-
Notifications
You must be signed in to change notification settings - Fork 26
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Sonarcloud #449
Comments
Maybe one more solution that can be considered is further tweaking sonarcloud. This would mean properly hooking up our tests to sonarcloud, telling sonarcloud about our naming conventions, and white/blacklisting files/sections of files that we do not want to be linted (took me a while to find, but sonarcloud supports it: https://sonarcloud.io/documentation/project-administration/narrowing-the-focus/). The only real blocker that you mentioned, in my opinion, is that sonarcloud does not work in forks. This could be worked around by only considering pull requests from local branches (something I as a fork user wouldn't mind), but this would still rule out checking forks with sonarcloud. We could also look at lighter alternatives that are easier to work with or can be run locally. But sonarcloud seems to be more or less the standard at the moment. (edit: CheckStyle comes to mind, but I do not have very good memories about that tool... Unless they've improved) |
One thing that I initially liked sonarcloud for is for insight in coverage. However, as that has been implemented in the vercors test suite, we don't necessarily need sonarcloud to check if the coverage is above a certain value. Besides that, if we want to do formatting/linting, we can also settle for scalafmt (and maybe checkstyle for java) or something similar. (Another thing I liked sonar cloud for is its static analysis capabilities, which I think we, as static analysis researchers, should employ. But sonarcloud only seems to catch the most mundane bugs, so we're not missing too much I think.) |
@Vescatur If you still remember, could you summarize what changes you have applied to achieve closing this issue? I remember you made sonarcloud more lenient but not exactly how anymore (might be nice for Pieter to read back later, and also for future reference) |
Changes to sonar are:
|
In my opinion, Sonarcloud is currently failing in its role as a useful step in our continuous integration, and we need it to change drastically. Please note that I'm not against linting in general, but these issues have become too annoying to leave them be. Some issues from the top of my head:
This leads to a situation where in almost any non-trivial pull request Sonarcloud fails the pull request. When deciding whether to merge the conversation usually is "oh it's just sonarcloud complaining? let's just merge it," so we merge pull requests even though the CI fails it. I think this, at a bare minimum, should not be the case.
Here are a few solutions:
I'm curious to hear your opinions!
The text was updated successfully, but these errors were encountered: