Keep track of trimmed variables or removed variables
Every so often, I get really confused when as to why the variables are not matching the errors. After some time debugging and reporting I'm told again that we trim things from the variables.
I know that Relay tracks what they trim via _meta. Could we do the same for the Python SDK?
We could then give hints to the customer that data has been removed in the UI.
Hey @armenzg, the SDK is already doing this. It's just a matter of the UI displaying it. FWIW this comes up periodically, we've been routing the feedback to this Sentry issue. I'll close this but please reopen if there's anything missing from the SDK POV.
@sentrivana we kinda don't track this for vars anymore since we started serializing them early, so the vars _meta is lost in the process.
Forgot about that. 😢 Let's keep this open.
Thank you! 🙏🏻
Related issue: https://github.com/getsentry/sentry-python/issues/3310
But this issue suggests to get rid of trimming as a whole. We tried this already, but never followed through with removing all the trimming.
We need to have a discussion and then make a final decision if we want to remove trimming, or add better _meta information everywhere we trim data. And then implement the decision and be done with it.
Related issue: #3310
But this issue suggests to get rid of trimming as a whole. We tried this already, but never followed through with removing all the trimming.
We need to have a discussion and then make a final decision if we want to remove trimming, or add better
_metainformation everywhere we trim data. And then implement the decision and be done with it.
Can we do some work in the UI now? Or do you want us to hold off until you have a plan?