Effectively, what I'm hoping to do is start merging soon, assuming
nobody objects, but to stop admitting new features after a couple of
weeks so that we can have most of April to look at what we've got and
fix bugs (but not try to add stuff). so the "freeze" period would be as
long as people need to feel confident.
OTOH if you want to keep adding features during April while we're
waiting for bug reports, we can start a "0.56" branch and merge it into
master once 0.55 is out. (and yes, I'm imagining numbering it 0.55).
But I can be flexible on all of this.
I'll look at the PRs suggested below... they all sound worthwhile.
cheers
Miller
On 3/12/24 11:49, Christof Ressi wrote:
Hi Miller,
good to hear from you!
Just a question: since you plan to add new features, this would be Pd
0.55 - and not Pd 0.54.2 - right?
I'm thinking of making a release mid April (assuming things go well)
and so I should probably call for a freeze late March.
That's a pretty narrow merge window, given that Easter holidays
already start on March 23, which some of us want to spend with our
families. I really appreciate the announcement, but it would be great
if it could come a bit more in advance (i.e. more than 2 weeks before
feature freeze). We all have jobs and/or families. I love to
contribute to Pd, but I really need to plan ahead of time.
In general, instead of a very dense 2-week merge window twice a year,
it would great to merge PRs on a mere regular basis. Not only would it
cause less stress, it would also give us more time to find bugs before
the actual release. That's just my personal opinion, of course.
---
A few things from my side:
1. Please consider my scheduler improvements:
https://urldefense.com/v3/__https://github.com/pure-data/pure-data/pull/1756__;!!Mih3wA!AAIlAP1eLcTDPzeu6GDEGJ9PsugMcsy9qUAViVvn-Fz-ppZh41nB1WC3rPywVlPjlBzPKX8MpBfXHg$
I have been using this for 1 1/2 years now, both in my daily patching
and in big concerts (including an opera production!), and I can't live
without it anymore. It would be nice if other people could enjoy these
improvements as well. Also, I wouldn't have to hand out custom Pd
versions to my performers anymore :)
---
2. There are quite a few missing multichannel features!
Here are my multichannel PRs:
* MC support for [print~], [snapshot~] and [sig~]:
https://urldefense.com/v3/__https://github.com/pure-data/pure-data/pull/1978__;!!Mih3wA!AAIlAP1eLcTDPzeu6GDEGJ9PsugMcsy9qUAViVvn-Fz-ppZh41nB1WC3rPywVlPjlBzPKX-U2CSP-g$
* MC support for [readsf~] and [writesf~]:
https://urldefense.com/v3/__https://github.com/pure-data/pure-data/pull/2052__;!!Mih3wA!AAIlAP1eLcTDPzeu6GDEGJ9PsugMcsy9qUAViVvn-Fz-ppZh41nB1WC3rPywVlPjlBzPKX-sJBQZ4g$
* MC support for [delwrite~], [delread~] and [delread4~]:
https://urldefense.com/v3/__https://github.com/pure-data/pure-data/pull/2049__;!!Mih3wA!AAIlAP1eLcTDPzeu6GDEGJ9PsugMcsy9qUAViVvn-Fz-ppZh41nB1WC3rPywVlPjlBzPKX8C17P5zA$
* allow to change the number of tables/channels in table DSP objects:
https://urldefense.com/v3/__https://github.com/pure-data/pure-data/pull/2058__;!!Mih3wA!AAIlAP1eLcTDPzeu6GDEGJ9PsugMcsy9qUAViVvn-Fz-ppZh41nB1WC3rPywVlPjlBzPKX8v1GLrmQ$
* signal comparison operators (finally!) with multichannel support:
https://urldefense.com/v3/__https://github.com/pure-data/pure-data/pull/2054__;!!Mih3wA!AAIlAP1eLcTDPzeu6GDEGJ9PsugMcsy9qUAViVvn-Fz-ppZh41nB1WC3rPywVlPjlBzPKX-3K-vwoA$
The [snake~] object is also missing a few crucial features, most
importantly:
* query the number of channels in a MC signal, e.g. [snake~ count]
* combine several MC signals into a single MC signal, e.g. [snake~
join], or extend [snake~ in to accept multichannel signals
* split a MC signal into several MC signals resp. get a subset of
channels, e.g. [snake~ split] resp. [snake~ get]
* sum a MC signal, e.g. [snake~ sum]
People are already implementing these as externals, but these features
seem so basic that they really should be part of Pd vanilla IMO.
