bugbug

bugbug copied to clipboard

Add a section about triage statistics in the quality notification

1

For example, # of S1, # of bugs with no priority, time to triage, etc., etc.. We could start with average time to set priority/severity, give a list of bugs...

marco-c

See also #2165.

marco-c

Try using doc2vec as similarity algorithm

36

After #637, we should try with doc2vec too and see what works better.

marco-c

In `extra-nlp-requirements.txt`, the gensim version is mentioned as `4.0.1`. But it looks like bugbug is using gensim `3.8` version. There are some major changes in gensim `4.0.1`, and therefore many...

rock420

Being able to fast-track (manually?) generating and deploying HTTP service on release

4

Right now, we are generating a Docker image with missing models for the HTTP service so we cannot deploy at the end of the release. But that means that all...

Lothiraldan

Currently whenever we add an option we need to add a new argument to run.py, add a new parameter to all model classes, etc.. We should find a better way...

marco-c

E.g. there is a big difference between a patch which is likely to cause a performance regression and a patch which is likely to cause a security bug.

marco-c

Write a test to make sure a new DB is downloaded when an already downloaded DB exists with an old schema

34

In the download method, https://github.com/mozilla/bugbug/blob/d1016a351603d948230188fa3f370a057a1a8127/bugbug/db.py#L89, we should handle the case where the DB is already downloaded but is using an older schema version than the current one.

marco-c

lead time is the time between when a patch lands to when it lands in release. This varies widely for Firefox based on if the patch landed early in the...

krupa-raj

It would be useful to see Regressions per team alongside - patches landed: is there a lot of churn in this component which is causing more regressions - Test coverage:...

krupa-raj