[latexrefman] split the .html's also?

Hefferon, Jim S. jhefferon at smcvt.edu
Fri Jul 6 02:06:35 CEST 2018


> My initial thought is that I wouldn't bother trying to include it in the
CTAN upload (-> TL). CTAN is not really set up to serve multifile .html
documents, as far as I've seen.

I had in mind a .zip, so a person could dump it in their web tree.

> Sheldon Green's

That's not the person who did latex2e-reference, right?  I had a look at the ctan-ann archives as far back as CTAN had them (Aug 2001) and did not find an uploader.

> I somehow didn't notice you had previously created a .css.
> Sorry, but I'm mostly unenthused.

> - code: should already be getting done in monospace, so why?

When I wrote to the Texinfo folks, I got a little help but not what I wanted, that @math{.. at code aaa}..} gives me monospace that is unslanted.

https://lists.gnu.org/archive/html/bug-texinfo/2018-03/msg00032.html

> - h3 color: no great objection, but I don't see the win.

I like the tiny bit of color.  Not married to it, of course, but it seems like an improvement, to me.

> If we are doing css, the two bits I've actually found useful (because
browsers and/or standards committees are stupid) and included on tug.org

OK.

But the real issue is getting something to come out in the footer.  I currently have a link to the tug.org page for the project, but it isn't in the footer, and the footer only appears on internal nodes of the html tree anyway.

The first of these has a footer while the second does not.

http://joshua.smcvt.edu/latex2e/Command-line.html#Command-line

http://joshua.smcvt.edu/latex2e/Recovering-from-errors.html#Recovering-from-errors

Anyway, I think I am going to do something else for a while.  :-)

Jim

----------
https://i.redd.it/cmsrqy1olq611.jpg

________________________________________
From: latexrefman <latexrefman-bounces at tug.org> on behalf of Karl Berry <karl at freefriends.org>
Sent: Thursday, July 5, 2018 6:24:21 PM
To: latexrefman at tug.org
Subject: Re: [latexrefman] split the .html's also?

    cannot ^F to search the entire doc.

That, along with general simplicity, was the big reason I thought
originally it would be better to have only the unsplit version. However,
I see that the .html file has more than doubled in size thanks largely
to your efforts :), so maybe now it is worth providing the split html
too.

    Google favors sites that load fast,

I have nothing against making the split version and linking to it from
our web site.

My initial thought is that I wouldn't bother trying to include it in the
CTAN upload (-> TL). CTAN is not really set up to serve multifile .html
documents, as far as I've seen. But I have no strong feelings either
way.

    people who are now hosting a collection of .html's might go for a
    replacement collection over one big page (a person can wish).

You dreamer you :). We could explicitly write to the people who are
serving Sheldon Green's pages now. Not fun, but might garner a change.

We could make the split html available via rsync, or as a .tar.gz, or
whatever seems most convenient. (Independent of CTAN, though maybe this
is a reason to include it in the CTAN upload.)

    I also did a very minimal amount of CSS styling.

I somehow didn't notice you had previously created a .css.
Sorry, but I'm mostly unenthused.

- code: should already be getting done in monospace, so why?

- h3 color: no great objection, but I don't see the win.

- a:{link,visited} color: I strongly think it is better not to tinker
with these. It is too easy to end up either with link text that blends
into the background, or is vibrant with it, or visited indistinguishable
from unvisited, etc. It happens all the time on web sites nowadays and
it drives me crazy. (I hate that we do it on tug.org, not done by me,
but seems too late to change it there.)

- hr: the gradient seems too fancy to me. Personally I'd just leave it
alone.

- div.referencetitle: ok, though again, I don't see a big win over
leaving it alone.

- div.referenceinfo: I don't understand what this is for, but in any
case, pixel dimensions aren't be a good idea in my experience.


If we are doing css, the two bits I've actually found useful (because
browsers and/or standards committees are stupid) and included on tug.org
are:

------
BODY {
  margin-top: 1em;
  margin-left: 1em;   /* auto results in two-digit <ol> lost off left */
  margin-right: 1em;
  margin-bottom: 1em;

  /* the idea is to use the whole window, unless it is ridiculously
     wide, probably with too-small fonts, too.  */
  max-width: 64em;
}
------

Thanks,
Karl



More information about the latexrefman mailing list