On Sun, Aug 20, 2023 at 10:44 PM John Levine <johnl@xxxxxxxxx> wrote:
It appears that John C Klensin <john-ietf@xxxxxxx> said:
>Now that particular argument is interesting, of only
>historically. The one piece of the puzzle the IETF mostly did
>invent was something called "Rich Text". IIR (it has been years
>since I looked at the spec and don't have time today) it would
>have given you bold and italics and, if wanted, section
>headings. I don't remember what it would have done about
>proportional fonts, but the want/need for that is probably more
>controversial than your statement above indicates. Nor do I
>know whether what Microsoft and others call Rich Text today is
>the same thing or compatible. ...
That's RFC 1896. It looks a lot like a subset of HTML. It assumes
proportional fonts unless you say <fixed>. There's no section
headings, though you can fake it with <bigger> and <bold>.
I think that what happened is that by the mid 1990s computer systems
all had shared libraries, and once you have a web browser, one of
those libraries is an HTML renderer. It's a lot easier to feed a
text/html part into that existing library than to write a separate
stripped down text/enriched renderer. Going the other way, there are
HTML editors for web tools which your MUA could just borrow.
I expect this isn't a very popular viewpoint in this crowd, but if you
can assume all of your readers can render HTML, you can do some nice
stuff in your mail. I often get mail from clients where they have a
question about what I wrote so they put the question at the top and
highlight the quoted text of interest in yellow. That is about two
clicks in an HTML editor.
Mesh Everything has its own document markup, it is XML but not HTML schema.
In order to render it, the EveryDoc markup has to be translated to HTML so it can render in Blazor. While the markup is essentially a subset HTML, there are security and accessibility concerns.
I do wonder if I could get away with my abbreviated HTML markup. Most tags have zero attributes or one attribute that is the main point. Paragraph oriented tags cannot contain other paragraphs so the close tags are superfluous. So why not be able to write
<h1="main">Main Heading
<p>Paragraph text with an <a="http://example.com/">anchor</a> inside.
Now in point of fact, this is getting back to what HTML/1.0 looked like. The main reason not to do this is that there is already an XML parser in the general toolset. The main reason to do it is so that implementations don't get lazy and just pipe the output to the renderer without filtering out CSS, _javascript_ etc. etc.
