I am currently trying Unison for Mindshifting May. As there is no syllabus for this track, I am a bit lost when it comes to choose my next exercise.
I think having the possibility to sort exercises by the number of validation would be a great feature, especially for tracks without syllabus. Edit: By “number of validation” I mean “the number of users who validated the exercise”. This indicator is used in Root Me and Leetcode.
I feel like it should not be too hard to add or too computationally expensive, but I may be wrong.
What do you think ?
Hey :) What is the “number of validation”?
Sorry, I forgot to precise that !
By “number of validation” I mean “the number of users who validated the exercise”.
This indicator is used in Root Me and Leetcode.
You mean the number of users who solved and submitted the exercise? That would be the number of public solutions.
Yes, that’s right.
I still have a hard time using the Exercism’s terms : solved, submitted and published.
The order is actually set manually by maintainers. It’s normally roughly based on the difficulty of the exercise. We are considering having some special sort of groupings of exercises too (e.g. a consistent core set across all languages that get highlighted). Sorting by submitted (effectively “most popular”) is a nice idea too.
Leetcode has something called “acceptance rate”, i believe it’s calculated as the number of accepted attemps divided by the number of attempts. Hard problems often have low acceptance rates, easy problems tend to have high rates.
Only root-me appears to have a “validation” number, but i don’t see something similar on leetcode.
Oh right, my memory is playing tricks on me.
Well, that could be another alternative.
It has the advantage be rapidly representative for new exercises, whereas the validation number would be off for some time.
But it can be affected by false fails, when the code is right but there is an error during the test run. I had a few with Unison today.