For reference, here's the discussion:
https://urldefense.com/v3/__https://github.com/pure-data/pure-data/issues/1996__;!!Mih3wA!AAIlAP1eLcTDPzeu6GDEGJ9PsugMcsy9qUAViVvn-Fz-ppZh41nB1WC3rPywVlPjlBzPKX-g0aWDLg$
---
A few other things I really want to see eventually (not necessarily
for this release):
* more clone improvements:
https://urldefense.com/v3/__https://github.com/pure-data/pure-data/pull/2053__;!!Mih3wA!AAIlAP1eLcTDPzeu6GDEGJ9PsugMcsy9qUAViVvn-Fz-ppZh41nB1WC3rPywVlPjlBzPKX-OXUxWsg$
* "goprect" method:
https://urldefense.com/v3/__https://github.com/pure-data/pure-data/pull/627__;!!Mih3wA!AAIlAP1eLcTDPzeu6GDEGJ9PsugMcsy9qUAViVvn-Fz-ppZh41nB1WC3rPywVlPjlBzPKX8r3LgY7Q$
. Solves a real issue and lying around for 5 years now.
* namespace constructors for all external objects:
https://urldefense.com/v3/__https://github.com/pure-data/pure-data/pull/630__;!!Mih3wA!AAIlAP1eLcTDPzeu6GDEGJ9PsugMcsy9qUAViVvn-Fz-ppZh41nB1WC3rPywVlPjlBzPKX8RsK_Cbg$
. Solves a real issue and lying around for 5 years.
Cheers,
Christof
PS: here is a full list of my open PRs, in case anyone is interested:
https://urldefense.com/v3/__https://github.com/pure-data/pure-data/pulls/Spacechild1?page=1&q=is*3Aopen*is*3Apr*author*3ASpacechild1__;JSslKyU!!Mih3wA!AAIlAP1eLcTDPzeu6GDEGJ9PsugMcsy9qUAViVvn-Fz-ppZh41nB1WC3rPywVlPjlBzPKX8aqw5N3w$
On 12.03.2024 08:47, Miller Puckette wrote:
To Pd dev -
I'm thinking of making a release mid April (assuming things go well)
and so I should probably call for a freeze late March. As usual I
plan to merge in "devel" and "Documentation" - in fact I should do a
first merge rather soon, assuming things are in a good state for
merging.
I'm planning to add a couple of features: 1. message to Pd to toggle
between GUI and no-GUI -- perhaps with a way to reset the GUI startup
command -- so that if you have a headless installation that's doing
something funny you can pop it open and look; and 2. improvements to
the "pointer" object to make it easier to get around data structures,
and possibly a menu extension for dragging new "data" onto the
screen; 3. an optional pop-up display showing (x,y) coordinates of
object or data knob being dragged.
Incidentally: I just noticed that the IEM slider object (and proabbly
other EM GUIs) spits out a number when clicked upon, even if not
dragged. Is this desirable behavior? It caught me out buit perhaps
other users are actually wanting to be able to click on a control to
repeat its value. Hmm..
cheers
Miller
_______________________________________________
Pd-dev mailing list
[email protected]
https://urldefense.com/v3/__https://lists.puredata.info/listinfo/pd-dev__;!!Mih3wA!AAIlAP1eLcTDPzeu6GDEGJ9PsugMcsy9qUAViVvn-Fz-ppZh41nB1WC3rPywVlPjlBzPKX-OcAJhxQ$
_______________________________________________
Pd-dev mailing list
[email protected]
https://urldefense.com/v3/__https://lists.puredata.info/listinfo/pd-dev__;!!Mih3wA!AAIlAP1eLcTDPzeu6GDEGJ9PsugMcsy9qUAViVvn-Fz-ppZh41nB1WC3rPywVlPjlBzPKX-OcAJhxQ$
_______________________________________________
Pd-dev mailing list
[email protected]
https://lists.puredata.info/listinfo/pd-dev
