To Pd-announce:
Pd version 0.55-0 test2 is available from http://msp.ucsd.edu/software.htm or (source only) via github: https://github.com/pure-data/pure-data
There was no "test1" - it never made it through the release process. Also, for Mac users, the app will show up as "Pd-0.55-0test2c" but in fact it's just test2.
cheers Miller
Pd-announce mailing list Pd-announce@lists.iem.at https://lists.puredata.info/listinfo/pd-announce
One observation:
The PD Window doesn’t display the text from the external libraries that are loaded at startup anymore. The libraries are still visible in Preferences > Startup and work as usual.
Best!
Edwin
On 14 May 2024, at 16:23, Miller Puckette mpuckette@cloud.ucsd.edu wrote:
To Pd-announce:
Pd version 0.55-0 test2 is available from http://msp.ucsd.edu/software.htm or (source only) via github: https://github.com/pure-data/pure-data
There was no "test1" - it never made it through the release process. Also, for Mac users, the app will show up as "Pd-0.55-0test2c" but in fact it's just test2.
cheers Miller
Pd-announce mailing list Pd-announce@lists.iem.at https://lists.puredata.info/listinfo/pd-announce _______________________________________________ Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list
Em ter., 14 de mai. de 2024 às 12:17, Edwin van der Heide pd@evdh.net escreveu:
The PD Window doesn’t display the text from the external libraries that are loaded at startup anymore. The libraries are still visible in Preferences > Startup and work as usual.
I can confirm it doesn't load the texts, but that is not the only issue. For 'ELSE', loading the library at startup also loads a plugin for object browsing and this doesn't work anymore. I think the order of loading things have changed and we're skipping an important step.
cheers
Hey
Thanks to everyone involved in that release. All the things mentioned in the release notes are much appreciated changes/improvements, especially the new audio back-end handling and scheduler.
On Tue, 2024-05-14 at 17:15 +0200, Edwin van der Heide wrote:
One observation:
The PD Window doesn’t display the text from the external libraries that are loaded at startup anymore. The libraries are still visible in Preferences > Startup and work as usual.
It appears the Pd console suppresses all output from start-up, including library messages, error messages, stuff printed with [print] triggerd by [loadbang]. Everything is printed normally to the terminal when using -stderr.
Hm.. it's not only suppressed at [loadbang]. Even a message triggered by loadbang sent to server and sent back is not printed.
Roman
Am 14. Mai 2024 22:57:01 MESZ schrieb Roman Haefeli reduzent@gmail.com:
On Tue, 2024-05-14 at 17:15 +0200, Edwin van der Heide wrote:
One observation:
The PD Window doesn’t display the text from the external libraries that are loaded at startup anymore. The libraries are still visible in Preferences > Startup and work as usual.
It appears the Pd console suppresses all output from start-up, including library messages, error messages, stuff printed with [print] triggerd by [loadbang]. Everything is printed normally to the terminal when using -stderr.
Hm.. it's not only suppressed at [loadbang]. Even a message triggered by loadbang sent to server and sent back is not printed.
Roman
https://github.com/pure-data/pure-data/issues/2300
mfg.sfg.jfd IOhannes
On Wed, 2024-05-15 at 06:04 +0200, IOhannes m zmölnig wrote:
Am 14. Mai 2024 22:57:01 MESZ schrieb Roman Haefeli reduzent@gmail.com:
It appears the Pd console suppresses all output from start-up, including library messages, error messages, stuff printed with [print] triggerd by [loadbang]. Everything is printed normally to the terminal when using -stderr.
Roman
👍
Roman
- Miller Puckette mpuckette@cloud.ucsd.edu [2024-05-14 16:24]:
To Pd-announce:
Pd version 0.55-0 test2 is available from http://msp.ucsd.edu/software.htm
Thanks Miller,
just reading the changelog at https://msp.ucsd.edu/Pd_documentation/x5.htm#s5.1 I wonder what the updates to the audio interfacing and scheduler are?
best, Peter
I wonder what the updates to the audio interfacing and scheduler are?
See https://github.com/pure-data/pure-data/pull/1756.
Christof
- Christof Ressi info@christofressi.com [2024-05-14 18:40]:
I wonder what the updates to the audio interfacing and scheduler are?
Thanks, amazing! best, P
What is the minimal Windows OS version you can install Pd (the 64 bit one)? Every other option seems to mention the minimum OS, but this one is missing.
cheers
Em ter., 14 de mai. de 2024 às 11:26, Miller Puckette < mpuckette@cloud.ucsd.edu> escreveu:
To Pd-announce:
Pd version 0.55-0 test2 is available from http://msp.ucsd.edu/software.htm or (source only) via github: https://github.com/pure-data/pure-data
There was no "test1" - it never made it through the release process. Also, for Mac users, the app will show up as "Pd-0.55-0test2c" but in fact it's just test2.
cheers Miller
Pd-announce mailing list Pd-announce@lists.iem.at https://lists.puredata.info/listinfo/pd-announce _______________________________________________ Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list
It used to be XP, but since Pd 0.54 it's Vista because of the new WASAPI backend. The 32-bit version, however, still runs on XP because it is built without WASAPI.
