Re: Proposal to include the Ef themes collection into Emacs core

[Protesilaos Stavrou <info@protesilaos.com>]

Good day Zoltan and Philip,

> From: Philip Kaludercic <philipk@posteo.net>
> Date: Tue,  1 Aug 2023 08:22:43 +0000
>
> Zoltan Kiraly <zolikydev@gmail.com> writes:
>
>> Dear Emacs Development Team,
>> I'm writing to bring your attention to an exceptional theme collection
>> named "Ef Themes," created by Protesilaos Stavrou, the same author
>> behind the Modus themes.
>>
>> Considering how well-received the Modus themes were and the positive
>> response they received from the community, I'm confident that
>> incorporating the "Ef Themes" collection into Emacs core would be of
>> benefit to users. These themes offer an excellent choice for
>> individuals who desire a more colorful Emacs environment while
>> ensuring good readability.
>
> My main concern is that there are a lot of themes, which might be
> overwhelming if they are all provided by default.

I understand this point.  There are lots of Ef themes as well.

> The other question is what the concrete use is of having them bundled
> in by default.

I don't have a strong opinion about this.  I guess the advantage is that
it is easier for new users to discover more themes.  The same argument
can be made for all the non-essential---but still useful---parts that
are built into Emacs.

> I am not sure what the argumentation was when adding the Modus Themes
> to the core, but I have to admit that I was not really a fan of adding
> the new Modus Themes (-tinted, -tritanopia, -deuteranopia) to the core
> as well, because of the above mentioned point of having too much
> provided by default, which can cause choice paralysis.

I asked before adding the new Modus variants.  From my side, I would
have added the new variants to the collection regardless.  It was just a
matter of whether core Emacs would have the same version as my Git
repository which, I believe, is the norm.

Generally though, the feeling of overwhelming options is not going away
by reducing the number of themes (which is the right number, anyway?).
Assuming this is the goal, concerted effort is needed across many parts
of Emacs.  For example, icomplete.el defines four minor modes and has to
explain to new users what Fido is all about.  I personally do not mind
this state of affairs.

> Perhaps a compromise could be to limit the number of themes that are
> added to three or four of the most popular ones.

Maybe this works for other themes.  Though when it comes to deuteranopia
or tritanopia popularity will not be a reliable criterion.

At any rate, we can take a look at the latest Emacs survey where there
are many entries for the Ef themes that were not consolidated in a
single bar: <https://emacssurvey.org/results/3425413930>.  At the time
of the survey, the Ef themes were only a few months old.

[ I don't read too much into the survey data, though it is a point about
  "popularity". ]

>> What makes the EF themes stand out is the awesome documentation,
>> covering everything a user needs to know, and the fact that it is
>> constantly updated and improved.
>
> IIRC EF Themes had fewer customisation options than Modus Themes, right?
> Is this something that would interest the "average" user who, I would
> guess usually is not that invested as to read the manual of a theme?

Yes, the Ef themes have fewer options than the Modus themes.  They still
are customisable and the manual explains all the details.  On the
flip-side, the Ef themes may be less confusing to the kind of user you
allude to here, because they look okay out-of-the-box, plus they cover a
wide range of preferences.

>> I kindly request you to consider evaluating the "Ef Themes" collection
>> for potential integration into Emacs core. By embracing this
>> collection, we can offer users a wider array of visually appealing
>> options while upholding the high standards of Emacs.
>
> On a tangent, what I was planning to work on for Emacs 30 was to add
> more "semantic" faces (along the lines of highlight, match, success,
> ...).  I had previously discussed the topic with Protesilaos (added in
> the CC's), and I think his work on various themes is useful in
> recognising common patterns in faces being defined in third-party faces
> that make creating themes so cumbersome (because of all the explicit
> support that has to be added for all the various known packages).  I
> reckon that ideally, a theme shouldn't have to concern itself with
> specific packages.

I agree and am happy to contribute towards that end.  If you already
have something, I am ready to contribute.  Otherwise, we can discuss it
whenever you want.

All the best,
Protesilaos (or simply "Prot")

-- 
Protesilaos Stavrou
https://protesilaos.com