path: root/2025/captions/emacsconf-2025-modern--some-problems-of-modernizing-emacs--eduardo-ochs--main.vtt

WEBVTT
Kind: captions
Language: en-GB

00:00:54.000 --> 00:00:55.000
 

00:00:55.000 --> 00:00:57.000
Hi! My name is Eduardo Ochs. I'm the

00:00:57.000 --> 00:01:00.000
author of an Emacs package called eev and

00:01:00.000 --> 00:01:03.000
the title of this video is

00:01:03.000 --> 00:01:05.000
"Some problems of modernizing Emacs".

00:01:05.000 --> 00:01:08.000
Here is a summary of the main themes

00:01:08.000 --> 00:01:10.000
of this video. I'm going to talk mainly

00:01:10.000 --> 00:01:12.000
about these four things here. The first

00:01:12.000 --> 00:01:15.000
one is that Emacs has changed a lot in its

00:01:15.000 --> 00:01:18.000
recent versions, and now it has lots of

00:01:18.000 --> 00:01:21.000
types... so if we want to look under the

00:01:21.000 --> 00:01:24.000
hood and to understand what Emacs

00:01:24.000 --> 00:01:27.000
really does we are going to stumble on

00:01:27.000 --> 00:01:30.000
lots of types... and the

00:01:30.000 --> 00:01:34.000
current tree of classes and types

00:01:34.000 --> 00:01:37.000
looks like this... that is,

00:01:37.000 --> 00:01:46.000
is quite big.

00:01:46.000 --> 00:01:49.000
The second theme is that people used

00:01:49.000 --> 00:01:53.000
to say things like "Anyone can learn Lisp

00:01:53.000 --> 00:01:56.000
in one day"... I'm going to explain

00:01:56.000 --> 00:02:01.000
this quote, and I'm also going to show

00:02:01.000 --> 00:02:04.000
that now this is gone... anyway. This is a

00:02:04.000 --> 00:02:08.000
very short summary... details soon.

00:02:08.000 --> 00:02:10.000
I will also show how to display

00:02:10.000 --> 00:02:13.000
better "inner views" of Emacs objects...

00:02:13.000 --> 00:02:16.000
I'm going to Define what is an inner view,

00:02:16.000 --> 00:02:18.000
of course.

00:02:18.000 --> 00:02:20.000
The main trick is that we are going

00:02:20.000 --> 00:02:24.000
to use one of the ways of displaying

00:02:24.000 --> 00:02:29.000
internal objects, that is the `cl-print'

00:02:29.000 --> 00:02:32.000
family of functions, for example,

00:02:32.000 --> 00:02:35.000
`cl-prin1-to-string', and here are some

00:02:35.000 --> 00:02:37.000
examples of the kind of output that we

00:02:37.000 --> 00:02:38.000
are going to see...

00:02:38.000 --> 00:02:44.000
for example, if we run these two lines

00:02:44.000 --> 00:02:47.000
here the first line defines a function `foo'

00:02:47.000 --> 00:02:52.000
and the second line sets `o' to the

00:02:52.000 --> 00:02:54.000
internal view of the definition of `foo'.

00:02:54.000 --> 00:02:59.000
In older Emacses `o' would be just a

00:02:59.000 --> 00:03:02.000
list that looks... that would look very

00:03:02.000 --> 00:03:05.000
similar to this line here... but in newer

00:03:05.000 --> 00:03:09.000
Emacses the result of this - I mean, the

00:03:09.000 --> 00:03:12.000
the contents of `o' is this thing here,

00:03:12.000 --> 00:03:15.000
that looks quite different

00:03:15.000 --> 00:03:18.000
from this definition.

00:03:18.000 --> 00:03:21.000
So, in older Emacses

00:03:21.000 --> 00:03:25.000
the contents of the

00:03:25.000 --> 00:03:28.000
function cell of `o'...

00:03:28.000 --> 00:03:30.000
sorry, of the function cell of `foo',

00:03:30.000 --> 00:03:32.000
would be an "old-style lambda",

00:03:32.000 --> 00:03:35.000
that would be just a list like this...

00:03:35.000 --> 00:03:39.000
and in newer Emacses uh the contents of O would

00:03:39.000 --> 00:03:42.000
be a "vector-like lambda"... look for the

00:03:42.000 --> 00:03:44.000
square brackets here - this is a

00:03:44.000 --> 00:03:47.000
vector, but it is preceded by a hash sign.