On 14.05.2024 20:13, Alexandre Torres Porres wrote:
What is the minimal Windows OS version you can install Pd (the 64 bit one)? Every other option seems to mention the minimum OS, but this one is missing.
cheers
Em ter., 14 de mai. de 2024 às 11:26, Miller Puckette mpuckette@cloud.ucsd.edu escreveu:
To Pd-announce: Pd version 0.55-0 test2 is available from http://msp.ucsd.edu/software.htm or (source only) via github: https://github.com/pure-data/pure-data There was no "test1" - it never made it through the release process. Also, for Mac users, the app will show up as "Pd-0.55-0test2c" but in fact it's just test2. cheers Miller _______________________________________________ Pd-announce mailing list Pd-announce@lists.iem.at https://lists.puredata.info/listinfo/pd-announce _______________________________________________ Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management ->https://lists.puredata.info/listinfo/pd-list
Em ter., 14 de mai. de 2024 às 15:19, Christof Ressi info@christofressi.com escreveu:
It used to be XP, but since Pd 0.54 it's Vista because of the new WASAPI backend. The 32-bit version, however, still runs on XP because it is built without WASAPI.
Awesome, Miller, I think it seems worth mentioning "Windows Vista or newer" in http://msp.ucsd.edu/software.htm
On 14.05.2024 20:13, Alexandre Torres Porres wrote:
What is the minimal Windows OS version you can install Pd (the 64 bit one)? Every other option seems to mention the minimum OS, but this one is missing.
cheers
Em ter., 14 de mai. de 2024 às 11:26, Miller Puckette < mpuckette@cloud.ucsd.edu> escreveu:
To Pd-announce:
Pd version 0.55-0 test2 is available from http://msp.ucsd.edu/software.htm or (source only) via github: https://github.com/pure-data/pure-data
There was no "test1" - it never made it through the release process. Also, for Mac users, the app will show up as "Pd-0.55-0test2c" but in fact it's just test2.
cheers Miller
Pd-announce mailing list Pd-announce@lists.iem.at https://lists.puredata.info/listinfo/pd-announce _______________________________________________ Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list
_______________________________________________Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list
thanks a lot for the test release! what i noticed here (apple m2) in addition to the missing console output: pd now takes ~5s to start. not a big deal obviously, but it's a huge difference compared to 0.54, where the gui is ready after ~1s.
cheers, ben
Am Di., 14. Mai 2024 um 16:26 Uhr schrieb Miller Puckette mpuckette@cloud.ucsd.edu:
To Pd-announce:
Pd version 0.55-0 test2 is available from http://msp.ucsd.edu/software.htm or (source only) via github: https://github.com/pure-data/pure-data
There was no "test1" - it never made it through the release process. Also, for Mac users, the app will show up as "Pd-0.55-0test2c" but in fact it's just test2.
cheers Miller
Pd-announce mailing list Pd-announce@lists.iem.at https://lists.puredata.info/listinfo/pd-announce _______________________________________________ Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list
Hmm... I'm not getting that on mine. The startup order is indeed different -
the GUI doesn't start up until after any startup stuff has been done (loading libraries,
etc). But I don't _think_ the overall time spent is different - do let me know if you're
finding otherwise :)
Miller
On 5/15/24 10:56 AM, Benjamin Wesch wrote:
thanks a lot for the test release! what i noticed here (apple m2) in addition to the missing console output: pd now takes ~5s to start. not a big deal obviously, but it's a huge difference compared to 0.54, where the gui is ready after ~1s.
cheers, ben
Am Di., 14. Mai 2024 um 16:26 Uhr schrieb Miller Puckette mpuckette@cloud.ucsd.edu:
To Pd-announce:
Pd version 0.55-0 test2 is available from http://msp.ucsd.edu/software.htm or (source only) via github: https://urldefense.com/v3/__https://github.com/pure-data/pure-data__;!!Mih3w...
There was no "test1" - it never made it through the release process. Also, for Mac users, the app will show up as "Pd-0.55-0test2c" but in fact it's just test2.
cheers Miller
Pd-announce mailing list Pd-announce@lists.iem.at https://urldefense.com/v3/__https://lists.puredata.info/listinfo/pd-announce... _______________________________________________ Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://urldefense.com/v3/__https://lists.puredata.info/listinfo/pd-list__;!...
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://urldefense.com/v3/__https://lists.puredata.info/listinfo/pd-list__;!...
hm, that's really weird. i tried to objectively measure here now with a script that waits for the "connected" output when starting both versions in parallel and i consistently get results quite similar to this:
pd-0.54-1.log: 0.99s pd-0.55-0.log: 2.72s
so my subjective impression was certainly wrong with 5s - but there's still a noticeable difference.
cheers, ben
Am Mi., 15. Mai 2024 um 11:51 Uhr schrieb Miller Puckette mpuckette@cloud.ucsd.edu:
Hmm... I'm not getting that on mine. The startup order is indeed different -
the GUI doesn't start up until after any startup stuff has been done (loading libraries,
etc). But I don't _think_ the overall time spent is different - do let me know if you're
finding otherwise :)
Miller
On 5/15/24 10:56 AM, Benjamin Wesch wrote:
thanks a lot for the test release! what i noticed here (apple m2) in addition to the missing console output: pd now takes ~5s to start. not a big deal obviously, but it's a huge difference compared to 0.54, where the gui is ready after ~1s.
cheers, ben
Am Di., 14. Mai 2024 um 16:26 Uhr schrieb Miller Puckette mpuckette@cloud.ucsd.edu:
To Pd-announce:
Pd version 0.55-0 test2 is available from http://msp.ucsd.edu/software.htm or (source only) via github: https://urldefense.com/v3/__https://github.com/pure-data/pure-data__;!!Mih3w...
There was no "test1" - it never made it through the release process. Also, for Mac users, the app will show up as "Pd-0.55-0test2c" but in fact it's just test2.
cheers Miller
Pd-announce mailing list Pd-announce@lists.iem.at https://urldefense.com/v3/__https://lists.puredata.info/listinfo/pd-announce... _______________________________________________ Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://urldefense.com/v3/__https://lists.puredata.info/listinfo/pd-list__;!...
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://urldefense.com/v3/__https://lists.puredata.info/listinfo/pd-list__;!...
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list
It feels slower to me too (on Mac at least) - not sure why but I'd like to figure it out.
cheers
Miller
On 5/15/24 1:25 PM, Benjamin Wesch wrote:
hm, that's really weird. i tried to objectively measure here now with a script that waits for the "connected" output when starting both versions in parallel and i consistently get results quite similar to this:
pd-0.54-1.log: 0.99s pd-0.55-0.log: 2.72s
so my subjective impression was certainly wrong with 5s - but there's still a noticeable difference.
cheers, ben
Am Mi., 15. Mai 2024 um 11:51 Uhr schrieb Miller Puckette mpuckette@cloud.ucsd.edu:
Hmm... I'm not getting that on mine. The startup order is indeed different -
the GUI doesn't start up until after any startup stuff has been done (loading libraries,
etc). But I don't _think_ the overall time spent is different - do let me know if you're
finding otherwise :)
Miller
On 5/15/24 10:56 AM, Benjamin Wesch wrote:
thanks a lot for the test release! what i noticed here (apple m2) in addition to the missing console output: pd now takes ~5s to start. not a big deal obviously, but it's a huge difference compared to 0.54, where the gui is ready after ~1s.
cheers, ben
Am Di., 14. Mai 2024 um 16:26 Uhr schrieb Miller Puckette mpuckette@cloud.ucsd.edu:
To Pd-announce:
Pd version 0.55-0 test2 is available from http://msp.ucsd.edu/software.htm or (source only) via github: https://urldefense.com/v3/__https://github.com/pure-data/pure-data__;!!Mih3w...
There was no "test1" - it never made it through the release process. Also, for Mac users, the app will show up as "Pd-0.55-0test2c" but in fact it's just test2.
cheers Miller
Pd-announce mailing list Pd-announce@lists.iem.at https://urldefense.com/v3/__https://lists.puredata.info/listinfo/pd-announce... _______________________________________________ Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://urldefense.com/v3/__https://lists.puredata.info/listinfo/pd-list__;!...
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://urldefense.com/v3/__https://lists.puredata.info/listinfo/pd-list__;!...
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list
I'm having trouble recreating this - I thik it was happening for a while but now I can't seem to get consistent differences (after launching both versions many times). Are you launching it by clicking on the Pd icon, or some other way? (That seems to make a difference).
thanks
M
On 5/15/24 1:25 PM, Benjamin Wesch wrote:
hm, that's really weird. i tried to objectively measure here now with a script that waits for the "connected" output when starting both versions in parallel and i consistently get results quite similar to this:
pd-0.54-1.log: 0.99s pd-0.55-0.log: 2.72s
so my subjective impression was certainly wrong with 5s - but there's still a noticeable difference.
cheers, ben
Am Mi., 15. Mai 2024 um 11:51 Uhr schrieb Miller Puckette mpuckette@cloud.ucsd.edu:
Hmm... I'm not getting that on mine. The startup order is indeed different -
the GUI doesn't start up until after any startup stuff has been done (loading libraries,
etc). But I don't _think_ the overall time spent is different - do let me know if you're
finding otherwise :)
Miller
On 5/15/24 10:56 AM, Benjamin Wesch wrote:
thanks a lot for the test release! what i noticed here (apple m2) in addition to the missing console output: pd now takes ~5s to start. not a big deal obviously, but it's a huge difference compared to 0.54, where the gui is ready after ~1s.
cheers, ben
Am Di., 14. Mai 2024 um 16:26 Uhr schrieb Miller Puckette mpuckette@cloud.ucsd.edu:
To Pd-announce:
Pd version 0.55-0 test2 is available from http://msp.ucsd.edu/software.htm or (source only) via github: https://urldefense.com/v3/__https://github.com/pure-data/pure-data__;!!Mih3w...
There was no "test1" - it never made it through the release process. Also, for Mac users, the app will show up as "Pd-0.55-0test2c" but in fact it's just test2.
cheers Miller
Pd-announce mailing list Pd-announce@lists.iem.at https://urldefense.com/v3/__https://lists.puredata.info/listinfo/pd-announce... _______________________________________________ Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://urldefense.com/v3/__https://lists.puredata.info/listinfo/pd-list__;!...
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://urldefense.com/v3/__https://lists.puredata.info/listinfo/pd-list__;!...
