Allowing HTML in EPUBS

[Sarmat89]

Originally Posted by DNSB

What exactly do you mean by a "proper heading"?

A heading like you see in book, or Latex. The HTML 'h' elements were intended for manuals, where the headings are numbered and consist of a single line.

[Sarmat89]

Originally Posted by KevinH

Allowing html as a core media type, means that your html using epub will no longer work with the installed user base of epub3 e-readers, and furthermore any epub3 that used html can not be made to be backwards compatible with epub2 anymore.

Any serious reader is built around an HTML control, which can handle HTML. So the real HTML support can be implemented with trivial changes.

[DNSB]

Originally Posted by Sarmat89

Any serious reader is built around an HTML control, which can handle HTML. So the real HTML support can be implemented with trivial changes.

Can I get a kilogram of what you've been smoking for my next party? It sounds like some righteous shit.

As for multi-line chapter headers? In ePub2, they were pretty trivial. In ePub3, they are even easier.

[pdurrant]

Originally Posted by Sarmat89

A heading like you see in book, or Latex. The HTML 'h' elements were intended for manuals, where the headings are numbered and consist of a single line.

Umm.... no.

1. <hn> tags do not force text onto a single line
2. <hn> tags do not imply any kind of numbering on the headers

<hn> tags are just like paragraph tags really, except that most (all?) renderers choose to have them in a different default format to <p>

[Quoth]

Multiline headers in Word/LO Writer are trivial to do and automatically translate to
<h2 class="docx-heading-style">A heading<br />
Heading continued</h2>.

Headings were not for manuals. By default they are bolded, but the CSS of docx-heading-style can over-ride that if it uses an already bold/heavy font.

In many applications Shift Enter creates a soft carriage return which translated to <br />
Typing Enter alone ends a paragraph or heading and starts a new one in the default style or current style or the style defined to follow a style.

A soft carriage return never starts a new style/paragraph.

[RbnJrg]

Originally Posted by Sarmat89

A heading like you see in book, or Latex. The HTML 'h' elements were intended for manuals, where the headings are numbered and consist of a single line.

You seem to have a very limited idea of what's possible in ePub. But could you post a screenshot or picture of a heading that you claim can't be duplicated in ePub?

[KevinH]

Originally Posted by Sarmat89

Any serious reader is built around an HTML control, which can handle HTML. So the real HTML support can be implemented with trivial changes.

And then you add in the costs of rewriting the publishing industry's epub toolchain software. And then the you add in the costs of rolling out firmware updates to your epub3 e-readers already in place. And of course that breaks any that are no longer providing firmware updates. Any you keep ignoring the costs associated with breaking backwards compatibly with epub2 only e-readers which are still a big part of the market, especially in Europe, and all of that for what benefit?

You really do not understand epub publishing processes at all.

[Sarmat89]

Originally Posted by pdurrant

<hn> tags are just like paragraph tags really

That's the problem; the headings are not paragraphs. A common header has from two to 4 parts. Only in manuals they write the header and the number in a single paragraph.

There is no provision in HTML for headers consisting of multiple parts.

Originally Posted by Quoth

Multiline headers in Word/LO Writer are trivial to do
...
Headings were not for manuals.

They are from the original HTML specification, which was made for man pages. Also, only American-style technical manuals can be adequately formatted with <h> tags.

[Quoth]

Originally Posted by Sarmat89

That's the problem; the headings are not paragraphs. A common header has from two to 4 parts. Only in manuals they write the header and the number in a single paragraph.

There is no provision in HTML for headers consisting of multiple parts.


They are from the original HTML specification, which was made for man pages. Also, only American-style technical manuals can be adequately formatted with <h> tags.

Sorry, but this is nonsense.

Headings work just like paragraphs with <hn (where n is 1 to 10, but ebooks are best limited to any 3 levels) instead of <p.

Headers can have as many parts as you want.

Original HTML was not for "man" pages. Original HTML spec was about 1998 or 1999 and websites using it from about 1992. It had no CSS. The Mobi format was already obsolete in 2015 when Amazon bought Mobipocket and more so on first Kindle in 2017. It uses HTML3 and there is no difficultly simulating any paper book multi-line headings.

"American style manuals"? HTML wasn't even an American design. It was by a British guy in Switzerland.

Edit:
In fact now with CSS and HTML you can interchange h and p tags and the appearance is identical if font-weight is specified. The only differences are automatic ToC building using <h tags and semantic (maybe some screen readers).

[JSWolf]

Originally Posted by Quoth

Edit:
In fact now with CSS and HTML you can interchange h and p tags and the appearance is identical if font-weight is specified. The only differences are automatic ToC building using <h tags and semantic (maybe some screen readers).

But in two of yhe most popular ePub editors, h tags are used for automatic ToC generation. a p will not work for that.

[DiapDealer]

What html header tags might have been originally created to do could not be more irrelevant to what can be accomplished with them today. Silly Putty was created to clean wallpaper.

[slm]

Originally Posted by DiapDealer

Silly Putty was created to clean wallpaper.

Actually, it was Play-Doh,

[Sarmat89]

Originally Posted by Quoth

In fact now with CSS and HTML you can interchange h and p tags and the appearance is identical

That's what I'm talking about: It is 2025, but you still have to build stupid CSS rules and write several paragraphs and ton of attributes manually just to make a simple chapter heading and make it to the TOC automatically. That's ridiculous.

[KevinH]

Originally Posted by Sarmat89

That's what I'm talking about: It is 2025, but you still have to build stupid CSS rules and write several paragraphs and ton of attributes manually just to make a simple chapter heading and make it to the TOC automatically. That's ridiculous.

It seems you do not understand CSS or html. So why are you attempting to post this here and derail this thread?

Moderators, perhaps we can move this discussion of html, css, and headers to its own thread? It is not relevant to this thread's topic.

[Quoth]

Originally Posted by Sarmat89

That's what I'm talking about: It is 2025, but you still have to build stupid CSS rules and write several paragraphs and ton of attributes manually just to make a simple chapter heading and make it to the TOC automatically. That's ridiculous.

CSS replaces inline styles. That works on HTML3 without CSS. It's much easier to use and edit CSS as the multiple Style strings need repeat on each instance and CSS replaces three to eight styles with one sensibly named class. The docx styles from LO Writer or MS Word usually map 1:1 to CSS classes for epub but have to be styles on mobi as it's HTML3. Thankfully old mobi is nearl gone now.

There are many ways to automatically do the ToC and at least 2 work with "headings" done using <p tags, though sematically that's wrong, even though it looks perfect.


You don't have to build anything, unless for some mad reason you are writing an entire novel directly in HTML. I've not done even websites that way for nearly 30 years. I do sometimes edit epubs or web content, but I'd not dream of writing them from scratch.

Do you use "Styles" in MS Word / LO Writer (=CSS), or use direct formatting (= HTML3 styles)?