00:03:47.000 --> 00:03:49.000
So this is what we call

00:03:49.000 --> 00:03:51.000
a "vector-like lambda",

00:03:51.000 --> 00:03:53.000
and vector-like lambas do not

00:03:53.000 --> 00:03:55.000
have a canonical printed representation -

00:03:55.000 --> 00:03:57.000
they have at least two semicanonical

00:03:57.000 --> 00:03:59.000
printed representations...

00:03:59.000 --> 00:04:01.000
The first semicanonical

00:04:01.000 --> 00:04:04.000
printed representation is this one, that is

00:04:04.000 --> 00:04:07.000
generated by a family of functions with

00:04:07.000 --> 00:04:09.000
names like `prin1'...

00:04:09.000 --> 00:04:13.000
and the second semicanonical printed

00:04:13.000 --> 00:04:17.000
representation is like this -

00:04:17.000 --> 00:04:20.000
it looks like a list...

00:04:20.000 --> 00:04:23.000
it looks somewhat like this definition

00:04:23.000 --> 00:04:27.000
of `foo' here, but it has this

00:04:27.000 --> 00:04:29.000
`:dynbind' symbol here...

00:04:29.000 --> 00:04:32.000
and it turns out that when we use

00:04:32.000 --> 00:04:35.000
the `cl-print' family of functions we can

00:04:35.000 --> 00:04:37.000
reconfigure how things are printed...

00:04:37.000 --> 00:04:40.000
and I'm going to show several interesting

00:04:40.000 --> 00:04:47.000
ways of reconfiguring how lambdas are printed,

00:04:47.000 --> 00:04:49.000
and one of the ways is going to

00:04:49.000 --> 00:04:52.000
be like this.

00:04:52.000 --> 00:04:56.000
We can also use the `cl-print'

00:04:56.000 --> 00:04:59.000
functions with my indentation tricks to

00:04:59.000 --> 00:05:04.000
to display how types, or classes, are

00:05:04.000 --> 00:05:07.000
viewed internally by Emacs, and this is a

00:05:07.000 --> 00:05:10.000
big example...

00:05:10.000 --> 00:05:14.000
This is what Emacs considers as being

00:05:14.000 --> 00:05:16.000
the definition of the type

00:05:16.000 --> 00:05:18.000
`cl-structure-class',

00:05:18.000 --> 00:05:21.000
class and it is this big thing here.

00:05:21.000 --> 00:05:24.000
I edited it very lightly...

00:05:24.000 --> 00:05:30.000
I just uh deleted some line breaks here.

00:05:30.000 --> 00:05:33.000
And another thing that I want to to

00:05:33.000 --> 00:05:35.000
explain is that Emacs

00:05:35.000 --> 00:05:37.000
has some help functions that

00:05:37.000 --> 00:05:39.000
I have never liked...

00:05:39.000 --> 00:05:41.000
for most people they are good enough,

00:05:41.000 --> 00:05:44.000
but for me they aren't... they...

00:05:44.000 --> 00:05:48.000
uh, well - I'm going to say

00:05:48.000 --> 00:05:50.000
more about this later...

00:05:50.000 --> 00:05:52.000
and, for example,

00:05:52.000 --> 00:05:54.000
if we want a description of what is

00:05:54.000 --> 00:05:58.000
this type here, that we just saw in

00:05:58.000 --> 00:06:00.000
its internal view here...

00:06:00.000 --> 00:06:02.000
we can run either `describe-type'

00:06:02.000 --> 00:06:04.000
or my variant of `describe-type',

00:06:04.000 --> 00:06:07.000
and we get a help buffer

00:06:07.000 --> 00:06:10.000
that looks like this, in which

00:06:10.000 --> 00:06:13.000
these blue things that are underlined

00:06:13.000 --> 00:06:15.000
are "buttons", in the classical sense...

00:06:15.000 --> 00:06:17.000
you can click on these buttons, or type

00:06:17.000 --> 00:06:19.000
RET on these buttons, and you will be

00:06:19.000 --> 00:06:22.000
taken to another help page, that is

00:06:22.000 --> 00:06:24.000
generated dynamically...

00:06:24.000 --> 00:06:28.000
and you can navigate back and forth...

00:06:28.000 --> 00:06:30.000
and well, whatever...

00:06:30.000 --> 00:06:33.000
and I'm going to explain my

00:06:33.000 --> 00:06:35.000
problems with these kinds of help buffers

00:06:35.000 --> 00:06:37.000
and what I'm trying to do to

00:06:37.000 --> 00:06:41.000
overcome these problems...

00:06:41.000 --> 00:06:43.000
One of my slogans in this video

00:06:43.000 --> 00:06:43.000
is going to be this one:

00:06:43.000 --> 00:06:45.000
"Anyone can learn Lisp in one day".

