[m-rev.] for post-commit review: maybe_succeeded

[Peter Wang]

On Thu, 12 Aug 2021 06:06:01 +1000 "Zoltan Somogyi" <zoltan.somogyi at runbox.com> wrote:
> The diff is long and boring. The main issue I seek feedback on
> is whether there is a better home for the new maybe_succeeded type
> than process_util.m. Many of the primitives that can succeed or fail
> are defined there, but not all, and with this diff, some modules import
> process_util that didn't import it before.
> 
> The new type does not fit in well anywhere else that I can see;
> e.g. it is quite different from everything else in maybe_error.m.
> The only two other places I can think of for it are a new module
> in the libs package (libs.succeeded or libs.maybe_succeeded),
> or (possibly) definition in libs.m itself.
> 
> Ideas? Opinions?

I think libs.maybe_succeeded is fine.

There could be a module for basic types with no/few dependencies and no
obvious home, e.g.

    libs.aux_types
    libs.basic_types
    libs.misc_types
    libs.simple_types

which would be fine as well.

Peter