So I'm not sure whether this is a qlab or eos issue (I think possibly qlab..) but I was doing some testing today to try and track down some intermittant missing of OSC cues in qlab fired by eos - what I've found odd is that often the same cue will be missed throughtout the run of a show - not every show but when a cue is missed it's often the same one..
To test this I did the following..
- Created a Qlab show which listens for cues, and logs when a cue is recieved. If the cue number is not the previous cue number +1 then it also logs the date and time, otherwise it just logs the cue number. Qlab file at https://www.dropbox.com/s/xqwbwo7brakdenw/Network%20Test.cues?dl=0 with 100 cues (1,2,...,100)
- Created an eos show file with 100 cues (1>100), each cue having a follow of 1 second and the last cue linked back to the start. set the "eos cue send string" in osc to /eos/cue/%1/start and then set it all off going
What I found is that sometimes (but not always) qlab would log cue 56 being fired twice! If you look at the logs at https://www.dropbox.com/s/5lrqyqloi85neht/NETWORK%20TEST.txt?dl=0 you will see that sometimes - but by no means always - the log shows:
55
56
Wednesday, 30 May 2018 at 22:47:00 56
57
Which means that cue 56 was fired twice by Qlab.
Part way through this logging I started wireshark - what I noticed in this is that - as far as I can tell - there is nothing different about the osc string that's sent at the times that the double go has been received, which does suggest to me that the issue is an Qlab issue - but I thought it worth highlighting here. The wireshark file is at www.dropbox.com/.../wireshark.pcapng
If useful the logs for the eos show file are at https://www.dropbox.com/s/s4xw6nhz0cxizt9/ETC_Logs_Mac_Offline_05312018_000646.zip?dl=0
Would be interesting to know if anyone else has seen a similar issue.
I tried restarting the system and was still getting the intermittent cue 56 issues..
I'm using eos version 2.7 (latest release) and I've tried it with Qlab 3 and Qlab 4 - both seem to exhibit the same issue, always with cue 56.
It's fully possible that I've done something silly somewhere - and it could be said that it's better to fire something twice rather than not at all, but it feels like this could be the reason I've had issues in the past..
Actually, while writing this I did a little more testing..
If I link Cue 60 back to cue 50 in eos and just run this bit, Qlab seems to miss out LX 56 and 58, and just shows LX 57 firing 3 times - see https://www.dropbox.com/s/1tnga4bh86v73ev/NETWORK%20TEST%202.txt?dl=0 - again the expected OSC appears to appear in wireshark..
Hopefully this makes some sense to someone!
Richard