Architect beta for macOS, Windows, and Linux. 0.10.5 now available

[lalo]

I'd add an option to [slider] to set wich kind of data generates. Casting with [to integer] is possible but i think that in case of UI objects would be better to have integer/float options.

Yeah, why didn't I think of that! Good call, thanks.

with such a complex work of art it's at least normal that something get missed.
I got another strange thing...
At first i casted the slider to int scalar like in array #1 jpg

And i got the same array build error as before

Then i casted to vector int after the tuple like in array #2 jpg


But as you see even if i move the slider always the first value is on outout

Then i got back to the first situation casting after te slider and everything worked as expected.
like in array #3 jpg

...mhhh

[colin@loomer]

There's no built-in LFO module, so you'll need to macro it yourself for now. I'll add some LFO macros when I start assembling the user library.

[cturner]

When I save an Architect patch in the plugin, I can then simply press CTRL-S to save it again over the previous version - great. When I close and re-open the Bitwig project and open Architect, it still holds the name of the patch in the preset field on top of Architect, but when I change something and hit CTRL-S, I get the save-as dialog instead of it saving over the existing one.

There's a related issue that may be unique to Reaper and the Architect plug. It seems like there's a conflict between Reaper's preset system and that of Architect. I can save a preset, but when I launch Reaper again, the preset values from what's shown in Reaper's preset dropdown are used. I believe I can restore the correct preset using Architect's preset system, but on next launch of Reaper, I'm back to what Reaper thinks is the correct preset. To fix this, I have to use the Reaper dropdown setting "Restore Factory Default" and select my preset in Architect, but this only works until the next save.

At least I think this is what's happening; it's very confusing. I had expected to do more research into the condition, but Tom had posted the above. I'll post again later with a better presentation.

[nordickvr]

There's no built-in LFO module, so you'll need to macro it yourself for now. I'll add some LFO macros when I start assembling the user library.

Many thanks!

[Distorted Horizon]

"early-bird" price, 25% of the standard price of £140 + VAT.

I believe you mean OFF rather than OF

[colin@loomer]

@lalo

Could you send me your preset so I can have a bit of a poke around and see where it's going wrong?

I also see in one of your pics you are using a `to integer (vector)`, and the other a `to integer (scalar)`. Also vector processing only works with arrays, so if you convert a tuple to an integer, it'll be 0 (if empty) or 1 (if not.)

I think I probably need to improve the "incorrect type" errors a little so at least you can see what it's getting as well as what it's expecting.

[colin@loomer]

I believe you mean OFF rather than OF

Dropping F's all over the place. Thanks!

[AZZIN]

Ok, I got my first "chaotic/recursive" equation in Architect done.

Greetings-

Do you know about the André Sier/Richard Dudas abstractions for Max?

http://s373.net/code/A-Chaos-Lib/A-Chaos.html

Max is a pretty different language from Architect's, and Siers' work is in C language, but you might find something of value there. (Dudas did the "abstractions" in Max for MIDI control.)

I believe these still work in Max 7 (tried a bit just a second ago), so perhaps you can get these going in a Max 8 demo.

HTH, Charles

Hello Charles, no I don't know A-Chaos-Lib (I don't use Max but I use a lot Pure Data). Actually the link you provided is pretty interesting, thanks for sharing it. As far as I can say LUA is much more effective for this kind of manipulations. I've already added two algorithms in LUA (Lorenz equations and Dejong attractors), I'm going to share them soon. But...

...for Architect v2.0 we need OpenGL graphical drawing capabilities Colin?

[colin@loomer]

It happens every time, yes. Haven't yet tried it in standalone. Will report back.

Please do. And after, when you get a moment, can I PM some things to try? I'd like to get this fixed for the next beta, if possible.

[colin@loomer]

...for Architect v2.0 we need OpenGL graphical drawing capabilities Colin?

One of the unfinished modules that I pulled before beta was a generic canvas for this kind of thing. It'll be back at some point!

[cturner]

I updated the Architect "tips" file: http://vze26m98.net/loomer/Loomer%20Arc ... 20Tips.txt

[colin@loomer]

How would everyone feel if I created a specific thread for "Could someone make me a * macro/script" type requests? Sounds like a decent idea? As this thread is as chaotic in topic as one of AZZIN's algorithms, it may given more visibility to people who just want to either have a specific macro, or know how one would be done?

[lalo]

@lalo

Could you send me your preset so I can have a bit of a poke around and see where it's going wrong?

I also see in one of your pics you are using a `to integer (vector)`, and the other a `to integer (scalar)`. Also vector processing only works with arrays, so if you convert a tuple to an integer, it'll be 0 (if empty) or 1 (if not.)

I think I probably need to improve the "incorrect type" errors a little so at least you can see what it's getting as well as what it's expecting.

So basically a tuple is not considered a vector. I guess beacause it has no indexes. It's just a collection of data.

[colin@loomer]

Yes. The key difference is that arrays are homogeneous (each element is of the same type); whilst tuples are heterogenous (elements can be different types), and the position of an element actually means something. So if you look at the tuple from say a noteon, you will see ("noteon", key, velocity, channel).

[lalo]

SysEx Value Send.zip

@lalo

Could you send me your preset so I can have a bit of a poke around and see where it's going wrong?

here it is
..and btw i still have the array pack errors...but not always...some init mess probably