You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
So far the Java analyzer only provides automated feedback on submitted solutions, but analyzers can also tag them.
There are multiple approaches to implement this:
We can start outputting tags in the exercise-specific analyzers, and exercises for which no analyzer is implemented yet will receive no tags on their submissions.
We can implement a 'global' analyzer that is able to identify paradigms, techniques, constructs etc regardless of which exercise a solution belongs to, so that all submissions will receive tags.
We can combine both options of course.
I found that the C# Analyzer took the second option. This makes sense, since a large portion of the tags currently implemented are exercise-agnostic. Perhaps it can serve as a good example to get started.
The text was updated successfully, but these errors were encountered:
@ErikSchierboom and @iHiD: before we get started on this, I would like to check with you in what direction Exercism is currently going regarding the tagging of solutions. I know that some work was done to get an ML model to tag exercises for all tracks, does that mean that implementing tagging ourselves may no longer be necessary?
I know that some work was done to get an ML model to tag exercises for all tracks, does that mean that implementing tagging ourselves may no longer be necessary?
In principle: yes. We've done large parts of the ML training work, but aren't yet fully done. Not sure when we will, but you might be better off spending your time elsewhere until we've seen the ML part working.
So far the Java analyzer only provides automated feedback on submitted solutions, but analyzers can also tag them.
There are multiple approaches to implement this:
I found that the C# Analyzer took the second option. This makes sense, since a large portion of the tags currently implemented are exercise-agnostic. Perhaps it can serve as a good example to get started.
The text was updated successfully, but these errors were encountered: