Re: html rich text colored

On Mon, 20 Mar 2023 21:43:47 +1100, Ian Smith wrote:
> On 20 March 2023 4:53:49 pm AEDT, white-wolf <white-wolf@blues-softwares.net> wrote:
>  [...]
>  > who have no X now on this desktop computer for see html+Markdown ?
> 
> Enough people who this community respects enough not to disrupt?

It's not just about reading messages. It's also about replying
to them (and malformed HTML can be a hazard to proper quoting
and to preserving structure), and it's about message archives,
as it has been mentioned.

In my opinion, it is important to understand that a modern day
mail client that includes a HTML renderer has reached the level
of complexity that in ye olden times has been reserved for
operating systems. Resource allocation, permission control,
sandboxing, 3D accelleration access... this doesn't come for
free. Keeping a web-enabled MUA up to date, especially regarding
the ongoing discovery of new (and old) security vulnerabilities
requires a steady flow in maintenance. On the other hand, reading
text is less risky; "less" hasn't required an update for
decades. ;-)

And it gets worse: Because you want color in your HTML message,
_how_ do you want it? A trivial approach would be to use the
designated HTML attributes for <span> and <div>, and you could
also use <quote> for inlining quotes. Okay so far. But now
people (probably correctly) tell you that this is wrong, and
you should use CSS for that. Now you have enabled another whole
level of complexity. Modern people will then tell you that you're
doing it all wrong, and you should use JavaScript. Not pure,
of course, but with the lastest frameworks. Bam! A new challenge.
No Internet connection? Sorry, you cannot even read your stored
messages! ;-)

Of course, using HTML properly gives you power of structuring
if (!) you use semantic markup instead of microformatting,
i. e., "this is a heading" instead of "this is bigger text
with underline". However, people often don't use this power.
Mailing programs that generate HTML usually also don't.

As mentioned above, the simple act of replying to a message
and adding statements can distort the whole "HTML landscape"
of the original, therefore lead to buggy output to the list.
While HTML _could_ be used with its power of semantic markup,
HTML generators will typically mess up the whole thing.

And... how about you want to include actual HTML tags in your
text, for example, you have a question regarding HTML substitution
or elimination using sed or awk? You need to pay attention to
quoting now. Or use a <pre> environment? And also keep an
eye of the usage of other "special symbols" that might be
causing undesired effects when used in HTML. HTML isn't a
good tool to render code snippets without further effort.

With HTML, list participants would probably also use the
ability to structure text as they want to. Some use a triple
empty line for paragraphing, others a single line, and some
prefer indentation, which all is okay, as it adds an individual
touch to messages. Some users indent code snippets with tabs,
others with spaces. All this doesn't map 1:1 into HTML and
requires more work. And just think about "ASCII art" used when
someone illustrates their home network to provide context for
a question...

And don't be fooled about ISO-8859-1 / -15 / UTF-8 and HTML.
Depending on the tool used to generate the list message, this
can lead to big trouble. Encoding hell is waiting for you,
located at Schlatildefracterstrasse in Berlin! :-)

There are other uses of mailing lists and their archives: the
automated processing. Searching for things, creating categories,
using snippets in notes - all this can easily be achieved with
pure text. The great thing about text is that it can combine
regular text with code and ASCII art without requring anything
else than your favorite editor - it doesn't even have to be
the one used by your mailing program, because anything it needs
to do is generate text.





-- 
Polytropon
Magdeburg, Germany
Happy FreeBSD user since 4.0
Andra moi ennepe, Mousa, ...