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list
thanks for looking into this! when comparing click-on-app-icon vs. starting "/Applications/Pd-0.55-0test2c.app/Contents/Resources/bin/pd" from the terminal, i also get slightly different results. the latter seems to be a 1 or 2s faster (this was also the way i measured). the former even causes the loading icon to appear, which i never saw in the pd context iirc.
but then again, i'm not sure if this is actually an issue at all! if you don't see it on your side on the same architecture, it may well be caused by some weird configuration on my side.
just to document what i did to check the difference - which i hope is a valid approach ... i have this little script "timer_pd.sh": #!/bin/bash LOG_FILE=$1 START_TIME=$(gdate +%s.%N) while : ; do if grep -q "connected" "$LOG_FILE"; then END_TIME=$(gdate +%s.%N) DURATION=$(echo "$END_TIME - $START_TIME" | bc -l) printf "%s: %.2fs\n" "$LOG_FILE" "$DURATION" break fi sleep 0.1 done
... and then starting these commands to reset the log and start timer and pd in parallel: truncate -s 0 pd-0.55-0.log ./timer_pd.sh pd-0.55-0.log & /Applications/Pd-0.55-0test2c.app/Contents/Resources/bin/pd 2> pd-0.55-0.log &
or obviously the same way: truncate -s 0 pd-0.54-1.log ./timer_pd.sh pd-0.54-1.log & /Applications/Pd-0.54-1.app/Contents/Resources/bin/pd 2> pd-0.54-1.log &
cheers, ben
Am Do., 16. Mai 2024 um 12:20 Uhr schrieb Miller Puckette mpuckette@cloud.ucsd.edu:
I'm having trouble recreating this - I thik it was happening for a while but now I can't seem to get consistent differences (after launching both versions many times). Are you launching it by clicking on the Pd icon, or some other way? (That seems to make a difference).
thanks
M
On 5/15/24 1:25 PM, Benjamin Wesch wrote:
hm, that's really weird. i tried to objectively measure here now with a script that waits for the "connected" output when starting both versions in parallel and i consistently get results quite similar to this:
pd-0.54-1.log: 0.99s pd-0.55-0.log: 2.72s
so my subjective impression was certainly wrong with 5s - but there's still a noticeable difference.
cheers, ben
Am Mi., 15. Mai 2024 um 11:51 Uhr schrieb Miller Puckette mpuckette@cloud.ucsd.edu:
Hmm... I'm not getting that on mine. The startup order is indeed different -
the GUI doesn't start up until after any startup stuff has been done (loading libraries,
etc). But I don't _think_ the overall time spent is different - do let me know if you're
finding otherwise :)
Miller
On 5/15/24 10:56 AM, Benjamin Wesch wrote:
thanks a lot for the test release! what i noticed here (apple m2) in addition to the missing console output: pd now takes ~5s to start. not a big deal obviously, but it's a huge difference compared to 0.54, where the gui is ready after ~1s.
cheers, ben
Am Di., 14. Mai 2024 um 16:26 Uhr schrieb Miller Puckette mpuckette@cloud.ucsd.edu:
To Pd-announce:
Pd version 0.55-0 test2 is available from http://msp.ucsd.edu/software.htm or (source only) via github: https://urldefense.com/v3/__https://github.com/pure-data/pure-data__;!!Mih3w...
There was no "test1" - it never made it through the release process. Also, for Mac users, the app will show up as "Pd-0.55-0test2c" but in fact it's just test2.
cheers Miller
Pd-announce mailing list Pd-announce@lists.iem.at https://urldefense.com/v3/__https://lists.puredata.info/listinfo/pd-announce... _______________________________________________ Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://urldefense.com/v3/__https://lists.puredata.info/listinfo/pd-list__;!...
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://urldefense.com/v3/__https://lists.puredata.info/listinfo/pd-list__;!...
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list
replying to myself here after realizing that i made this quite complicated, sorry. i get similar results when just doing this on macos here:
time /Applications/Pd-0.54-1.app/Contents/Resources/bin/pd -send "pd quit" vs. time /Applications/Pd-0.55-0test2c.app/Contents/Resources/bin/pd -send "pd quit"
cheers, ben
Am Do., 16. Mai 2024 um 13:21 Uhr schrieb Benjamin Wesch benjamin.wesch@gmail.com:
thanks for looking into this! when comparing click-on-app-icon vs. starting "/Applications/Pd-0.55-0test2c.app/Contents/Resources/bin/pd" from the terminal, i also get slightly different results. the latter seems to be a 1 or 2s faster (this was also the way i measured). the former even causes the loading icon to appear, which i never saw in the pd context iirc.
but then again, i'm not sure if this is actually an issue at all! if you don't see it on your side on the same architecture, it may well be caused by some weird configuration on my side.
just to document what i did to check the difference - which i hope is a valid approach ... i have this little script "timer_pd.sh": #!/bin/bash LOG_FILE=$1 START_TIME=$(gdate +%s.%N) while : ; do if grep -q "connected" "$LOG_FILE"; then END_TIME=$(gdate +%s.%N) DURATION=$(echo "$END_TIME - $START_TIME" | bc -l) printf "%s: %.2fs\n" "$LOG_FILE" "$DURATION" break fi sleep 0.1 done
... and then starting these commands to reset the log and start timer and pd in parallel: truncate -s 0 pd-0.55-0.log ./timer_pd.sh pd-0.55-0.log & /Applications/Pd-0.55-0test2c.app/Contents/Resources/bin/pd 2> pd-0.55-0.log &
or obviously the same way: truncate -s 0 pd-0.54-1.log ./timer_pd.sh pd-0.54-1.log & /Applications/Pd-0.54-1.app/Contents/Resources/bin/pd 2> pd-0.54-1.log &
cheers, ben
Am Do., 16. Mai 2024 um 12:20 Uhr schrieb Miller Puckette mpuckette@cloud.ucsd.edu:
I'm having trouble recreating this - I thik it was happening for a while but now I can't seem to get consistent differences (after launching both versions many times). Are you launching it by clicking on the Pd icon, or some other way? (That seems to make a difference).
