DSP Robotics Support

I Submit My Test Module!



Download & export it as a VST then try it in reaper. The pulldown which has its interaction in Ruby does not work for me. Lets see where this goes!

Drnkhobo wrote:
Download & export it as a VST then try it in reaper. The pulldown which has its interaction in Ruby does not work for me. Lets see where this goes!

i don`t have reaper, but it works on cubase, live and renoise..

Testing .... [WinXPPro - 32]

Reaper[v:4.32] - both work

WaveLAB 6 - both work. no checkmark on SM version but also do see it in FS either.

StudioONE [v:2.06]. both work

Adobe Audition [v:3.01] ... did NOT load the VST plug

Sonar [v:8.5] - SM version works RUBY ---- NOT work

in SONAR ... the RUBY version is not working. Clicking on it does nothing !

Audition could even find the test plug even after re-scanning


I should also mention that I have had issue with the Ruby version of pull-down. In one project, clicking and selecting an item ... the Mouse Arrow would disappear completely. The only way back was to blindly locate the QUIT FlowStone. I copy my old SM version in and replaced the FS one ... no further problem.

I thought I read in the late FS update notes that this issue had been spotted and fixed. It seems the disappearing mouse arrow hasn't happened again ... but the FS RUBY menu is completely dead in SONAR

These are not all the latest DAW versions ... just the ones I stopped at. Reaper has slowly become my default DAW [but even it is not the latest version].

Post DAW Version Numbers !!!! this could be another variable to consider.

Something else noticed during testing of a couple of my project [VST].

Some of my VST were not made with the current version of FS, they were with the previous version.

It was strange that in WaveLAB [i believe], my newest plugin would lockup WaveLAB when I tried to load a text file. The VST screen would turn white [blank]. I had to force a shutdown of WaveLAB.

When I loaded a different VST I made with the previous version, I was able to load my configuration file with out issue ... just like it should !

I need to check this closer ... but this really makes me wonder if there is a BUG floating around. The 2 VST tested are different, in that they have a different look GUI-wise ... but the function of the Save & Load routines are not really different ... yes, my latest plug uses the Marshal RUBY design ... the other was prims with plain text files.

Another strange thing ... I have a Registration file that is searched for when plugin loads ... in some of the DAW's my VST is loaded as a DEMO ... this is not right. My LED indicator shows that the Registration file was NOT found, thus not loaded. But in other DAWs [including my testbed REAPER] ... it all works.

Is this confusing or what

It appears that I was right about ruby.

As for loading plugins as demo.
Guess 1) Loaded plugin is copied somewhere else than the location of serial number.
Guess 2) You may need to "delay" (via timer) afterload prim (an old error, since SM) when loading external file (like serial number) - thus it should be triggered when filesystem is ready.
Guess 3) You may need to add an option "retrig" for reloading serial number file (if S/N is latched in some rigid way, then loading after latch - will not work).

Guess 1) Loaded plugin is copied somewhere else than the location of serial number.

no ... the plugin location hasn't changed. Each DAW tested is loading the same DLL. I'm using the PRIM that identifies where the plug was launched from [FOLDER]. The entire Registration routine is controlled from Trog's module 'Save Load Application Data settings file', with AUTO-LOAD set to TRUE.

Guess 2) You may need to "delay" (via timer) afterload prim (an old error, since SM) when loading external file (like serial number) - thus it should be triggered when filesystem is ready.

hmmm ... did not know about this In Trog's module, he has an 'AfterLoad' prim that was doing the trigger to auto-load the file. This has a problem ???
I looked at the TIMER prim ... I have this in a 'delayed trigger' module ... guess I could try this. Any suggested 'delay time' ???

Guess 3) You may need to add an option "retrig" for reloading serial number file (if S/N is latched in some rigid way, then loading after latch - will not work).

Not understanding this. Might this be using the 'De-Threader' module to sent a clean trigger ?? need clarification please.

Thanks 'tester' for giving ideas to try ...

Still some questions about what are these issues from. Like the Pull-Down menu [RUBY] not working in Sonar, yet the SM version does ?? Is the problem with Ruby ??? ... or the Ruby coding ???? We need one of the Guru's to identify what the problem is. At this point, we should replace all Pull-Downs with the original SM version I must say that the FS version has a better look to it ... but if it doesn't work across all test systems, looks must take a back seat.

