#714 "Long life to Pagure" awarded in error
Opened a year ago by dreua. Modified 8 months ago

I just got awarded with all the "Long life to Pagure" badges. I guess it should be level I and probably II, but definitely not V. Don't know if it is worth the investigation but I wanted to report it just in case.

My badges: https://badges.fedoraproject.org/user/dreua

Event for III to V: https://apps.fedoraproject.org/datagrepper/id?id=2020-7238450b-a110-4133-956b-17255d396e9c&is_raw=true&size=extra-large

Maybe the algorithm got confused because I re-commited a commit by releng? (I just merged master - including the f32 rebuild commit by releng - into f31, can't imagine I'm the only one doing that.)

Metadata Update from @riecatnor:
- Custom field artwork adjusted to None
- Custom field concept_review_passed adjusted to None
- Issue tagged with: bug

a year ago

Although I really like having "Long life to Pagure", I've got from level I to level VI with one push too ;)

First level I've got today with one commit (although I've added commits before), and the next 5 levels with a push after merging master into epel8.

First one: https://apps.fedoraproject.org/datagrepper/id?id=2020-71613048-4f16-4d98-b522-505d67e798c1&is_raw=true&size=extra-large (1 commit)
Next five: https://apps.fedoraproject.org/datagrepper/id?id=2020-2a96cb64-07dd-465a-b7a0-d025b43da266&is_raw=true&size=extra-large (65 commits)

Counting previous commits I'd estimate myself being at level IV, as I've already pushed another merge sometime with >100 commits, but level VI seems too high.
(Error in merge-related counting is more likely though)

Same issue here, in my case I can only count one commit, awarded for this activity:


I was awarded levels II, III and IV for which I don't have commits.

Here is the rule for level V:

The other levels are exactly the same except for the obvious changes of name, description, image_url and the number in line 32.

Did this ever work as intended? The filter for users is "%(msg.authors)s" and all reports here (as well as some other users with all tiers awarded on the same day) have multiple entries in "authors", therefore it is quite likely that it is counting all commits of these users together but awarding badges to the single user. (This happens all the time when people push other people's or releng's commits to other branches.)

Possible fix but I have no clue how to test it:
Matching just for msg.username instead of msg.authors should work.

Note that we can't copy from "you can call me patches" here since the commit object that is used there is not present in messages from pagure for some reason.

You can call me patches (Pushed commits to Fedora package repository):
Users are filtered for "%(msg.commit.username)s", see rule:
Example message: https://apps.fedoraproject.org/datagrepper/id?id=2020-deb447bb-7a85-40c1-a531-b3807c321bcf&is_raw=true&size=extra-large

(This happens all the time when people push other people's or releng's commits to other branches.)

Seems like that's what has happened with me when I rebased one my fork on src.fp.o,

It's the same event for all the badges, II to V, https://apps.fedoraproject.org/datagrepper/id?id=2020-503a0131-7fa9-47d5-b248-09d20c297b75&is_raw=true&size=extra-large

On the other hand I came to know about the "you can call me patches" badge recently and realized I never received one (I do have 5 commits but all those were merged in form of pull requests). That should be reported as a different issue I believe.

Login to comment on this ticket.