[m-dev.] proposal: rename the modules of the first termination analyzer

[Zoltan Somogyi]

On Sat, 29 Aug 2026 21:59:01 +1000, Julien Fischer <jfischer at opturion.com> wrote:

> On Sat, 29 Aug 2026 at 21:16, Zoltan Somogyi <zoltan.somogyi at runbox.com> wrote:
> >
> > The names of the modules of our second termination analyzer
> > all follow a common pattern: term_constr_X. This is not true
> > for the first (since we didn't know then that there would BE a second :-)
> > but it would be nice if it WERE true.
> 
> Most of the modules for the original termination analyser have the form term_X.
> The exception is termination.m.

Yes, but unfortunately term_constr_X also has the form term_X :-(

> > The main difference is that the second system uses polyhedra, while
> > the first used linear inequalities, which compare the size of the outputs
> > to the size of a subset of the inputs plus a constant. The subset was
> > the "output supplier" inputs, meaning the inputs that control the size
> > of the outputs. I therefore propose that new naming scheme should be
> > one of these three: term_osi_X (short for output supplier inputs),
> > term_ineq_X, or term_lineq_X (inequalities and linear inequalities
> > respectively).
> >
> > Does anyone object to the renaming, and if not, what are your preferences
> > about the naming scheme?
> 
> Of the three you suggest, my preference would be term_osi_X.
>
> (Incidentally, both analysers could probably be restructured to use submodules.)

Agreed.

Zoltan.