The 2nd major issue is my beta-testers 2 machines failing DRNKHBO's 'Save/Load' test. He's verified his computers system files, no background scanning software to interrupt, runs in ADMIN mode so permissions should not be an issue ... the thing is .... the config file he saved and sent to me is corrupt.

so 2 things with this ... he can Save a file with my project, but when he runs this test VST, not only does it fail to read it ... it never got written or he can't find the darn thing

So ... what can cause a file to be corrupted when written ??? My Ruby module was posted in this thread for all to see ... no one has mentioned seeing an error in it ... and of course, works perfect on my XP system

I'm thinking maybe I should go back and test the Marshal routine, but with a standard format rather than Raw Binary. I don't know ... the standard format worked just fine for me too ... I just need to be able to write a simple text file and read it back This is not suppose to be the hard part, but it has stop my project before it had a chance.

Last question [yeah ... probably not] ... This latest change to FlowStone, and in particular the change in how things changed from the previous version [Ruby ???]. Is it remotely possible that some library conflict could happen if someone has 2 VST programs [one written on each of these 2 FS versions] ??? Could an older library be causing a problem on my testers machine ??? I don't even know what I'm asking, cause this is not making sense.

If this is NOT a FS issue [how would I know] ... any idea/suggestions I can relay to my beta-tester ? Is there another test app we could get together that might generate some type of error message ? Right now we got nothing.

I hope we get more test results posted ... it can't be just a handful who care if their project works on other systems.
I'm not scolding ... but lets at least identify any potential surprises. [like the pull-down Ruby menu].

really need help ... for all of us

i think its time to make two testschematics, the one of drnkhobo and one with the marshal methode, test it on the different hosts and report the results including the testschematics to the FS devs.. so they know exactly where it works and where not

a.d.2.) I use timer delay with "whatever time" (like 1ms?), just to add "trigger separation", not liked by some gurus here It changes the order of events in SM/FS, and thus - afterload trigger is executed after the filesystem is loaded within FS. Af far I remember, that issue wasn't fixed (I just use that solution since SM, to load external files on start).

ad.3.) some sort of "distorted event order" idea.

A) Can you add a small clicker/area/button to gui, to manually retrig loading file with serial number? So that you can check if the plugin will reload serial number after clicking?

B) Will the plugin become unlocked anyway - after such retrig, or there is some fixed order of events, that prevent that?

4. If ruby does not works correctly under some hosts, and ruby is involved in loading or decrypting (or routing) the serial number, then it is obvious, thatthere an error will occur.

Hi Tester,

4. If ruby does not works correctly under some hosts, and ruby is involved in loading or decrypting (or routing) the serial number, then it is obvious, thatthere an error will occur.

thanks for the additional details ... I'll test the 'delayed trigger' suggestion for AfterLoad.

The serial number loading/processing has NO Ruby involvement at all. It is all through PRIMS.

I was looking over the module that handles the serial loading and noticed that the file 'extension' was giving unexpected results. Then I remember ... I changed the filename to NOT have an extension so it came down to idiot programmer [me] .... NO there

... I'm sending the pull-down menu test to my guy to run on his systems ... try to get as much test results as we can.

Still the big question is the cause of the corrupt file writing on my [tester's] system ... that is really strange.

Crazy ... was checking the FS version of 'Pull-Down' menu.

Pulled it into a new blank schematic ... make a selection from one of the default list ... the display does NOT update

The OUTPUT connections do output the correct values, but the actual display just stays 'MENU'.

This is within FS !



Can anyone verify this ?!?

***** EDIT ****

Actually, I found a 2nd Menu Selector [all RUBY] ... now THIS one works [at least in FS]. It's called 'Selector'.

The broken one is called 'Menu' ... maybe I got this one from the forum ... don't know ... anyway ... do you guys have both ? What the status if so ?

Of course ... these may be 2 different tools. The Properties pages are different. The one that didn't update the name uses a LABEL. Maybe this is for a very different design/need.