thanks
M
On 5/15/24 1:25 PM, Benjamin Wesch wrote:
hm, that's really weird. i tried to objectively measure here now with a script that waits for the "connected" output when starting both versions in parallel and i consistently get results quite similar to this:
pd-0.54-1.log: 0.99s pd-0.55-0.log: 2.72s
so my subjective impression was certainly wrong with 5s - but there's still a noticeable difference.
cheers, ben
Am Mi., 15. Mai 2024 um 11:51 Uhr schrieb Miller Puckette mpuckette@cloud.ucsd.edu:
Hmm... I'm not getting that on mine. The startup order is indeed different -
the GUI doesn't start up until after any startup stuff has been done (loading libraries,
etc). But I don't _think_ the overall time spent is different - do let me know if you're
finding otherwise :)
Miller
On 5/15/24 10:56 AM, Benjamin Wesch wrote:
thanks a lot for the test release! what i noticed here (apple m2) in addition to the missing console output: pd now takes ~5s to start. not a big deal obviously, but it's a huge difference compared to 0.54, where the gui is ready after ~1s.
cheers, ben
Am Di., 14. Mai 2024 um 16:26 Uhr schrieb Miller Puckette mpuckette@cloud.ucsd.edu:
To Pd-announce:
Pd version 0.55-0 test2 is available from http://msp.ucsd.edu/software.htm or (source only) via github: https://urldefense.com/v3/__https://github.com/pure-data/pure-data__;!!Mih3w...
There was no "test1" - it never made it through the release process. Also, for Mac users, the app will show up as "Pd-0.55-0test2c" but in fact it's just test2.
cheers Miller
Pd-announce mailing list Pd-announce@lists.iem.at https://urldefense.com/v3/__https://lists.puredata.info/listinfo/pd-announce... _______________________________________________ Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://urldefense.com/v3/__https://lists.puredata.info/listinfo/pd-list__;!...
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://urldefense.com/v3/__https://lists.puredata.info/listinfo/pd-list__;!...
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list
On 5/27/24 09:24, Benjamin Wesch wrote:
replying to myself here after realizing that i made this quite complicated, sorry.
it depends what you actually want to measure.
i get similar results when just doing this on macos here:
time /Applications/Pd-0.54-1.app/Contents/Resources/bin/pd -send "pd quit" vs. time /Applications/Pd-0.55-0test2c.app/Contents/Resources/bin/pd -send "pd quit"
which measures the time for the core to start and quit. this is not the same as the time it takes for Pd to startup with a functional GUI. esp. with Pd-0.55 (where the GUI is started late), the numbers you get might significantly differ from what you are experiencing when you start Pd interactively.
here's what i get on my Debian system using hyperfine (which runs the
application multiple times, to cater for disk-caching and whatnot)
| version | flags | time | |---------|--------|---------------------| | 0.54-1 | | 131.7 ms ± 21.8 ms | | 0.54-1 | -nogui | 4.7 ms ± 2.6 ms | | 618b63 | | 38.4 ms ± 3.6 ms | | 618b63 | -nogui | 4.1 ms ± 2.3 ms |
the 0.54-1 version is the one packaged in Debian.
the "618b63" version is current master (slightly after 0.55-0test3), but
i haven't found any significant differences between master,
0.55-0test3 and 0.55-0test2c)
tests are run on a Debian/amd64 system (i7-13700K).
if you want to measure the time until the GUI is fully available, you probably have to use a GUI plugin that quits Pd from the GUI side.
mfdgrasd IOhannes
On 5/27/24 12:15, IOhannes m zmoelnig wrote:
here's what i get on my Debian system using
hyperfine(which runs the application multiple times, to cater for disk-caching and whatnot)| version | flags | time | |---------|--------|---------------------| | 0.54-1 | | 131.7 ms ± 21.8 ms | | 0.54-1 | -nogui | 4.7 ms ± 2.6 ms | | 618b63 | | 38.4 ms ± 3.6 ms | | 618b63 | -nogui | 4.1 ms ± 2.3 ms |
[...]
if you want to measure the time until the GUI is fully available, you probably have to use a GUI plugin that quits Pd from the GUI side.
so i went ahead and created the following GUI-plugin named "quitter-plugin.tcl":
after idle {pdsend "pd quit"}
now Pd will quit once the GUI has become "idle" (right after startup). which gives me the following stats:
| version | time | |---------|---------------------| | 0.54-1 | 249.9 ms ± 61.8 ms | | 618b63 | 202.3 ms ± 119.3 ms |
so you can see, it is much slower than the naive "-send 'pd quit'" test.
