@Josefo18 There was a discussion long ago..... https://forum.pdpatchrepo.info/topic/6003/rap-realtime-analysis-project where [sigmund~] and [bonk~] were mentioned.
@katjav investigated SNAC on her site...... https://www.katjaas.nl/helmholtz/helmholtz.html
... and at the bottom of that page she posted [helmholz~] which might be what you are looking for.
David.

@ddw_music https://mitp-content-server.mit.edu/books/content/sectbyfn/books_pres_0/8375/designing_sound.zip/practical15.html has a little more description and states for [rain_on_water] .... Rain - rain on water, alternative droplet model.
But fortunately, [raindrops] is included in the tarball of all files here....... http://aspress.co.uk/sd/ so it is an omission mistake on the page you linked to....
David.

@gentleclockdivider It is replaced when a symbol...... Bug_select_or_not.pd but once created for symbols cannot become [sel float]
[select] and [route] become different objects when they are created for floats or symbols.
David.

@elden Help... Find Externals..
and search for list-abs.
As it is a library of abstractions it is not platform dependent.
David.

@elden list-abs.... here it is... list-random.zip although it is worth having the whole library.
The library consists almost totally of abstractions.
David.

@tildebrow Then you should get the else library and see.......\else\Live-Electronics-Tutorial\Part.12-Advanced.Pd\39-Data.Structures .... as that is an up-to-date tutorial.
All of the examples need [declare -path else] and that is missing from some. But I installed it long ago and that has probably been corrected.
David.

@tildebrow Unsure whether you are looking to put "words" ..... [drawsymbol].... not editable..... [drawtext].... working.
Or graphical symbols.
If the latter then have a look at ..... https://forum.pdpatchrepo.info/topic/10349/drawing-objects-simplified/2
Particulary object-draw-help.pd ...... 1478428764693-object-draw-updated.zip

Or the else library contains a tutorial for data structures.
David.

@elden You will have seen....... https://www.uni-weimar.de/kunst-und-gestaltung/wiki/images/Dynamic_Time_Warping_for_Pure_Data.pdf

The paper concludes with an application for body gesture recognition, but it is working with arrays so should be equally applicable to audio.
The code still exists as C++ to be compiled for a Pd external, but of course we have since moved to 64bit.
https://github.com/lemire/lbimproved

And there is this....... which is older, by the same same author....
https://github.com/felixr/dtw_lbimproved

David.

@oid I can correct the object numbers in the test patch for vanilla and then it loads correctly, so I don't think it is the same problem. [matrix~] also creates correctly with the [loadbang] in vanilla, but with object numbers wrong...... and I can correct that.
I will try with [initbang] instead when I get time... thank you for the idea..... but I am not convinced for now that object numbers are wrong because of [loadbang] and the ordering will be changed again when I try.... so a lot of work.
If [loadbang] in later vanilla is sorted after all other objects regardless of its creation order then that would explain the change of object numbers..... but that would be in the "breaks old patches" category... and unusual especially for a foundational object. But only for dynamic patching, and we were warned that it is unsupported.

So I will first check whether moving [loadbang] to the end of the object list solves the problem.... an easier fix.

You are correct that this patch worked correctly in older vanilla...... it was built to create a replacement for [matrix~] during the 32/64 changeover while the external was unavailable, and worked at the time.
David.

@FFW You can dynamically re-create the connections, but of course you need to know the object numbers of the objects in the main patch that you want to connect to the inlet/outlet.
I have had to do this building patchable input /output cords in a matrix..... matrix~.zip
Everything inside [matrix~] is created according to its arguments and then the test patch connects to the abstraction.
The test (matrix-test~.pd) works in extended.... but strangely the object numbers have changed when opened in vanilla. Inside [matrix~] everything creates correctly in both versions.
It should create this....... which demonstrates the principle...

It is best to recreate your main patch with the objects you will want to connect being the first objects created... and then copy back the rest of the patch and connect those objects.
Then if you modify the patch later you will not change those object numbers and break the dynamic cord creation.
David.