I use multiple “log.debug” entries to debug my custom Handlers.
But obviously the SmartThings IDE Log is subject to multiple race conditions, and I can never predict which of my log.debug entries will actually be displayed, or not, in the Live Logging window. And it is completely random : the same code executed twice gives different lists of log entries.
It does not happen 1 time out of 100 (which I could live with), more like 1 out of 4…
SInce the whole purpose of a trace is to check where your code goes or not, missing entries can make the whole debugging process intensively more painful and way too much “hit and miss”.
This problem is not new, although I got the impression the percentage of missed entries was increasing with time those last 2 years.
I was wondering if anybody had found a way of using the IDE Log in a way making it less forgetful and more reliable ?
I’ve found that Live Logging is more reliable (i.e. more of the messages displayed more of the time) when I’m using a wired ethernet connection to my router, compared with a wireless-n connection on my laptop. Though it’s still not 100%.
I agree that this lack of completeness and consistency in the Live Logging output is highly undesirable considering it’s intended purpose for debugging. It is no doubt causing a lot of confusion and unnecessary headaches for community developers.
Not sure if the issue has been covered in this thread: Tell me your IDE issues
FYI, to combat the problem, I implemented a form log concatenation/pagination in my Z-Wave Tweaker DTH.
I have seen live logging “drop” a lot of messages. I don’t know if it is rate or queue limited, but if you send a lot of messages, you are sure not to see a lot of them.
I’ve been coding a new DH over the last couple of days, logging has been the worst I’ve seen it. It has never been 100%, which is ok, I guess. My experience over the last 72 hours has been more like 30%.
I confirm the same. like 70% of my many log.debugs are missing. When I comment out most of them, I can see the rest. But then it’s hard to trace the big picture. This is so bad. I am wondering if three is any 3rd party log solution.
Sad that 3 years later the log is still not working properly…
The only workaround, assuming you can “loop” around the code you want to trace, is that after enough executions, you will probably have seen all your log entries if you “merge” them across multiple loops.
And yes, I know, not really a solution…
I’ll chime in here, Live logs completely stopped over a month ago. Maybe because I’m on a V1 hub but it would nice to have some indication on what is going on.
Live Logging is still working for me and my dozen of V1 Hubs.