00:06:45.000 --> 00:06:49.000
this is a part of a bigger quote

00:06:49.000 --> 00:06:51.000
that I took from a keynote presentation

00:06:51.000 --> 00:06:54.000
by Abelson and Sussman, who

00:06:54.000 --> 00:06:58.000
are two dinosaurs of Computer Science...

00:06:58.000 --> 00:07:00.000
Here is the full quote:

00:07:00.000 --> 00:07:04.000
"Anyone can learn Lisp in one day -

00:07:04.000 --> 00:07:06.000
except that if they already know Fortran

00:07:06.000 --> 00:07:11.000
then it would take three days."

00:07:11.000 --> 00:07:24.000
This is a frame of the video...

00:07:24.000 --> 00:07:28.000
By the way I am going to to add

00:07:28.000 --> 00:07:32.000
this... "and if the person is starting

00:07:32.000 --> 00:07:34.000
with Doom Emacs then it would take 5 years."

00:07:34.000 --> 00:07:39.000
why? I'm going to explain why.

00:07:39.000 --> 00:07:43.000
This is how Emacs used to be.

00:07:43.000 --> 00:07:46.000
If we execute these two expressions here

00:07:46.000 --> 00:07:51.000
the first one... sorry, each symbol can

00:07:51.000 --> 00:07:53.000
have two "values",

00:07:53.000 --> 00:07:54.000
one is its "value as a variable"

00:07:54.000 --> 00:07:58.000
and another one is its "value as a function"...

00:07:58.000 --> 00:08:02.000
and if we run this we store 42

00:08:02.000 --> 00:08:07.000
in the "value cell" of the symbol `foo', and

00:08:07.000 --> 00:08:11.000
if we run this defun here it stores a

00:08:11.000 --> 00:08:14.000
certain anonymous function in the

00:08:14.000 --> 00:08:18.000
"function cell" of the symbol `foo'...

00:08:18.000 --> 00:08:22.000
and in Emacs, until some time ago

00:08:22.000 --> 00:08:27.000
if we did that and and if we ran

00:08:27.000 --> 00:08:30.000
this expression here the result

00:08:30.000 --> 00:08:31.000
would be 42,

00:08:31.000 --> 00:08:35.000
because of this line here, and if we

00:08:35.000 --> 00:08:37.000
ran this line here the result would be

00:08:37.000 --> 00:08:40.000
the anonymous function corresponding to

00:08:40.000 --> 00:08:41.000
this defun here...

00:08:41.000 --> 00:08:45.000
but now this has changed...

00:08:45.000 --> 00:08:48.000
the result of this thing here is this

00:08:48.000 --> 00:08:51.000
vector-like lambda here - but that doesn't

00:08:51.000 --> 00:08:54.000
matter much now...

00:08:54.000 --> 00:08:56.000
So, until some time ago

00:08:56.000 --> 00:08:58.000
if we did that and if we ran

00:08:58.000 --> 00:09:01.000
this expression here, (foo foo)...

00:09:01.000 --> 00:09:04.000
Emacs would do this: it would

00:09:04.000 --> 00:09:06.000
replace the first `foo' by this

00:09:06.000 --> 00:09:09.000
anonymous function here, it would replace

00:09:09.000 --> 00:09:11.000
the second `foo' by the value of `foo' as a

00:09:11.000 --> 00:09:13.000
variable, that is 42,

00:09:13.000 --> 00:09:16.000
and it would evaluate this, and the

00:09:16.000 --> 00:09:20.000
result would be 420.

00:09:20.000 --> 00:09:23.000
So, again, we used to have this slogan

00:09:23.000 --> 00:09:26.000
here, "anyone can learn Lisp in one day"...

00:09:26.000 --> 00:09:28.000
but now this is gone.

00:09:28.000 --> 00:09:30.000
Let me show... let me talk

00:09:30.000 --> 00:09:34.000
a bit more about why...

00:09:34.000 --> 00:09:36.000
the title of this slide is

00:09:36.000 --> 00:09:38.000
"Lambdas for beginners broken"...

00:09:38.000 --> 00:09:41.000
if we run this, as I've shown

00:09:41.000 --> 00:09:43.000
in the previous slide...

00:09:43.000 --> 00:09:45.000
in the old style, in old Emacses,

00:09:45.000 --> 00:09:47.000
the result of (symbol-function 'foo)

00:09:47.000 --> 00:09:49.000
would be this anonymous function here...

00:09:49.000 --> 00:09:54.000
and now we get this strange thing here.