Pd-0.55 is still much faster that 0.54-1, at least for me.
gdmfar IOhannes
thanks a lot for pointing out 'hyperfine' and for providing this somewhat self-defeating plugin.
comparing 0.54-1 to the artifact from the iem pipeline for 618b6325 now, i get the results below. side notes:
- the plugin is present in all cases, but should only affect the last
comparison, i assume
- it's a bit irritating to me that quitting 618b63 is slower via
'-send' flag than through the plugin, but this was obviously caused by outliers (although i got similar results with another 10 runs)
| version | flags | time | |---------|------------------------|-----------------------| | 0.54-1 | -nogui -send "pd quit" | 78.5 ms ± 0.6 ms | | 618b63 | -nogui -send "pd quit" | 77.9 ms ± 1.3 ms | | | | | | 0.54-1 | -send "pd quit" | 327.2 ms ± 5.6 ms | | 618b63 | -send "pd quit" | 2512.9 ms ± 1523.3 ms | | | | | | 0.54-1 | | 396.5 ms ± 4.7 ms | | 618b63 | | 2117.4 ms ± 518.0 ms |
... and a bit of system info:
Model Name: MacBook Air Model Identifier: Mac14,2 Model Number: MLY33D/A Chip: Apple M2 Total Number of Cores: 8 (4 performance and 4 efficiency) Memory: 8 GB
System Version: macOS 14.5 (23F79) Kernel Version: Darwin 23.5.0
cheers, ben
Am Mo., 27. Mai 2024 um 15:08 Uhr schrieb IOhannes m zmoelnig zmoelnig@iem.at:
On 5/27/24 12:15, IOhannes m zmoelnig wrote:
here's what i get on my Debian system using
hyperfine(which runs the application multiple times, to cater for disk-caching and whatnot)| version | flags | time | |---------|--------|---------------------| | 0.54-1 | | 131.7 ms ± 21.8 ms | | 0.54-1 | -nogui | 4.7 ms ± 2.6 ms | | 618b63 | | 38.4 ms ± 3.6 ms | | 618b63 | -nogui | 4.1 ms ± 2.3 ms |
[...]
if you want to measure the time until the GUI is fully available, you probably have to use a GUI plugin that quits Pd from the GUI side.
so i went ahead and created the following GUI-plugin named "quitter-plugin.tcl":
after idle {pdsend "pd quit"}now Pd will quit once the GUI has become "idle" (right after startup). which gives me the following stats:
| version | time | |---------|---------------------| | 0.54-1 | 249.9 ms ± 61.8 ms | | 618b63 | 202.3 ms ± 119.3 ms |
so you can see, it is much slower than the naive "-send 'pd quit'" test.
Pd-0.55 is still much faster that 0.54-1, at least for me.
gdmfar IOhannes _______________________________________________ Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list
On 5/27/24 21:58, Benjamin Wesch wrote:
thanks a lot for pointing out 'hyperfine' and for providing this somewhat self-defeating plugin.
thanks for the additional data.
====== recap ======
i think the most promising paths for investigation are:
- the code
- the build environment
- the signing process
ad signing process: macOS spends some time during startup to check whether the application can be trusted. in theory this should give some performance penalty at the very first startup of an application, but from then on things should be pretty fast (as the results are hashed). in any case, i'm not aware that anything has changed in the way we sign/notarize Pd since 0.54 (apart from Apple having switched to a new signing/notarizing infrastructure,...)
ad build environment: about two months ago i switched the CI builders for macOS from an aging Mac Mini 5,2 (mid 2011) running Big Sur (within a VM on a Debian hypervisor), to a Mac Studio 13,1 M1 (2022) running Monterey (within a VM on a Ventura hypervisor). this sped up builds considerably (from about ~7 minutes to ~1 minute).
(sidenote: this change only affects the build process itself; signing/notarization is still done on the old VMs)
ad code: my gut feeling tells me that there are two promising candidates:
- the changed GUI startup protocol
- something about the audio backends
i *guess* that the changed GUI startup will only affect Pd if it is started via the core (as opposed to starting via the GUI, e.g. by clicking on the app icon). i would have assumed that the changed GUI startup would actually speed up things (at least for our benchmark tests, where we start via the core and try to quit asap)
====== more benchamrks ======
in the meantime i've learned some more about hyperfine (which can do argument variations on its own), and ran some more tests with something like the following:
hyperfine \
--export-csv benchmark.csv \
-N --warmup 3 \
-L pd 0.54-1,0.55-0test3-4-g618b6325 \
-L audio nosound,jack,pa \
-L callback ,no \
-L quit '-send "pd quit" -nogui','-send "pd quit"','-path quitter/' \
"Pd-{pd}.app/Contents/Resources/bin/pd -noprefs -nostdpath -{audio}
-{callback}callback {quit}"
this tests:
- multiple versions of Pd
- different audio backends (w/wo callbacks)
- w/wo gui
- quiting from either GUI or core
(it assumes that there are a couple of Pd-*.app in the current directory, so we only need vary the actual version number. also the quitter-plugin is in the quitter/ directory of the cwd and is *explicitly* activated).
i did one run on our Mac Studio (Ventura) via VNC(!), and another on my Debian desktop PC (obviously with slightly different invocation to cater for the changed path layout and different available backends).
no check has been made to see if the audio backends are actually useable (if Pd tries to open a blocked audio device, this could stall the startup considerably!)
