Have you, or anybody else to your knowledge, authored a non-trivial 100% hand-made website with multiple pages, one where all HTML code is
actually typed in the editor, rather than generated by any sort of
automation tool or template engine?

The Kiss Linux website seems close to it:

<https://kisslinux.org>

but they still have some minimal custom templating:

<https://github.com/kisslinux/kisslinux.github.io/tree/master/site>
--- Synchronet 3.19b-Linux NewsLink 1.113

Anton Shepelev <ant@tilde.culb> writes:

Have you, or anybody else to your knowledge, authored a non-trivial
100% hand-made website with multiple pages, one where all HTML code is actually typed in the editor, rather than generated by any sort of
automation tool or template engine?

Sure ... please don't remind me ... too late ... that's a mean task if
you mainly write about tech stuff with code examples and lots of other <pre>...</pre> stuff. My torture goes back even to the days when HTML
really was young and we had to emulate tables via <pre>.

I won't go back!

Now I'm really happy that I can do it in emacs/org-babel and even
execute the example code inline before exporting that to other formats including HTML. That really makes my life easier.

I won't go back!

A stripped down example:

<https://yeti.tilde.institute/plain/> (currently has cert problems)
<http://yeti.freeshell.org/plain/> (always http without the "s")

I won't go back!
--
2. Hitchhiker 17: (110) "Careful with that hammer, sir," he said.
--- Synchronet 3.19b-Linux NewsLink 1.113

yeti <yeti@tilde.institute> wrote:

Anton Shepelev <ant@tilde.culb> writes:

Have you, or anybody else to your knowledge, authored a non-trivial
100% hand-made website with multiple pages, one where all HTML code is
actually typed in the editor, rather than generated by any sort of
automation tool or template engine?

Sure ... please don't remind me ... too late ... that's a mean task if
you mainly write about tech stuff with code examples and lots of other <pre>...</pre> stuff.

This part does not seem too difficult, markup in such articles being
tolerably sparse. A friend of mine, for example, writes some good HTML:

<http://inversed.ru/InvMem.htm>

My misginvings, however, are about the reuse of repetitive HTML
fragments, e.g. for header and footer. That nearly inevitable task
already requires some mechanism of macro-substitution and therefore
prevents the manual writing of final .html files.

My torture goes back even to the days when HTML really was young and
we had to emulate tables via <pre>.

HTML 2.0, I think. Right. Or you could format those tables with
*roff's tbl:

<https://wolfram.schneider.org/bsd/7thEdManVol2/tbl/tbl.pdf>

Now I'm really happy that I can do it in emacs/org-babel and even
execute the example code inline before exporting that to other formats including HTML. That really makes my life easier.

Emacs is so vast a universe that I never plucked courage to plunge into
it :-)

[...]

<https://yeti.tilde.institute/plain/> (currently has cert problems)
<http://yeti.freeshell.org/plain/> (always http without the "s")

I won't go back!

Are they examples of what you did using org-babel? They look pretty
neat. What languages did you use in org-babel to produce those pages?
May I have a look at the org-babel source of that page?
--- Synchronet 3.19b-Linux NewsLink 1.113

Anton Shepelev <ant@tilde.culb> writes:

Are they examples of what you did using org-babel? They look pretty
neat. What languages did you use in org-babel to produce those pages?
May I have a look at the org-babel source of that page?

The main page's source:

<http://yeti.freeshell.org/plain/index.org>

And some lines about the other ingredients:

<http://yeti.freeshell.org/plain/under-the-hood.htmlhttp://yeti.freeshell.org/plain/under-the-hood.html>
--
"Emacs" rhymes with "Yessssss!". \o/ -- 20240328/yeti
--- Synchronet 3.19b-Linux NewsLink 1.113

Anton Shepelev <ant@tilde.culb> writes:

Are they examples of what you did using org-babel? They look pretty
neat. What languages did you use in org-babel to produce those pages?
May I have a look at the org-babel source of that page?

The main page's source:

<http://yeti.freeshell.org/plain/index.org>

And some lines about the other ingredients:

<http://yeti.freeshell.org/plain/under-the-hood.html>
--
"Emacs" rhymes with "Yessssss!". \o/ -- 20240328/yeti
--- Synchronet 3.19b-Linux NewsLink 1.113

On Thu, 12 Sep 2024, Anton Shepelev wrote:

Have you, or anybody else to your knowledge,
authored a non-trivial
100% hand-made website with multiple pages,
one where all HTML code
is actually typed in the editor

https://tilde.club/~xwindows/
http://tilde.club/~xwindows/
gopher://tilde.club/1/~xwindows/

All the pages [1] you are seeing there
at the time of this writing,
--whether HTTP(S) or Gopher--
are authored using regular text editor [2]
with no automation whatsoever
beyond basic text editor functions
and basic filesystem operations
(i.e. cut/copy/paste).

If you look at the source code
of the HTTP(S) version,
you would see that I have written them
in a very opinionated way as well. [3]

These are by no means
the only group of sites that I hand-author:
they are only just ones I could share in this circle.
As I treat web authoring as a craftwork,
most of my websites
--including ones I wrote for my software--
are written and maintained this way;
and needless to say
that I personally make sure
that they are usable
down to even the lowest common denominator
HTML-only [4] text mode browser
that one could find in this millennium as well.

Regards,
~xwindows


[1]
Including a pair of Atom feeds
provided on the HTTP(S) version as well.
(One for HTTP,
one for HTTPS)

[2]
With no auto-complete functionality of any sort.

[3]
Apart from it being intentionally designed
in legacy-first fashion
(HTML 4.01, CSS 2.1+, no client-side scripts),
it is also typed up using semantic linefeed
a.k.a. ventilated prose technique,
just like this netnews post you are reading: https://rhodesmill.org/brandon/2012/one-sentence-per-line/ https://vanemden.wordpress.com/2009/01/01/ventilated-prose/

[4]
No plugins,
no scripts,
no styles,
no images/audios/videos.
--
Contains ventilated prose,
manually typed up in RFC 2646
"text/plain; format=flowed" whitespace encoding;
might read like poetry in some newsreaders.
--- Synchronet 3.19b-Linux NewsLink 1.113

xwindows <xwindows@tilde.club> wrote:

On Thu, 12 Sep 2024, Anton Shepelev wrote:

Have you, or anybody else to your knowledge,
authored a non-trivial
100% hand-made website with multiple pages,
one where all HTML code
is actually typed in the editor

https://tilde.club/~xwindows/
http://tilde.club/~xwindows/
gopher://tilde.club/1/~xwindows/

Thanks. I don't use Gopher (yet), but I have already seen your club
site. I did not linger there, however, lookng for hard, text documents
rather than comics and image gallery.

All the pages [1] you are seeing there
at the time of this writing,
--whether HTTP(S) or Gopher--
are authored using regular text editor [2]
with no automation whatsoever
beyond basic text editor functions
and basic filesystem operations
(i.e. cut/copy/paste).

Did you manually copy-paste the header and footer for each .html file, including the <head> element and the variable part at the top with a
sort of a "path" to the current page from "root" of your website?

These are by no means
the only group of sites that I hand-author:
they are only just ones I could share in this circle.
As I treat web authoring as a craftwork,

I am glad for you, for this approach makes writing a joy rather than a
chore: typing every tag and every sentence with deliberation.

most of my websites
--including ones I wrote for my software--
are written and maintained this way;
and needless to say
that I personally make sure
that they are usable
down to even the lowest common denominator
HTML-only [4] text mode browser
that one could find in this millennium as well.

Modern text-mode browsers are pretty advanced, perhaps even more so than
Dillo.

[2]
With no auto-complete functionality of any sort.

I dislike autocomplete for coding, and hate so-called templates and
snippets. I like to write my code myself.

[3]
Apart from it being intentionally designed
in legacy-first fashion
(HTML 4.01, CSS 2.1+, no client-side scripts),

Your source looks very good to me. I have started a web-site in HTML
2.0, but the major problem is that I have not much put there so far, and
now I seem to need some minimal automation for header and footer.

it is also typed up using semantic linefeed
a.k.a. ventilated prose technique,
just like this netnews post you are reading:

Tin is blissfully ignorant of format=flowed;
and I,
as an erstwhile Groff user,
have no problem with it.

https://rhodesmill.org/brandon/2012/one-sentence-per-line/

So, Kernigan, and after him Brandon, advocate semantic linefeed for
/internal code/, on the ground of its convenience for editing and
version control, whereas you seem to like reading it too, don't you?

https://vanemden.wordpress.com/2009/01/01/ventilated-prose/

That's interesting: the taking advantage of the two-dimensional nature
of the media. Perhaps I will read Fuller -- thank you.

[4]
No plugins,
no scripts,

These are sources of such colossal bloat that low-grade computers such as
RPi struggle and sag under the weight of it. So-called responsive design
is actually non-responsive: anti-human, anti-software (how many indy browsers?), and anti-hardware.

no styles,

Of course: HTML is a language for structural markup, and the browser
should render bare HTML nicely, according to a sensible default style,
or to the one the user has configured. Modern browsers generally do not
support user styles, but special plugins (e.g. /Stylus/ for Chrome and Filrefox) supply that need.

no images/audios/videos.

They are cruelly abused these days, when people blog on Youtube, and
publish video instructions about software, when a text document would
have been more efficient.
--- Synchronet 3.19b-Linux NewsLink 1.113