Replacing linum-mode by nlinum-mode

Replacing linum-mode by nlinum-mode

[Stefan Monnier]

Could someone try to take care of replacing linum-mode with nlinum-mode?
I guess this would involve renaming linum.el (and its functions/vars) to
olinum.el, move it to lisp/obsolete, and then try to rename the nlinum
vars and functions to try and maximize backward compatibility.

Any taker?


        Stefan

RE: Replacing linum-mode by nlinum-mode

[Drew Adams]

> Could someone try to take care of replacing linum-mode with nlinum-mode?
> I guess this would involve renaming linum.el (and its functions/vars) to
> olinum.el, move it to lisp/obsolete, and then try to rename the nlinum
> vars and functions to try and maximize backward compatibility.

If you want to "maximize backward compatibility", why rename anything?

Why not just use nlinum functions wherever you want, in place of linum
functions?  Why gratuitously change user code at the same time?

(To be clear, I don't knowingly use either, and I have no code that
does.  My question has nothing to do with my personal use of Emacs.)

Re: Replacing linum-mode by nlinum-mode

[Stefan Monnier]

> If you want to "maximize backward compatibility", why rename anything?

> Why not just use nlinum functions wherever you want, in place of linum
> functions?  Why gratuitously change user code at the same time?

> (To be clear, I don't knowingly use either, and I have no code that
> does.  My question has nothing to do with my personal use of Emacs.)

Because the main reason to install nlinum-mode is to fix the bugs of
linum-mode, so users of linum-mode will only benefit from those
bug-fixes if they get transparently "upgraded" to nlinum-mode.


        Stefan

RE: Replacing linum-mode by nlinum-mode

[Drew Adams]

> > If you want to "maximize backward compatibility", why rename anything?
> >
> > Why not just use nlinum functions wherever you want, in place of linum
> > functions?  Why gratuitously change user code at the same time?
> >
> > (To be clear, I don't knowingly use either, and I have no code that
> > does.  My question has nothing to do with my personal use of Emacs.)
> 
> Because the main reason to install nlinum-mode is to fix the bugs of
> linum-mode, so users of linum-mode will only benefit from those
> bug-fixes if they get transparently "upgraded" to nlinum-mode.

They will benefit if they are "upgraded" to nlinum-mode, whether
that happens "transparently" or they do it intentionally & knowingly.

Still sounds more user-friendly to me to (a) fix all Emacs source code
to use the nlinum functions and (b) encourage 3rd-party code to move
to using those functions also.

Changing the names is not so nice.  3rd-party programmers will not
know without digging a bit that completely different code is now being
run under the same names (even the library name is taken over).

Not a big deal, but not super considerate of users.  And you did say
you wanted to "maximize backward compatibility".

Re: Replacing linum-mode by nlinum-mode

[Juanma Barranquero]

On Wed, Aug 14, 2013 at 4:35 AM, Stefan Monnier
<monnier@iro.umontreal.ca> wrote:

> Because the main reason to install nlinum-mode is to fix the bugs of
> linum-mode, so users of linum-mode will only benefit from those
> bug-fixes if they get transparently "upgraded" to nlinum-mode.

Some important differences betwen nlinum and linum were already
reported. Have they been fixed in nlinum?

   J

Re: Replacing linum-mode by nlinum-mode

[Stefan Monnier]

> Some important differences betwen nlinum and linum were already
> reported. Have they been fixed in nlinum?

No.


        Stefan

Re: Replacing linum-mode by nlinum-mode

[Juanma Barranquero]

On Wed, Aug 14, 2013 at 5:40 PM, Stefan Monnier
<monnier@iro.umontreal.ca> wrote:

> No.

Then, replacing linum.el with nlinum.el can hardly be called an "upgrade".

  J

Re: Replacing linum-mode by nlinum-mode

[Stefan Monnier]

>> No.
> Then, replacing linum.el with nlinum.el can hardly be called an "upgrade".

Well, AFAIK it's faster and more reliable.


        Stefan

Re: Replacing linum-mode by nlinum-mode

[Juanma Barranquero]

On Wed, Aug 14, 2013 at 6:00 PM, Stefan Monnier
<monnier@iro.umontreal.ca> wrote:

> Well, AFAIK it's faster and more reliable.

I've never had problems with linum's speed. Nor with reliability,
truth be told; though I know there are a few bug reports about it, so
perhaps I've just been lucky.

But, in any case, a functionality-removing change could only be called
at most a "sidegrade". Also, I seem to remember from the last time we
talked about this, that implementing the missing linum features in
nlinum wasn't exactly trivial, if at all possible.

   J

Re: Replacing linum-mode by nlinum-mode

[Stefan Monnier]

> But, in any case, a functionality-removing change could only be called
> at most a "sidegrade". Also, I seem to remember from the last time we
> talked about this, that implementing the missing linum features in
> nlinum wasn't exactly trivial, if at all possible.

There's no doubt that in some cases it's currently a downgrade.
If those can all be fixed, I don't know, but I think the important cases
can be fixed.

Would you be willing to try and be in charge of the change, then?
You seem to have a good grasp of what it would take to make the package
good enough to be a real upgrade.

I'd provide any help I can, of course.


        Stefan