Re: master 12578d6: Change how (declare (modes store the data

Re: master 12578d6: Change how (declare (modes store the data

[Stefan Monnier]

>     * lisp/emacs-lisp/byte-run.el (byte-run--set-modes): Change from
>     being a predicate to storing the modes.  This allows using the
>     modes for positive command discovery, too.

BTW, maybe the generalization is to replace the "completion-predicate"
and the "command-modes" into a function that returns a *score*, with
some convention where a score lower than X means the command can't
possibly work, and score lower than Y means the command is too unlikely
to be useful to be worth listing, a score higher than Y means that the
command has a special affinity with the current situation (like "written
for the current mode"), ...


        Stefan

Re: master 12578d6: Change how (declare (modes store the data

[Lars Ingebrigtsen]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

> BTW, maybe the generalization is to replace the "completion-predicate"
> and the "command-modes" into a function that returns a *score*, with
> some convention where a score lower than X means the command can't
> possibly work, and score lower than Y means the command is too unlikely
> to be useful to be worth listing, a score higher than Y means that the
> command has a special affinity with the current situation (like "written
> for the current mode"), ...

That sounds useful, but it sounds like a really difficult thing to
relate to.

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no

Re: master 12578d6: Change how (declare (modes store the data

[Richard Stallman]

[[[ To any NSA and FBI agents reading my email: please consider    ]]]
[[[ whether defending the US Constitution against all enemies,     ]]]
[[[ foreign or domestic, requires you to follow Snowden's example. ]]]

  > BTW, maybe the generalization is to replace the "completion-predicate"
  > and the "command-modes" into a function that returns a *score*, with
  > some convention where a score lower than X means the command can't
  > possibly work, and score lower than Y means the command is too unlikely
  > to be useful to be worth listing, a score higher than Y means that the
  > command has a special affinity with the current situation (like "written
  > for the current mode"), ...

There must be no end to the levels of complexity we could put into this
feature, but this could be past the point of diminishing returns.

I suggest people stop making this more complicated and see how people
like it.  In the mean time, we could work on improving other aspects
of Emacs.

-- 
Dr Richard Stallman
Chief GNUisance of the GNU Project (https://gnu.org)
Founder, Free Software Foundation (https://fsf.org)
Internet Hall-of-Famer (https://internethalloffame.org)

Re: master 12578d6: Change how (declare (modes store the data

[Stefan Monnier]

> There must be no end to the levels of complexity we could put into this
> feature, but this could be past the point of diminishing returns.

Given the efforts needed to mark up all the relevant ELisp code, I think
there's something to be said for trying to get as much as possible
from it.


        Stefan

Re: master 12578d6: Change how (declare (modes store the data

[Richard Stallman]

[[[ To any NSA and FBI agents reading my email: please consider    ]]]
[[[ whether defending the US Constitution against all enemies,     ]]]
[[[ foreign or domestic, requires you to follow Snowden's example. ]]]

  > BTW, maybe the generalization is to replace the "completion-predicate"
  > and the "command-modes" into a function that returns a *score*, with
  > some convention where a score lower than X means the command can't
  > possibly work, and score lower than Y means the command is too unlikely
  > to be useful to be worth listing, a score higher than Y means that the
  > command has a special affinity with the current situation (like "written
  > for the current mode"), ...

There must be no end to the levels of complexity we could put into this
feature, but this could be past the point of diminishing returns.

I suggest people stop making this more complicated and see how people
like it.  In the mean time, we could work on improving other aspects
of Emacs.

-- 
Dr Richard Stallman
Chief GNUisance of the GNU Project (https://gnu.org)
Founder, Free Software Foundation (https://fsf.org)
Internet Hall-of-Famer (https://internethalloffame.org)