Born-hidden knowls reborn

[Rob Beezer]

For the cognesceti. Born-hidden knowls, whether automatic (e.g. "hint") or
elected by your publisher (e.g. "theorem"), are now implemented with the HTML
"details" disclosure widget. Thanks to David Farmer, Andrew Scholer, and Alex
Jordan (and maybe people I'm missing) for their contributions to this. Website
samples have been updated. Includes an increment in the CSS and JS numbers
(@Alex - the WW JS file will need a symlink or however you handle that).

My favorite sort of upgrade - authors don't need to do *anything* to their source.

These samples are *not* the knowl-everything "beta" testing samples we've been
working with. So if you can get the CLI nightly tomorrow and rebuild *your*
project as HTML locally, there is a good chance you will see areas for
improvement that directly impact your work and the work of those who don't
follow this group slavishly.

One change to mention here. A variety of faux born-hidden knowls that lived in
files are no longer needed. This is the consequence of bulk of the 350 lines of
removed code (2.5% of the HTML conversion). But if you push your HTML to some
site for hosting, you might want to trash your "knowl/" directory at the site
before your next upload. The no-longer needed files look like
"knowl/*-hidden.html" and there are not too many of them. There will be massive
file-storage savings next, so you could wait a week or two in this if you like.

[Rob Beezer]

For the CLI folks. I should know the answer, and I could run a test (maybe),
but it is esier to ask. ;-)

When a publisher makes HTML output publicly available (deploying?) do the files
get refreshed, or do they all get deleted and then new versions uploaded? In
other words, as I fiddle with the directory of knowl files, do publishers need
to take any extra steps to make sure stale files aren't left around on hosts?

[Steven Clontz]

It depends: the contents of the deployed materials are ignored upon a new deployment. However, they still might be cached in the user's local output directory, unless a `--clean` build was run.

So in general, anyone acting as a publisher should run `pretext build TARGET_NAME --clean` in order to ensure cached files from previous builds aren't deployed. (Oscar can tell us if this is exposed in pretext-tools GUI.)

[Rob Beezer]

Thanks. Sounds like --clean will be just the thing. Next big change, maybe in a few days, will benefit greatly from it. And maybe I'll suggest some coordination with CLI releases. Not too fast, not too slow. We can discuss later.

Thanks,
Rob

[Oscar Levin]

Today I got an issue on the pretext-codespace repo from an author who thought something was broken because of these (much nicer) knowls.  So perhaps:

1. They still need some styling help.  I notice that my mouse doesn't change to a link-like pointer when hovering, and the knowl formatting is not being applied.  

2. Once this is updated, a post on -announce might be helpful.

[David W. Farmer]

> 1. They still need some styling help.  I notice that my mouse doesn't change to a link-like pointer when hovering, and the
> knowl formatting is not being applied.  

I get the pointer in the sample article. Is there another place I
should look?

[Rob Beezer]

I see the pointer on Checkpoint 4.4, but I do not see it for
Hint|Answer|Solution in Activity 4.3. Both Firefox and Chrome.

Rob

PS: Any chance of directing support to -support when it is off on some other
repo? ;-) Thanks for doing it manually, Oscar.

[David W. Farmer]

Thanks for the "pointer". Fixed.

> --
> You received this message because you are subscribed to the Google Groups
> "PreTeXt development" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to pretext-dev...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/pretext-dev/MTAwMDAzNi5iZWV6ZXI.1697510594%40quikprotect.
>
>

[Rob Beezer]

On 10/17/23 06:53, David W. Farmer wrote:
> Thanks for the "pointer".

Hah!

And for the record, I just checked "biblio/note" and "proof", which are similar
appendages and they are fine. DISCUSSION-LIKE would be similar (for open
problems) but we don't have such a thing in the sample article yet. Holler if I
should add one and I can test/post an update.

Rob

[David W. Farmer]

It is no problem^* for me to adjust the CSS if something comes up,
so no need to post an update.


----

* until we change the procedure so that I am no longer allowed to
edit the live CSS!

> --
> You received this message because you are subscribed to the Google Groups
> "PreTeXt development" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to pretext-dev...@googlegroups.com.
> To view this discussion on the web visit

> https://groups.google.com/d/msgid/pretext-dev/MTAwMDAwMC5iZWV6ZXI.1697562004%40quikprotect.
>
>