WARNING! v2.05 has mystic Trigger/Timer BUGS

[sunsynth]

Strange behaviour gets even stranger...

since v2.03 my schematics (Midi Controllers with many timers) behave randomly wrong...

I even don't use the new trigger system trying to prevent this... (So my workarounds aren't useless)

maybe it's the size of the project (700000 mods/prims + 10% more links) but it appears in SM and in the resulting EXE's...
Here an actual screen shot:
(even a heartbeat trigger every second and/or a reset/start doesn't reactivate the timer...)
P.S.: Forgot 2 mention that it has a de-threader at the input...of course

[infuzion]

Just by looking & without an OSM to actually play with, I'd say the 2nd trigger counter will never increment due to link order.

[aliasant]

Just by looking & without an OSM to actually play with, I'd say the 2nd trigger counter will never increment due to link order.

So you mean the Stop kills that last trigger that the rightmost triggercounter is waiting for?

[infuzion]

Just by looking & without an OSM to actually play with, I'd say the 2nd trigger counter will never increment due to link order.

So you mean the Stop kills that last trigger that the rightmost triggercounter is waiting for?

Yes maybe; I'd have to test for sure but I have to go to work now & won't have time for days.

SM's trigger system is very powerful, but tricky to outsmart sometimes.

[aliasant]

SM's trigger system is very powerful, but tricky to outsmart sometimes.

This weeks understatement!

[trogluddite]

I doubt it is a link order problem; I use trigger delays linked just like this all the time without any problem (so far!)

I rigged up a schematic for testing...

Trigger Timer Test.osm

(4.85 KiB) Downloaded 116 times

...the timer receives flurries of 50 triggers, followed by a 50 tick break (long enough for the timer to reach it's end). So far, the 'finished' triggers match the number of input 'flurries' every time I run it, even after leaving it for half an hour while doing some other browsing.
Tried it with the new and old style triggers - same both ways.

My guess would be a "crossed threads" problem - the MIDI 'de-threading' problem is well known, but other things can cause it too (e.g. triggers from VST preset automation).

Or maybe there is a limit on how many windows timers can be active simultaneously?

[sunsynth]

Hmm...

thanks guys 4 testing - but:

i should have mentioned that not the trigger counter is the problem...
it's the led flipper!!!

The red LED should be off after 100ms but it keeps ON = red
AND this worked for years and month and hours...
now i get random drop outs!

A new hickup: crash on a XP while Mouse over with a timer 4 a popup (but only after a while running the App)
I managed the bus prob like always - get rid of it

Or maybe there is a limit on how many windows timers can be active simultaneously?

@malc: please give ur statement...

[tektoog]

I just tested it with Ver 1.1.7 (which I still use) and it works flawless.... for info.....

[trogluddite]

The red LED should be off after 100ms but it keeps ON = red

Tested here now, up into the 100's of trigger count.
SM2.0.5, new triggers on or off, two different machines - always works just fine.

I have also used Mouse Over + Timer for my help displays, has always been stable here. Do you have these problems with any other schematic? Could be like my arrays bug - this schematic making memory or thread problems, possibly?

[infuzion]

The red LED should be off after 100ms but it keeps ON = red
AND this worked for years and month and hours... now i get random drop outs!

I ran it while web-surfing for about 15 minutes; no visible issues. I also removed the De-threader (it is redundant) & ran it for a while, same no issues.
Sorry!

[stw]

works flawless here for more than an hour...

[infuzion]

Idea: Overall system CPU spikes causes windows timers to be thrown off?

[philter5]

Idea: Overall system CPU spikes causes windows timers to be thrown off?

isn´t this ONE of the main reasons WHY the green timer system is so unstable?....

[sunsynth]

remember:

even a heartbeat trigger every second and/or a reset/start doesn't reactivate the timer...

so even a simple trigger by hand didn't brought the timer back to live in my big schematic...

Idea: Overall system CPU spikes causes windows timers to be thrown off?

...maybe infuzion

new idea: even a midi loop within a midi network driver could be the reason - have to investigate further
will report later ...later

[McBarGig]

Idea: Overall system CPU spikes causes windows timers to be thrown off?

+1