find the results in the attached CSV files (where i've turned the startup options in separate columns. an empty value in the "audio" column means that Pd was started with "-nosound")
my prelaminary findings are:
- the build environment is not to blame for the slow down (the startup
times range between 70ms and almost 2 seconds (on macOS) for both the old and the new binaries (on Debian it's between 2ms and 300ms))
- the most important factors for the entire startup-shutdown cycle are
- whether there's a GUI (slow) or not (fast)
- whether the shutdown is initiated from the core (fast) or from the
GUI (slow)
- as far as the audio backends are concerned:
- whether one uses callbacks or not, does not seem to make much of a
difference 2. using PortAudio adds a considerable overhead. (JACK timings and the PortAudio timings are clearly in two different leagues, separated by a factor of 1.5 up to 5). the separation is most prominent in the tests that do not involve the GUI (suggesting that the GUI overhead hides the overhead from the audio backend)
my numbers differ greatly from ben's. esp. when running Pd-0.54 with GUI it takes much longer (e.g. using the quit-plugin, the minimum turnaround time is ~1.4secs!, rather than 0.4secs as seen by ben) this can probably be explained in part by the different CPUs and OS version. without a GUI i'm seeing comparable numbers (~70ms), but only when using JACK.
the default audio backend for macOS is PortAudio.
gfdasr IOhannes
2. using PortAudio adds a considerable overhead. (JACK timings and the PortAudio timings are clearly in two different leagues, separated by a factor of 1.5 up to 5). the separation is most prominent in the tests that do not involve the GUI (suggesting that the GUI overhead hides the overhead from the audio backend)
So it seems like portaudio is a bit slow to initialize on macOS. FYI, the initial pa_init() call happens in sys_get_audio_devs() via sys_set_audio_settings() in sys_arg_parse() resp. sys_loadpreferences(). I don't think this can be avoided in a reasonable way. It hasn't really been a problem in the past neither.
ad code: my gut feeling tells me that there are two promising candidates:
- the changed GUI startup protocol
- something about the audio backends
I would rule out the latter because AFAICT the audio setting and preference logic hasn't really changed between Pd 0.54 and 0.55.
Christof
Well, I can't reproduce that here - I tried with 0.54-0, 0.55-0test2c. and the latest build from CI:
time ~/Desktop/Pd-0.55-0test2c.app/Contents/Resources/bin/pd -send "pd quit" --> 0.03s user 0.08s system 13% cpu 0.854 total
time ~/Applications/Pd-0.54-1.app/Contents/Resources/bin/pd -send "pd quit" --> 0.03s user 0.09s system 13% cpu 0.882 total
time ~/Desktop/Pd-0.55-0test3-21-gb6476e02.app/Contents/Resources/bin/pd -send "pd quit" --> 0.03s user 0.08s system 12% cpu 0.860 total
I got very little variation in the execution times, except that the first time in a series of running the same app there was an additional 0.06 sec. (The first time of all that I ran each app it was a bit slower yet).
This is on an M2 Mac MINI with 8MB memory. My guess is this would be a faster machine than a Macbook air so these might be consistent with your 0.55 behavior, but not with your (faster) 0.54 results.
Oh - I didn't include any plug-in, just straight Pd distribution.
On 5/27/24 9:58 PM, Benjamin Wesch wrote:
thanks a lot for pointing out 'hyperfine' and for providing this somewhat self-defeating plugin.
comparing 0.54-1 to the artifact from the iem pipeline for 618b6325 now, i get the results below. side notes:
- the plugin is present in all cases, but should only affect the last
comparison, i assume
- it's a bit irritating to me that quitting 618b63 is slower via
'-send' flag than through the plugin, but this was obviously caused by outliers (although i got similar results with another 10 runs)
| version | flags | time | |---------|------------------------|-----------------------| | 0.54-1 | -nogui -send "pd quit" | 78.5 ms ± 0.6 ms | | 618b63 | -nogui -send "pd quit" | 77.9 ms ± 1.3 ms | | | | | | 0.54-1 | -send "pd quit" | 327.2 ms ± 5.6 ms | | 618b63 | -send "pd quit" | 2512.9 ms ± 1523.3 ms | | | | | | 0.54-1 | | 396.5 ms ± 4.7 ms | | 618b63 | | 2117.4 ms ± 518.0 ms |
... and a bit of system info:
Model Name: MacBook Air Model Identifier: Mac14,2 Model Number: MLY33D/A Chip: Apple M2 Total Number of Cores: 8 (4 performance and 4 efficiency) Memory: 8 GB
System Version: macOS 14.5 (23F79) Kernel Version: Darwin 23.5.0
cheers, ben
Am Mo., 27. Mai 2024 um 15:08 Uhr schrieb IOhannes m zmoelnig zmoelnig@iem.at:
On 5/27/24 12:15, IOhannes m zmoelnig wrote:
here's what i get on my Debian system using
hyperfine(which runs the application multiple times, to cater for disk-caching and whatnot)| version | flags | time | |---------|--------|---------------------| | 0.54-1 | | 131.7 ms ± 21.8 ms | | 0.54-1 | -nogui | 4.7 ms ± 2.6 ms | | 618b63 | | 38.4 ms ± 3.6 ms | | 618b63 | -nogui | 4.1 ms ± 2.3 ms |
[...]
if you want to measure the time until the GUI is fully available, you probably have to use a GUI plugin that quits Pd from the GUI side.
so i went ahead and created the following GUI-plugin named "quitter-plugin.tcl":
after idle {pdsend "pd quit"}now Pd will quit once the GUI has become "idle" (right after startup). which gives me the following stats:
| version | time | |---------|---------------------| | 0.54-1 | 249.9 ms ± 61.8 ms | | 618b63 | 202.3 ms ± 119.3 ms |
so you can see, it is much slower than the naive "-send 'pd quit'" test.
Pd-0.55 is still much faster that 0.54-1, at least for me.
gdmfar IOhannes _______________________________________________ Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://urldefense.com/v3/__https://lists.puredata.info/listinfo/pd-list__;!...
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://urldefense.com/v3/__https://lists.puredata.info/listinfo/pd-list__;!...
thanks for your tests and response! these numbers look rather encouraging and at the same time let me assume that something might be weird with my setup here.
after learning about 'hyperfine' from iohannes in the other thread, i ran tests once more now with the new test release and get slightly better results (~30% faster).
since this startup behaviour doesn't seem to affect pd's functionality anyway, it probably can't be considered a blocker ... so my suggestion would be to move that issue from the list to github. i'd create an issue noting the current status and measurements - if there are no objections concerning this procedure?
i collected/summarized my current measurements compared with previous results and the ones from iohannes' csvs here for now (the first table header row is a bit confusing in this generated pivot table): https://docs.google.com/spreadsheets/d/e/2PACX-1vS8_dgJXwY62YMJtwQoZ5o-mSvzq...
cheers, ben
On Fri, May 31, 2024 at 2:40 PM Miller Puckette mpuckette@cloud.ucsd.edu wrote:
Well, I can't reproduce that here - I tried with 0.54-0, 0.55-0test2c. and the latest build from CI:
time ~/Desktop/Pd-0.55-0test2c.app/Contents/Resources/bin/pd -send "pd quit" --> 0.03s user 0.08s system 13% cpu 0.854 total
time ~/Applications/Pd-0.54-1.app/Contents/Resources/bin/pd -send "pd quit" --> 0.03s user 0.09s system 13% cpu 0.882 total
time ~/Desktop/Pd-0.55-0test3-21-gb6476e02.app/Contents/Resources/bin/pd -send "pd quit" --> 0.03s user 0.08s system 12% cpu 0.860 total
I got very little variation in the execution times, except that the first time in a series of running the same app there was an additional 0.06 sec. (The first time of all that I ran each app it was a bit slower yet).
This is on an M2 Mac MINI with 8MB memory. My guess is this would be a faster machine than a Macbook air so these might be consistent with your 0.55 behavior, but not with your (faster) 0.54 results.
Oh - I didn't include any plug-in, just straight Pd distribution.
On 5/27/24 9:58 PM, Benjamin Wesch wrote:
thanks a lot for pointing out 'hyperfine' and for providing this somewhat self-defeating plugin.
comparing 0.54-1 to the artifact from the iem pipeline for 618b6325 now, i get the results below. side notes:
- the plugin is present in all cases, but should only affect the last
comparison, i assume
- it's a bit irritating to me that quitting 618b63 is slower via
'-send' flag than through the plugin, but this was obviously caused by outliers (although i got similar results with another 10 runs)
| version | flags | time | |---------|------------------------|-----------------------| | 0.54-1 | -nogui -send "pd quit" | 78.5 ms ± 0.6 ms | | 618b63 | -nogui -send "pd quit" | 77.9 ms ± 1.3 ms | | | | | | 0.54-1 | -send "pd quit" | 327.2 ms ± 5.6 ms | | 618b63 | -send "pd quit" | 2512.9 ms ± 1523.3 ms | | | | | | 0.54-1 | | 396.5 ms ± 4.7 ms | | 618b63 | | 2117.4 ms ± 518.0 ms |
... and a bit of system info:
Model Name: MacBook Air Model Identifier: Mac14,2 Model Number: MLY33D/A Chip: Apple M2 Total Number of Cores: 8 (4 performance and 4 efficiency) Memory: 8 GB
System Version: macOS 14.5 (23F79) Kernel Version: Darwin 23.5.0
cheers, ben
Am Mo., 27. Mai 2024 um 15:08 Uhr schrieb IOhannes m zmoelnig zmoelnig@iem.at:
On 5/27/24 12:15, IOhannes m zmoelnig wrote:
here's what i get on my Debian system using
hyperfine(which runs the application multiple times, to cater for disk-caching and whatnot)| version | flags | time | |---------|--------|---------------------| | 0.54-1 | | 131.7 ms ± 21.8 ms | | 0.54-1 | -nogui | 4.7 ms ± 2.6 ms | | 618b63 | | 38.4 ms ± 3.6 ms | | 618b63 | -nogui | 4.1 ms ± 2.3 ms |
[...]
if you want to measure the time until the GUI is fully available, you probably have to use a GUI plugin that quits Pd from the GUI side.
so i went ahead and created the following GUI-plugin named "quitter-plugin.tcl":
after idle {pdsend "pd quit"}now Pd will quit once the GUI has become "idle" (right after startup). which gives me the following stats:
| version | time | |---------|---------------------| | 0.54-1 | 249.9 ms ± 61.8 ms | | 618b63 | 202.3 ms ± 119.3 ms |
so you can see, it is much slower than the naive "-send 'pd quit'" test.
Pd-0.55 is still much faster that 0.54-1, at least for me.
gdmfar IOhannes _______________________________________________ Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://urldefense.com/v3/__https://lists.puredata.info/listinfo/pd-list__;!...
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://urldefense.com/v3/__https://lists.puredata.info/listinfo/pd-list__;!...
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list
-
Alexandre Torres Porres -
Benjamin Wesch -
Christof Ressi -
Edwin van der Heide -
IOhannes m zmoelnig -
IOhannes m zmölnig
-
Miller Puckette -
Peter P. -
Roman Haefeli