Re: Is Elisp really that slow?

[Dmitry Gutov]

On 17.05.2019 17:36, Eli Zaretskii wrote:

> You said you thought the competition was more popular because they are
> more consistent.  We need evidence to support that, if we are to
> believe that this is a reason.

I think you're asking too much. The only way to undoubtedly prove 
Emacs's popularity is hurt by inconsistency is to create a mass A/B 
trial with volunteers spending time with different editors, as well as 
with a "fixed" version of Emacs, for some definition of "fixed". And 
then observing the behavior of al participants over the years.

The only things we can really do is to make the improvements (or, say, 
particular changes, to call them neutrally) and then somehow observe 
whether that improves things over time.

I never said that the competition was more popular *only because* it was 
more consistent, because there are a lot of things at play: installation 
base (for Vim), lots of marketing (mostly VS Code lately, but Atom too), 
market capture, certain features, of course. But user friendliness a 
large part of that. I think most people would agree that the user of a 
familiar editing paradigm is a serious advantage in terms of popularity.

We're probably not going to change our keybindings wholesale, but 
improving consistency can at least help learn and become comfortable 
with them faster.

And technically, changing key bindings is an easy task, so that wouldn't 
require too much actual work. I might also improve our image a little 
within the existing community, where the prevailing thinking is, 
unfortunately, that we're afraid to change long-standing behaviors, even 
if it's ultimately for the better.

> And if it is a reason, we should work
> to eliminate those inconsistencies.

So my vote is to simply work toward better UI, one step at a time. As 
long as the total capability of the features that we provide is not 
diminished, I do not think most existing users would really become 
upset. Similar to how John Yates put, it's going to take more than some 
changed/removed key bindings to drive away old and experienced users.

> Thus, a single example certainly is not enough to identify a general
> problem that could explain why we are less popular.

It's not the question I've been trying to answer.

> > Why not fix at least one thing, then?
>
> I didn't say I objected to such a change.  So this question is not for
> me to answer.
Who then? Alan, who never saw this thread? Would you say that somebody 
should finally file the bug report?