Update circular_buffer.h

[lucaato]

I've added function definition in header files to let the user know what are the function that need to be created.

[github-actions[bot]]

Hello. Thanks for opening a PR on Exercism. We are currently in a phase of our journey where we have paused community contributions to allow us to take a breather and redesign our community model. You can learn more in this blog post. As such, all issues and PRs in this repository are being automatically closed.

That doesn't mean we're not interested in your ideas, or that if you're stuck on something we don't want to help. The best place to discuss things is with our community on the Exercism Community Forum. You can use this link to copy this into a new topic there.

---

Note: If this PR has been pre-approved, please link back to this PR on the forum thread and a maintainer or staff member will reopen it.

[siebenschlaefer]

I'm not really opposed to this PR, but:

As far as I know it was a deliberate decision not to provide any declarations in most exercises. Students have to read the tests anyway to understand the requirements, extracting the definitions/declarations of the public types/functions should not be that hard, and there are a few choices that students can make (public definitions of struct in the .h file vs. opaque types, sometimes void return types vs. some sort of error code, etc.), and (in general) writing .h files is part of the development process in C.

Also, Exercism prefers to discuss these kinds of changes in the forum first to get feedback and reach some consensus. (Tracks can opt out and keep the discussion here on Github, but AFAIK that's not the case for this C track, right @wolf99 or @ryanplusplus?)

[lucaato]

Hi,

it is true that writing .h files is part of the development process, but if I already have an idea of what a circular buffer is I just need to know what are the function name to get going. It's also true that users should look at the tests but from the website small side panel is not always easy to do, so i think the header file should at least give an idea of what are the function names, and maybe a little explanation of what they should do, everything can be written down in a minute comment. Later when test fail users will have the chance to look at testcases in more details.

It's just an idea, if I have time I'll write about this in the forum. Thanks for the support.

[ryanplusplus]

Also, Exercism prefers to discuss these kinds of changes in the forum first to get feedback and reach some consensus. (Tracks can opt out and keep the discussion here on Github, but AFAIK that's not the case for this C track, right @wolf99 or @ryanplusplus?)

I personally don't make it into the forum much and like to have these conversations in a GitHub issue so we can discuss the path we want to take prior to a PR.

[ryanplusplus]

Hi,

it is true that writing .h files is part of the development process, but if I already have an idea of what a circular buffer is I just need to know what are the function name to get going. It's also true that users should look at the tests but from the website small side panel is not always easy to do, so i think the header file should at least give an idea of what are the function names, and maybe a little explanation of what they should do, everything can be written down in a minute comment. Later when test fail users will have the chance to look at testcases in more details.

It's just an idea, if I have time I'll write about this in the forum. Thanks for the support.

Thanks for this information, it helped me to understand why it's not already easy to figure out what the function names are by looking at the tests. I think this is good feedback to bring to the Exercism website team. Reading the tests and understanding the requirements that they encode is an important part of the development process in Exercism. If the online editor is not making that easy, then we should make a fix in the online editor instead of trying to design the tracks so that you don't need to see the tests.

[siebenschlaefer]

I personally don't make it into the forum much and like to have these conversations in a GitHub issue so we can discuss the path we want to take prior to a PR.

If the other three maintainers (@ryanplusplus, @bobahop, @patricksjackson) agree you can let @ErikSchierboom know so that the automatic closing of PRs that directs to the forum gets disabled.

[ryanplusplus]

If the other three maintainers (@ryanplusplus, @bobahop, @patricksjackson) agree you can let @ErikSchierboom know so that the automatic closing of PRs that directs to the forum gets disabled.

I'm on board. I've just been reopening issues and PRs that the bot closed.

[wolf99]

I personally don't make it into the forum much and like to have these conversations in a GitHub issue so we can discuss the path we want to take prior to a PR.

If the other three maintainers (@ryanplusplus, @bobahop, @patricksjackson) agree you can let @ErikSchierboom know so that the automatic closing of PRs that directs to the forum gets disabled.

I'm ok with this also.

[bobahop]

If the other three maintainers (@ryanplusplus, @bobahop, @patricksjackson) agree you can let @ErikSchierboom know so that the automatic closing of PRs that directs to the forum gets disabled.

Okay by me as well.

[ErikSchierboom]

Seems to have been done in https://github.com/exercism/c/commit/f9dfff1501e462e7550f3bd44307b43801b2a1c3?

[siebenschlaefer]

@ErikSchierboom Yes. See issue #904

[wolf99]

I had a quick try of the circular-buffer exercise on the website. For me the website UI looks OK, but the barrier to entry for this exercise is high because so much implementation needs to be in place before the code will even compile without errors.

Suggestions for how to resolve this:

1. Break up the tests to simpler steps. This could be tests additional, but earlier, to those from teh exercise specification, or we could work to update the specification.
2. Raise the difficulty level from easy to something above medium.
3. Provide header stub for this exercise.

Personally I think a mix of 1 & 2 would be the correct approach, but I'm open to 3.

My reasoning:

• 1 is better than 3 because it fixes the issue rather than "hiding" the issue.
• 2 is better than 3 because it gives correct information rather than "hiding" the issue (also, IIRC we dont have many, iexercises marked above medium difficulty).

[siebenschlaefer]

I'm not sure I understand what you mean with "1. Break up the tests to simpler steps."
Do you want to add additional tests at the beginning of test_circular_buffer.c that make the required interface easier to extract?

And I agree with (2), this is not an easy exercise, But I'm a mentor, not a maintainer, so take this just as input, not some sort of vote.

[bobahop]

It's certainly worth considering these things.

A common suggestion I make when mentoring circular buffer is to define the buffer as an array as the last field of the struct, instead of as a pointer, so that it doesn't need to be separately allocated and freed. By providing the header stub, the student may not know why it is like that, unless they learn by the example, or we provide some explanation in an About, or they ask about it. So I'm a bit wary of providing the header stub.

I definitely don't think Circular Buffer is "Easy", so I think at least "Medium" is better.

As for fiddling with the tests, as a student who's done 870 exercises, I do not do a happy dance when an exercise becomes outdated. If it is not substantive, it's annoying, And if it is a substantial change that breaks my solution, I curse that it just wasn't made a different exercise. So I would hope that any test changes bring some real benefit. It might be helpful to have stats on how many try circular buffer and abandon it. I think they are available on the site somewhere, and I always forget where they are. I'm not sure that CB has been a problem to more than a few students over the years.

Edit: Build Status

About a 48% completion rate, which is even higher than Armstrong Numbers. If we're not looking to change Armstrong Numbers, then perhaps CB is okay?

[siebenschlaefer]

@bobahop armstrong-number gets displayed as the first exercise after hello-world. I guess it has a low completion rate because this is the first time students have to solve a problem in C on their own and some give up. (but that's just a guess)

[bobahop]

Perhaps that's all the more reason to look at Armstrong Numbers first? Maybe move it further back? It has over 80% completion rate on C++ track.