.ml stubs for exercises

[sshine]

Some language tracks have a {{ stub }}.ml that the student can fill out.

The Ocaml track doesn't. Ideally, add stubs to all exercises.

To begin with, start with ones that use Map from Base, like the four exercises in #277.

Issue spawned from this message I wrote on Slack:

I just reviewed the new version of Nucleotide Count and asked if there were any hard points, and [the student] said that the (module Char) annotations were difficult to figure out because they're not very well-documented. That was exactly the pain point that made it hard for me to convert the exercises from Core(_kernel) to Base. A solution could be to add a template .ml file with a stub in it that throws an exception. For some exercises, e.g. ones that require a data type definition repeated from the .mli, or in case of Nucleotide Count, a line like let empty = Map.empty (module Char), that'd be really neat.

[stevejb71]

If it's too much effort, perhaps just prioritise the initial few exercises, and nucleotide count.

[sshine]

Good point to have this order.

[sshine]

Another incident with a student spending way too much time (4 days) figuring out Map.empty (module Char) has pushed me to address this issue for Nucleotide Count. I'll look at the other exercises later.

[LucasMembrane]

I just spent a bunch of time on this with poor results, too. One thing I did not get from the readme is that Map is in Base when Base is opened (which I think is true, now). I was looking at the standard library, not aware that Base replaces Map. The standard library base documentation works ok for a while on this problem, but after a while the differences become more than apparent.

Another odd thing is that when I googled for solutions, the first three that I saw were apparently not doing the problem now out there -- I didn't see anything in them about the (Ok count) (Error c) (Ok map) returns, just counts and maps.

I have so far found these exercises a fine way to learn, so I have nothing but gratitude for your work, but since I came across this effort to improve things, I figured I should share my POV. Thanks much!

[sshine]

I have postponed this for nucleotide count for too long now.

Thanks for reminding me, @LucasMembrane.