@totoetlititi I don't see a flatpak app for 0.55.2 on the Pd 0.55.2 download page.
You can find "apt-get" versions for Linux/debian etc. here......
https://deb.debian.org/debian/pool/main/p/puredata/
David.

@ddw_music It is a little more complicated.
If you are writing a paper then you should look at @jameslo's conclusions re a previous discussion on the building of the dsp graph and the use of subpatches to control the ordering and ensure repeatability..... https://forum.pdpatchrepo.info/topic/13512/s-r-throw-catch-latency-and-object-creation-order/21
There is a lot of detailed info in the thread.... maybe nothing new for you but...?
David.

@soundproofskin Pd specific shortcut key bindings are in pd_bindings.tcl in the Pd/tcl folder.
You can change them (carefully) to have what you need, or better still write a pd-plugin file.
I don't know (as yet) how to do that correctly.

OS standard bindings seem to be in tk.tcl in the same folder and are set according to the the window system detected.

I am running windows on a Mac so a Mac keyboard. Key identifiers seem to be wrong for some modifying keys.
For me Fn+left and Fn+right jump to the beginning and end of the selected text. But I find no mention of Fn as a modifier anywhere in the .tcl files..... so...?
And the up and down arrows do the same.... jump to the beginning and end.

Sorry... not much help..
David.

@oid A recent change (since the doc was written).
https://puredata.info/docs/developer/PdFileFormat

I don't know how long ago but it doesn't work in Pd extended.
Interesting, and it persists through a save..... so is now part of the file format.
It also works for objects and even some gui's even though gui's have fields already for setting width.
David.

@jameslo Not sure..... if you are banging the slider and moving it at the same time then it is hard to interpret.
The fader can be moved by hand fast enough to cause trouble as the OP found.
Yes, certainly, if moving the slider faster than the speed that would create an output every 0.125 msecs it might not be reliable, but your test is mixing 2 effects so....?
I found that all data points were produced by my test at higher speeds than [metro 0.125] but the console print was failing.
Caveat.... was [tabwrite] still writing to the array?
I will try again at higher rates but if [tabwrite] fails to update the [array] then I will never be certain of the limit without console output.
Except that if a whole 400 point table is written it should indicate success.... unless the screen fails to update because Pd has effectively hung of course.

No idea whether cpu speed has relevance at todays speeds..... probably not.
Mine ... Processor Intel(R) Core(TM) i7-3615QM CPU @ 2.30GHz, 2301 Mhz, 4 Core(s), 8 Logical Processor(s)

David.

@jameslo Comms starts to fail with the console below about [metro 0.125] but all values are still produced by the slider........

David.

@jameslo My test........ as I remember [timer] reporting "The timer object measures elapsed logical time. Logical time moves forward as if all computation were instantaneous and as if all "delay" and "metro" objects were exact."
try.zip
All values are sent by the slider banged at a 1msec rate, and are fast in 0.55.2
Keep the slider sub-patch open or [metro] cannot be stopped.

Much earlier versions of Pd were not so fast, with some values repeated or not sent.
Pd better? Wish86 better than Wish85?.

It looks as though a [metro] fine tuned to the actual reported time could maybe achieve a stable [realtime] value... I will try when I get a moment..
It gives 3 values per 3msecs with jitter.
DSP off.

With DSP on the timings are very stable though and suggest 10 sub-divisions per block.... but none of the maths I have tried so far correlates to samplerate and/or screen refresh.
10 values per 10msecs, so not per block either.

Some thoughts I wrote before actually testing.... which seemed correct at the time of writing but are not confirmed by the control rate test..... so a large pinch of salt required....

"As said above, the intermediate values cannot be recovered, but can be approximated by an audio rate object.
As @jameslo says [lop~] will not arrive at the final value, and introduces a delay (phase shift).... https://forum.pdpatchrepo.info/topic/11850/explanation-for-lop-object-in-this-patch-requested

[line~] and [vline~] will approximate in "realtime" at the samplerate. [vline~] better than [line~] as it works across boundaries and so values between ticks can be obtained.

And anyway, the audio samplerate is 64 values every 1.45msecs.
The control rate samplerate is 1 value every 1.45msecs.
If you drag your number box fast enough then you will exceed even the audio rate possibilities for capture, in much the same way as the values generated by your [osc~] at a high enough frequency.
It is only once back in the analog domain via [dac~] that you can have a smooth waveform.

David.

@PLDS "\" is an escape character....... https://en.wikipedia.org/wiki/Escape_character
Older versions of Pd didn't display them and they could not be typed.
(This forum behaves in a similar manner..... I had to escape the backslash above to make it display)

It says.... what follows... $1 in the message shown in the screenshot below... is a literal string rather than a variable or a special character.

But since some version of Pd (about 0.51) they can be typed, so you can do this......

Which is slightly more elegant... but also much easier......
David.

@cfry You can chain [list split 2]'s.... middle outlet of each to the inlet of the next + left outlets in parallel ..... so as to have longer name/value lists on each line.

And also If you use [route symbol] with just the one argument ( multiple [route symbol] objects in parallel ) it can be updated through the right inlet if necessary.
David.

@cfry Might help...... no-comma.zip
Will remove commas from e.g.
A, 3;
B, 6;
etc.
.... but depends on the lines being always of 2 elements.
[text define] seems to display the data very differently in 0.55 and 0.50.... but results are the same...
David.