I had a dream

[tester]

I think I'm moving into the next state of FS programing.

I thought that I finished one of my largest projects. Everything finished, everything polished. So I even stopped thinking about doing anything else inside. I just let go.

But today, during a nap, I had a dream. It was dream about programming, but not a kind of dreams that you have when you work too much on something. Anyway. I dreamed about how to optimize the simple 1:1 preset system I use since the beginning (c.a. 2000 automated parameters). I was accustomed that it loads programs/banks very slow (about 40 seconds, but no matter whether single program or all programs), so I lived with that with no pressure. The dream was very complex, multiple steps, and very precise. I woke up and my first reaction was - "hmmm... really??".

I started to check that idea and it appeared, that the preset system after that change can indeed speed up from 40 seconds to maybe 3 secs...

Anyway the whole idea is, to keep preset parameter prims for automation only, and send all user values to preset string primitive, by making (comma) separated long strings (no mess like with arrays). Then - just load these strings, split and send values via wireless. Because I have almost all values sent to similar modules for other needs, this concept should be easy to do.

What a nice dream.
(puff... - again back to work)

So - I wish you sweet dreams tonight.

p.s.: are there any prims to wire FS directly to the brain, to do such stuff during dreaming?

[tektoog]

Not only I dreamt of it, but I've made it... last year...
there...
I called it (the dream): "Multi type data array 2 VST string"
yeah... quite a useful trick...

[tester]

Checked. Somewhat similar to what I do, but my way seems to be a little bit simpler. I connect output parameters to array builders (any data -> sample and hold with trigger on save -> format if needed -> string to array builder), convert array to string and send to preset string. Then from preset string - string is reconverted into array, and (instead of creating a bunch of various wireless links) - modules receive "string get at" with proper array link and ID. Thus - preset value prim is used only in cases where automation is needed (to pass through data without saving), and everywhere else - is removed.

Right now I'm rebuilding it to that.

[Tronic]

I have delegated the management of my presets to an external DLL,
with which I also created a system of encripting and decrypts as the protection system.

all values ​​are sent to a single ruby module, which sends them to my DLL, it encrypts the data and saves them.

I also provides the ability to handle large data quickly, all at the end is executed in C + +.

when a preset is recalled, it decrypts the file and sends the data to ruby module,
it sends the data in the same way as the preset manager makes now, via a wireless.

that's why I discuss much on having the ruby or something that communicates with the outside.

Edit:
to forgot the whole thing is written in C++ but directly to the ruby C, so it does not need the Win32API.
and you can call it using only a:
require 'preset_manager.so'

[tester]

The thing I'm rebuilding right now - should allow loading presets in "no-time". I noticed, that while saving 2000 values goes quickly (both, single programs and large program sets), loading them - takes enormous amount of time (but the same amount for both - single programs and program sets). This led me to the conclusion, that there is some other processing going on inside preset manager/receivers. So having only 100-200 longer segmented entries should be fine. The only thing not glued that way - will be wavetable paths.

Another advantage of this string-based approach - will be quick exchange of data between cloned units inside one application - just redirecting strings (thus - it should give me the answer how to make some things that did not worked old way).

Year ago, I did not approached this preset related part, because in general - I didn't knew how much primitives and tiny processings SM/FS can handle without outputing artifacts or crashes. But right now, after I made extensive cross-exchange section - I see that this is very powerful and stable engine in terms of that - you just need to pay attentnion to stream related things (sometimes you need to disconnect streamy parts before you load some data).

As for encrypting - I see no reason to use it. I prefer human friendly readable values, so that you can handle them with external editors when needed (reporting tools for comparative analysis).

[Tronic]

the encrypting part is used as protection system, so when you save the preset and not mach whit your algo, it not work if the license is not valid.

[tester]

I understand, I would just not go so deep. Using multiple computers (even at home) - I don't like over-protected software (or over-protected data exported from it). It's a paradox - even if I buy something - sometimes I just prefer unofficial (or portable) version, because it's easier to use (and I don't need to contact anyone, to recover my data back if something crashes deeper; nor I don't need internet connection). That's why I like philosophies like in Reaper or Flowstone - they do understand modern reality.

[Tronic]

your speeches are based on not to take any kind of business from what you do,
not everyone plays only.

[tester]

You don't like that people don't follow your assumptions about artificial ethics? It's your issue. I'm not going to please you. Doing business, I do have a legal licence for what I use. But it does not means, that I must use official version of the same thing, if the official version causes headaches and more trouble than it's supposed to. We don't live in a totalitarian world. So personally - why should I support paranoia? But analyzing in second direction - what can I do (how can I organize a project), to avoid being paranoiac on what people do with my own work? How to do something and feel comfortable with the results? What strategy to take to avoid unnecessary mess? I guess, such thoughts were a part of origins of open sourcing versus blind patenting, and creative commons versus "don't even dare to think about it". I'm not developing new Windows or OSX, to worry about certain things.

A simple thought on protections. What happens when you die (or have an accident) and your website is down?

And one more thing on business. If you are focused on products (project = product) and I'm focused on services made with products (project = tool for creating a product), then there is different prioritization in the game.

[Tronic]

A simple thought on protections. What happens when you die (or have an accident) and your website is down?

happen what may happen to you too. obvious.

but not always, everything stops, life goes the same even without us,
wanting to give some examples, the creator of http://www.discodsp.com, unfortunately is no longer with us, but its plugins are still sold and developed, just because there was the project of a business, and when you want to do a business you also have people who are interested in it, so you're not alone, also I'm lucky, even if I die, I have a business partner.

sorry, your vision is based in a small context, maybe mine is more commercial than ethical.