[chirp_devel] Sharing reports
I was just looking over reporting.py, and was curious if you could open up the data for others to look at. Here's a few ideas I had:
we could see exceptions thrown as variants of existing models are brought out by the manufacturers
It would be nice to find common exception stacks and see if they're bugs that we can fix
Adding version info would be a big help. it would be good to know if there's a new exception hit on dailies that's not in 0.4.1, or vice versa
1+2 would also help catch if memory locations are wrong or have changed
Cheers,
Patrick (KG7NSC)
Sent from Windows Mail
I was just looking over reporting.py, and was curious if you could open up the data for others to look at. Here's a few ideas I had:
I've shared access to it in the past, it's not a secret. However, it really comes from a time when there were many fewer chirp users and thus fewer reports. At this point, there is so much being reported that sharing it too widely would be quite a stress on my machine (the reports themselves actually create a fair bit of load). Also, I have to purge the exception captures pretty regularly just for space reasons.
Since I last purged, there is 5.5G of exception reports. I'd be happy to get you a dump of all that if you'd like to browse it, but you can probably see that providing general access to that long-term is a pretty heavy commitment :)
- we could see exceptions thrown as variants of existing models are brought out by the manufacturers
Unless we upload the entire radio image, or some amount of identifying information, this will likely be hard to discern I think.
- It would be nice to find common exception stacks and see if they're bugs that we can fix
Yep, we report exception stacks now and there is probably quite a bit to be learned by scraping through the data for common bits.
- Adding version info would be a big help. it would be good to know if there's a new exception hit on dailies that's not in 0.4.1, or vice versa
Version information is reported already, so this should be possible.
--Dan
participants (2)
-
Dan Smith
-
stickpatrick@gmail.com