A tag that's new to me <div/>

[DyckBook]

I've recently run across this self closing tag in a few books, but it looks like it should actually be </div> and not <div/>. In fact the check book tool throws an error, so clearly it's not effective in closing the <div> tag.

Is this to be used in some specific situation? Can you cast any light?

[kovidgoyal]

<div/> is the same as <div></div>

[phossler]

Originally Posted by kovidgoyal

<div/> is the same as <div></div>

I can see <br/> as a self closing tag being useful, how (why) would anyone use something the same as <div></div>, i.e. nothing being grouped?

Just curious

[Tex2002ans]

Originally Posted by phossler

I can see <br/> as a self closing tag being useful, how (why) would anyone use something the same as <div></div>, i.e. nothing being grouped?

Just curious

I would never use it, but perhaps they could use it for a scenebreak.

You can use CSS's ::before + ::after + content to insert stuff like images, asterisks, or have certain text be spoken out loud during Text-to-Speech:

https://codersblock.com/blog/diving-...eudo-elements/

[thiago.eec]

Originally Posted by Tex2002ans

I would never use it, but perhaps they could use it for a scenebreak.

You can use CSS's ::before + ::after + content to insert stuff like images, asterisks, or have certain text be spoken out loud during Text-to-Speech:

https://codersblock.com/blog/diving-...eudo-elements/

For a scene break I think the correct tag would be <hr/>.
I use <div/> for separators (a line below a chapter title, for example). This is just decorative and you should not use a <hr/> tag, since it means a thematic break in HTML5.

[hobnail]

Many tags that you see as a begin and end pair can also be given as the singular with the slash at the end. For example, I tried all of these and none gave an error with EpubCheck.

<p/>
<em/>
<b/>
<span/>
<body/>

As for why, I agree with Tex2002ans, you could use them with a class and a css combinator. (I'd be interested to see useful examples with the em and b singular tags because I can't think of any at the moment.)

Not related to this, but speaking of hr tags and combinators, in my stylesheet.css I have

hr + p {
  text-indent: 0;
}

So that paragraphs after an hr aren't indented. All of my p tags are naked/classless.

[kovidgoyal]

Originally Posted by phossler

I can see <br/> as a self closing tag being useful, how (why) would anyone use something the same as <div></div>, i.e. nothing being grouped?

Just curious

There is no how or why In EPUB you dont have HTMl you have XHTML which is XML. In XML *any* tag can be self closing, regardless of whether it is useful or not.

[Tex2002ans]

Originally Posted by thiago.eec

For a scene break I think the correct tag would be <hr/>.
I use <div/> for separators (a line below a chapter title, for example). This is just decorative and you should not use a <hr/> tag, since it means a thematic break in HTML5.

Attachment 190182

Yes, I wrote a lot about this. See the discussion in:

• 2020: "<i>, <em> or <span> for italics ?"

especially posts #128 and beyond. Me + Hitch went breaking down pros/cons of "<hr/> as scenebreaks".

Personally, I just use simple, centered asterisks. Not relying on CSS3 or complicated overrides (these will fail across many older readers/devices).

Side Note: The one nice thing about <hr/> in HTML5 is it generates a new <section> in the Accessibility Tree. I described some of that in post #148 of that thread. This would allow a screen reader to easily jump between scenes (or sections of an article) without the need for extra markup.

Originally Posted by kovidgoyal

There is no how or why In EPUB you dont have HTMl you have XHTML which is XML. In XML *any* tag can be self closing, regardless of whether it is useful or not.

EPUB3 is HTML5... and in HTML, you can do all this really funky/obscure stuff.

Like I remember running across a Sigil "bug" with:

<p">

(It was a bad search/replace I did.)

Sigil would completely crash with Reports, Spellcheck Lists, etc., but other things like Preview worked fine...

KevinH dug into this, and turns out, it's actually "valid" (but VERY horrible) HTML5. I convinced KevinH to match Calibre's parsing of this, by changing it into an underscore:

<p_>

so at least things would still work. Then epubcheck can warn you about non-standard tags at a later date.

Note: EPUB2 is XHTML1.1, and EPUB3... many tools like Sigil/Calibre conform to XHTML5—accepting HTML5's new tags, but still enforcing XHTML's stricter rules... like always closing your tags and using quotes around your attributes.

This is just good practice, and I have no idea why you'd NOT want to use XHTML5.