Page 1 of 3

Going from 3.08.1 to 3.06

PostPosted: Mon Feb 12, 2018 8:51 am
by Spogg
Hello!

Based on advice from others I installed 3.06 but I got a crash every time when I tried to drag in a Wave Player module. I returned to 3.08 and it was fine with this module.
Then I renamed the Flowstone program folder (not the App data one) and re-installed 3.06. The rogue Wave Player was gone (a “good” one was still there) and 3.06 seems to work fine now. I must assume that this module was left over from 3.08 and the installer for 3.06 didn’t remove it.

As far as I can tell all my toolbox is still intact and as it was before (this is in the Appdata folder).

The edit environment now seems more responsive somehow, on my large schematic, plus I gather from tulamide that Ruby runs “better” in 3.06, so I’m happy so far.

If I do get any issues that I think may be version-related I'll post in this topic.

Cheers

Spogg

Re: Going from 3.08.1 to 3.06

PostPosted: Sun Feb 18, 2018 11:06 am
by kortezzzz
Thanks Spogg,

My partner did it as well, and yes, 3.0.6 version is far more stable with my big project, even though one of the most important ruby blocks that tulamide made for me doesn't work with 3.0.6 :?

Re: Going from 3.08.1 to 3.06

PostPosted: Sun Feb 18, 2018 12:07 pm
by tulamide
kortezzzz wrote:... even though one of the most important ruby blocks that tulamide made for me doesn't work with 3.0.6 :?

That is literally impossible, since I work with 3.0.6

So, there's only two possibilities: You messed things up, or I messed things up. Please pm me with a link to a schematic, that contains my non-working Ruby code, demonstrating the issue. I will have a look at it.

Re: Going from 3.08.1 to 3.06

PostPosted: Mon Feb 19, 2018 10:35 am
by kortezzzz
Thanks for the answer, tulamide :)

The schemating is attached to the post. After re-loading it, the Ruby module won't work for me.

The basic idea behind the the ruby module is to output "1" every time a matched midi key is pressed. For instance, if key C is pressed, the Ruby module will output "1" from it's first output. If key #C is pressed, it will output "1" from it's second output and so on.

Re: Going from 3.08.1 to 3.06

PostPosted: Mon Feb 19, 2018 11:56 am
by tulamide
kortezzzz wrote:Thanks for the answer, tulamide :)

The schemating is attached to the post. After re-loading it, the Ruby module won't work for me.

The basic idea behind the the ruby module is to output "1" every time a matched midi key is pressed. For instance, if key C is pressed, the Ruby module will output "1" from it's first output. If key #C is pressed, it will output "1" from it's second output and so on.

I see the issue. Needs a workaround. When you open the schematic, for whatever reason, two midi triggers are sent with a note off event. That messes up the array counter that I implemented assuming that we won't receive midi when nothing is played :roll:

I'll upload as soon as the workaround is ready.

Re: Going from 3.08.1 to 3.06

PostPosted: Mon Feb 19, 2018 3:22 pm
by kortezzzz
THANKS :D

Just a one more question: how can the midi triggers mess with the array? or it's something related to Ruby midi stuff only? It's never happened to me before.

Re: Going from 3.08.1 to 3.06

PostPosted: Tue Feb 20, 2018 5:07 am
by tulamide
kortezzzz wrote:THANKS :D

Just a one more question: how can the midi triggers mess with the array? or it's something related to Ruby midi stuff only? It's never happened to me before.

No, it is not related to Ruby Midi. It is an error of the incoming midi stream. It sends midi note offs without prior midi note ons. That's a violation of the midi protocol => a bug.

I could track down the issue to that keyboard module. It produces those note offs. The issue is buried somewhere in the "GUI notes" structure. I tried my best to fix it, but wasn't successful. You should contact the author of that module. He makes a couple of odd things, like sending a midi note off for every incoming midi message, whose second data byte is 0 (regardless of status byte) :shock: .

Anyway, maybe Spogg should split our parts from his thread into a new one, as it has nothing to do with going from 3.0.8.1 to 3.0.6

Re: Going from 3.08.1 to 3.06

PostPosted: Tue Feb 20, 2018 10:41 am
by kortezzzz
tulamide wrote:The issue is buried somewhere in the "GUI notes" structure. I tried my best to fix it, but wasn't successful. You should contact the author of that module. He makes a couple of odd things, like sending a midi note off for every incoming midi message, whose second data byte is 0 (regardless of status byte) :shock: .
[/quote]

Thanks for your efforts, tula. I went back to my old trusted green solution, although its a lot more clumsy, but seem like I can never get everything all the time. Only some of the things, some of time, and I'm still happy with it, as long as things keep working somehow :)

Re: Going from 3.08.1 to 3.06

PostPosted: Tue Feb 20, 2018 11:02 am
by tulamide
kortezzzz wrote:
tulamide wrote:The issue is buried somewhere in the "GUI notes" structure. I tried my best to fix it, but wasn't successful. You should contact the author of that module. He makes a couple of odd things, like sending a midi note off for every incoming midi message, whose second data byte is 0 (regardless of status byte) :shock: .


Thanks for your efforts, tula. I went back to my old trusted green solution, although its a lot more clumsy, but seem like I can never get everything all the time. Only some of the things, some of time, and I'm still happy with it, as long as things keep working somehow :)

That's totally fine.

But, while my module works flawlessly, the keyboard module produces bugs and propagates them through your schematic. Yet, you prefer to keep that buggy module and instead let go off a perfectly fine one. That's really strange. It's not that there wouldn't be lots of keyboard modules (for example, Nubeat made a Ruby based one). So why not using another one? Or trying to solve the issue in that keyboard module?

My module works fine as long as the provided midi data sticks to the midi protocol.

Re: Going from 3.08.1 to 3.06

PostPosted: Tue Feb 20, 2018 12:43 pm
by Spogg
tulamide wrote:Anyway, maybe Spogg should split our parts from his thread into a new one, as it has nothing to do with going from 3.0.8.1 to 3.0.6


Yeah I thought about this but it would be tricky to split it without losing some context.

I did it successfully once before but the last time I tried the whole lot got lost. This board can be a bit unpredictable at times. Or maybe I did some bad clicking :lol:

Cheers

Spogg