Here is an updated workshop file, with my examples. I can't put some examples of other persons, but they're available online.

João

http://www.pdpatchrepo.info/hurleur/data-s.workshop.jmmmp.zip

Cool ! Thanks João.

João, thanks for all your explanations. I admire your perseverance, to make useful patches with these convoluted methods. If I were not so deeply involved with Pd already, I would just forget about data structures. But I still have a feeling that it is worth the effort. Weird idea, that Pure Data was conceived for exactly this aspect, which hardly anyone uses.

Katja

I don't make that many so much convoluted patches
I guess both [bezier] and [swatch] were a kind of inspiration, and concentrated problem-solving (each was done in around 1 day). Actually it's not that different from other people who use a clock of some kind to update a gui - since [struct] outputs the operation+pointer, that serves as clock (which is an advantage to "normal" Pd).

Yes, reading Miller's original paper on Pd, this was created (mainly?) to add data structures to Max. All these years gone by, and not so much has been done - new features come very slowly, one or two in each release.
I use data structures some times when I need an interaction between a dynamic graphic + user actions. But still, they're quite basic.

I repeat again, Pd would gain a lot if someone would port ftm to Pd (or continue IOhannes' work on it). This would give us a very much more potent and elegant data structures framework.

Until then, we'll have to make due with what we have. It works nevertheless, but it is quite slow to program.

João

@katjav said:

But now the surprise: on my computer (OSX) the fat point is invisible!

Hmm that are bad news. I thought pd would be the same on each OS.. at least vanilla pd..
Could you try if instead of [drawpolygon 900 8 a b a b]
something like [drawpolygon 900 8 a b a(0:10000)(0:10001) b] works??
..maybe one can trick pd ..once again..

Jmmmp, thanks again for the patches. ...how did you find something like
[send-window vis 0/1(--[pointer] ??!
If one would ask me I'd say we should add those files to the forums tutorial-section..:)
..and yeah all your files seem to come with the new pd-versions..

Hi,

since I don't know how to read C, I can only wait until a feature is documented, or mentioned in the pd-list. Someone else said something about [send-window vis 0/1(--[pointer] in the list, and I took note.

I would prefer to add my files in the puredata.info website. But as I said before, I'm still waiting for another chance of presenting them, and do some more changes in the content (or maybe even in the presentation and examples, if I have the time).

João

Yop,

@jmmmp said:

• nau, that's a complicated circle you drew. Maybe you can adapt the circle I did in [swatch]?

You were right, João. But I was looking forward.

Here is a lissajous tryout, visualizing two sine signals of different frequencies.
The result is not trying to artificially compensate for the fact that it doesn't act as an oscilloscope would.

I'll continue to play with lissajous, as i'd like to use it to make a fun gui element that renders synchro quality between two speeds, a bit in the fashion of turntables 'strobes' :

Nau

http://www.pdpatchrepo.info/hurleur/lissajous_tryout.pd

Hi nau,

that's a very nice concept. Only one problem, data structures is very unefficient for it, you'll need lots of cpu just for the display.
Do you know cyclone/Scope~? That's the same object as max/msp, and I think it works well. Unless you need all the infos in the arrays for something else, then it's a good point to use data structures.

Yop,

thanks for reminding me of Scope~, João.
My version costs a little more than twice as [Scope~].
Just having fun with data structures, anyway.

Nau