00:09:54.000 --> 00:09:58.000
So, this is an "old-style lambda",

00:09:58.000 --> 00:10:02.000
this is a "vector-like lambda",

00:10:02.000 --> 00:10:05.000
and until the middle of 2024

00:10:05.000 --> 00:10:08.000
beginners could learn a lot of Lisp

00:10:08.000 --> 00:10:11.000
by thinking only in terms of

00:10:11.000 --> 00:10:13.000
objects like these...

00:10:13.000 --> 00:10:15.000
this is a function and this

00:10:15.000 --> 00:10:17.000
is an anonymous function, and

00:10:17.000 --> 00:10:20.000
they would learn how to draw cons cell

00:10:20.000 --> 00:10:23.000
diagrams like this thing here and this

00:10:23.000 --> 00:10:25.000
thing here...

00:10:25.000 --> 00:10:27.000
they would think on lists as

00:10:27.000 --> 00:10:29.000
being these trees here, and they

00:10:29.000 --> 00:10:32.000
would be able to understand a lot of

00:10:32.000 --> 00:10:35.000
Lisp just by thinking in these terms...

00:10:35.000 --> 00:10:39.000
and then vector-like lambdas started

00:10:39.000 --> 00:10:43.000
to appear in many places... and if we use

00:10:43.000 --> 00:10:46.000
"vector-like lambdas" in a wide sense,

00:10:46.000 --> 00:10:50.000
to mean all the new objects,

00:10:50.000 --> 00:10:54.000
these new objects, that are

00:10:54.000 --> 00:10:56.000
difficult to visualize... they also started

00:10:56.000 --> 00:10:58.000
to appear in many places.

00:10:58.000 --> 00:11:01.000
This is a continuation of the

00:11:01.000 --> 00:11:04.000
previous slide - this part here is a copy

00:11:04.000 --> 00:11:06.000
of things that were in the previous slide...

00:11:06.000 --> 00:11:12.000
before 2024 beginners could

00:11:12.000 --> 00:11:17.000
open black boxes like this...

00:11:17.000 --> 00:11:20.000
they could try to see what was in the

00:11:20.000 --> 00:11:24.000
function cell of the symbol `foo'...

00:11:24.000 --> 00:11:27.000
and they would see something elegant and

00:11:27.000 --> 00:11:29.000
mind-blowing... and they would start to love

00:11:29.000 --> 00:11:31.000
Lisp immediately.

00:11:31.000 --> 00:11:33.000
Now what they get - what they see -

00:11:33.000 --> 00:11:35.000
is a tiny part of a very complex structure

00:11:35.000 --> 00:11:39.000
that is very powerful but that is

00:11:39.000 --> 00:11:41.000
very difficult to understand...

00:11:41.000 --> 00:11:44.000
and now our beginners are overwhelmed

00:11:44.000 --> 00:11:46.000
instead of mind-blown.

00:11:46.000 --> 00:11:48.000
Note that I said "black box" here.

00:11:48.000 --> 00:11:52.000
Let me explain the term.

00:11:52.000 --> 00:11:57.000
We can open what's inside of `foo'...

00:11:57.000 --> 00:11:59.000
we can open `foo' to see the contents of

00:11:59.000 --> 00:12:02.000
the symbol `foo', and we can try to see

00:12:02.000 --> 00:12:06.000
what's in the function cell of the

00:12:06.000 --> 00:12:08.000
symbol `foo'...

00:12:08.000 --> 00:12:10.000
so we can open the box, but what we get

00:12:10.000 --> 00:12:13.000
is something very difficult to understand,

00:12:13.000 --> 00:12:17.000
and so I'm going to say that

00:12:17.000 --> 00:12:21.000
when this happens that box is black.

00:12:21.000 --> 00:12:23.000
It is not totally black - we can open open it -

00:12:23.000 --> 00:12:26.000
but we don't understand what is going on there,

00:12:26.000 --> 00:12:30.000
so we declare that that is black.

00:12:30.000 --> 00:12:33.000
And... when these things started to happen

00:12:33.000 --> 00:12:38.000
_I_ was overwhelmed -

00:12:38.000 --> 00:12:40.000
and in this video I'm going to pretend

00:12:40.000 --> 00:12:44.000
that I was not the only person

00:12:44.000 --> 00:12:46.000
that was overwhelmed

00:12:46.000 --> 00:12:50.000
by these new structures

00:12:50.000 --> 00:12:52.000
that are not so elegant

00:12:52.000 --> 00:12:54.000
as the ones that we had before.

00:12:54.000 --> 00:12:56.000
Anyway...

00:12:56.000 --> 00:20:38.000