizWE0dNyr0Vdmg

--
You received this message because you are subscribed to the Google Groups "pyo-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pyo-discuss...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/pyo-discuss/d2d5ef31-09d6-4c3b-a2b5-3aaad9f6b3b2%40gmail.com.

Aaron Krister Johnson

unread,

Feb 14, 2026, 1:32:55 PM (10 days ago) Feb 14

to pyo-d...@googlegroups.com

I did some work on my fork this morning. I've verified the building is fine on `python-3.13`. It will take more work to bring pyo up to 3.14, b/c there are some breaking API changes for C-modules.

However, the following is visible on my fork-branch as of git commit `3c774be`:

`setup.py` is a bit more flexible now, allowing for dynamically working with a python instance (and its includes and libs) that happened to be installed via `pyenv`
A new script `scripts/get-pyo.sh` that can be pulled from the repo separately, put into a download directory, and run, so that it clones `pyo` and builds it for Linux or Mac. (N.B. you're own you own, `Windows` users, I won't touch your platform with a 10-foot pole, sorry -- for starters, it's a nightmare platform to build software on, even worse than the already poor `MacOS`...)
Documentation cleanup, fixing grammar and spelling errors
Reformatting the entire codebase of `python` via `ruff format --line-length 100`. This prettifies the code quite a bit

Next weekend or thereabout, I'll attempt 3.14 integration and builds, but I'm burnt for the day on the maintenance stuff.

Feel free to pull from my fork: https://github.com/akjmicro/pyo

Aaron Krister Johnson
Music, etc.:
https://soundcloud.com/aaron-krister-johnson

https://www.youtube.com/channel/UC_utjGYbSizWE0dNyr0Vdmg

Gabriele Battaglia

unread,

Feb 15, 2026, 12:10:27 AM (9 days ago) Feb 15

to pyo-d...@googlegroups.com

Il 14/02/2026 17:05, Aaron Krister Johnson ha scritto:
> Hi all,
>
> Before I go too far down a rabbit-hole with this idea of getting pyo
> "bleeding-edge" to python-3.14 (I am busy with many other projects
> ATM, and I need to know my time spent on this is actually used by
> others) I noticed Gabriele Battaglia said that pyo wasn't working for
> him on python > 3.11.
>
> However, in my own experience, this is decidedly not true, and others
> have confirmed they have no problem with pyo, as advertised, going up
> to 3.13.
>
> Perhaps, Mr. Battaglia, you need assistance getting to 3.13? Before we
> take on 3.14, which is the work of at least an afternoon, we want the
> entire community to be comfortable and working on 3.13. 3.13 isn't
> bleeding edge, but sufficiently new enough. I'm using 3.13 at work,
> for instance.
>
>

Hi Aaron, hi everyone,
Thank you, Aaron, for your message and for the offer to help.
I would like to clarify what happened and, most importantly, apologize
to you and the entire mailing list. When I first tried to install pyo, I
was using Python 3.11.9 and pip kept returning errors. At that time, I
asked here on the list and was under the impression that a compatible
version for 3.11.9 wasn't available yet.
However, after reading your email, I did a fresh test just a few minutes
ago:
1.
pyenv local 3.11.9
2.
pip install pyo
Everything worked perfectly and version 1.0.5 installed without any issues.
It's clear that my previous statement was incorrect and based on a
personal misinterpretation or a temporary local issue. I am very sorry
if this caused any confusion or extra work for the developers. It was
certainly not my intention to provide misinformation; I simply made a
mistake in my evaluation.
I hope you and the community can accept my apologies. I will be much
more careful with my reports in the future.
Best regards,
Gabriele

Aaron Krister Johnson

unread,

Feb 15, 2026, 10:44:09 AM (9 days ago) Feb 15

to pyo-d...@googlegroups.com

Gabriele -- no worries, not at all! I'm just happy you are able to use Pyo again!

And, this has spurred important conversations about getting Pyo unblocked for Python >= 3.14.

Aaron Krister Johnson
Music, etc.:
https://soundcloud.com/aaron-krister-johnson

https://www.youtube.com/channel/UC_utjGYbSizWE0dNyr0Vdmg