@samth if you’re around and have 5 minutes to look at the coverity dashboard, can you please let me know?

@jacob.d.mitchell @greg Interesting, I was not aware of exercism. I have registered as a mentor, but it’s not clear what the next steps are.

@pocmatos sure

do you have 5 mins now?

yes

@samth great, the issue with the coverity scan needs to be sorted by someone with write permissions to the racket repo.

oh ok

So, if you could, register with coverity start a project and then you can add me as a member and I will take it from there.

https://scan.coverity.com/users/sign_up

Then you can signup and after that in the dashboard, you can create a project from GitHub.

I’m confused

do you have some existing Racket project that’s not working?

can we fix that?

also, have you actually gotten useful scan results yet?

So, I have created a project for my fork which is working - because I own it.

But you are not allowed to change the url of the project after project creation.

Then I created another one manually (not directly through the github integration page) for racket/racket because I didn’t have write permissions.

Was just told I cannot integrate that one with Github because was created manually.

I have scan results for my fork. Shall I share that with you?

yes, i’d be interested

That’s, as I pointed out in the bug, one way to do it. I get my fork to track upstream, check the problems on my fork and create new issues upstream.

then we won’t have all of this annoyance of regular merges upstream, etc. I can do it on my fork.

I think if the scan results are already useful, then we should figure out how to do it automatically

If the scan results will take more work to become useful, we should fix that first

I have added you as a member of LinkiTools/racket. You should be able to see the last results I ran.

This is the direct link: https://scan.coverity.com/projects/linkitools-racket/view_defects

Here’s an example:

That one looks interesting, maybe @mflatt can chime in as well on this.

@samth let me know what you prefer todo: keeping track of this through the fork, or trying to get this working upstream.

I believe that guard should indeed be >= 31 instead of > 31.

That’s a 1–0 score for coverity then. < @samth

@mflatt @samth submitted #2281 https://github.com/racket/racket/pull/2281

@pocmatos my guess is that the best thing to do is go through the current failures — that will provide more bang for the buck than the CI issues

@samth Sure, I was skimming through it and there are quite a few issues that deserve a close inspection. I will transform them into github issues, prs or into the false positive mask files.

awesome! thanks a bunch for doing this

@samth happy for you to close the pr if you want to leave it for the foreseeable future.

Awesome, thanks. Not having gone through the process I’m not completely sure. I get the impression there’s an approval process for mentors and once finalized you should expect to be listed on the Racket track page I linked earlier and perhaps receive more details from Exercism about next steps.

Scribble docs searching question: how can I restrict the search to only module names? H:and R: work for #lang and #reader implementations, but M: shows both module names and bindings from that module. I don’t want to see bindings, I just want module names.

@notjack I think the statement “‘M:’ by itself will restrict results to module names only” is a documentation bug, it should say “‘L:’ by itself”.

_gasp_bamboozled again!

that worked

so it seems I can get what I want by combining L: and M:str

e.g. L: M:typed for tons of TR modules and some other stuff