On 27.02.2019 19:12, Stefan Monnier wrote:
> I see. I was taking the point of view of a generic `flex` method for
> use anywhere you like (I think the original motivation was to use it in
> conjunction with `icomplete-mode` to more closely match to `ido`
> behavior).
That's a worthy goal as well, of course. Not sure it will be ideal for
e.g. eglot's completion table, but that's a separate discussion, I guess.
I've been very busy and just caught up with this discussion. I can't contribute much more right now other than confirming Stefan's intuition for the original motivation. I wanted, and want, icomplete to be as usable as ido.el (with the added advantage of being well behaved). But I also want eglot.el and elisp's completion-at-point to be able to flex-match like, i.e. Sly does. So basically I want to be able to flex match everywhere.
Regarding sorting, after using it in icomplete, C-x b, C-x f, M-x for a while now, I can say this: There seems to be no case where I wouldn't like flex's scoring to be the dominant, but I haven't turned it on right now (mostly because there's no easy way yet due to this analysis-paralysis), and the table's scoring for those three seems sufficient in a reasonable number of the situations. When completing larger sets, such as company-completing all elisp symbols for writing code, flex scoring becomes indispensable, because the secondary sorting there is lexicographic.
Not sure what you can make of this, but if you can, make something :), and thanks very much for working on this.
João