Prime Marks Question

[Quoth]

Most word processors not only get leading apostrophes wrong (like using ‘85 for “back in ’85”) but also use double closing quote and single closing quote on measurements like feet and inches or minutes and seconds (time or angular).
5’ 3” is wrong.
5' 3" is better.
5′ 3″ is correct. The Prime and Double Prime marks.
You can fake the marks by italic straight quotes: 5' 3".


Done in Georgia and larger so to show it.

Obviously if you embed a font and the ereader device or app supports it you can use the professional Prime marks (not related to Prime Numbers).

But if you don't embed a font or it's an old ereader? What happens if you you use real prime marks?

Should we always fake it with italics and straight quotes and then the fall back is " and ' rather than [] and [] or space?

No issue on paper. Use real prime marks. The " and ' are only for programming languages, in English. A throw back to typewriters as actually is the underline. IMO real ebooks and paper books should never use the underline except for URLs, they meant "typeset this in italic".

[Jellby]

Originally Posted by Quoth

But if you don't embed a font or it's an old ereader? What happens if you you use real prime marks?

I guess the same as with any other character that's not available in the font: you'll get a square, or a question mark, or nothing.

Should we always fake it with italics and straight quotes and then the fall back is " and ' rather than [] and [] or space?

No, we should use the proper characters and blame the reader (the device, not the person) if they don't work.

IMO real ebooks and paper books should never use the underline except for URLs, they meant "typeset this in italic".

Underlines can be used for other stylistic purposes. But I wouldn't use them for URLs, especially not in paper books.

[Quoth]

I'd not underline a URL in a paperbook, you can't tap/click it. It badly reduces readibility The Underline is certainly a typewriter thing and should be rarely if ever used in fiction. There might be some use for it, but I can't think of it.

"No, we should use the proper characters and blame the reader (the device, not the person) if they don't work."

That's not useful. The reader (the human) must not be penalised when they buy a book due to limitations of the reader hardware/software.

I guess I'll test a Kindle DXG and Palm PDA. Maybe a Binatone.

What I meant was does anyone know what devices don't support Prime marks?

[DNSB]

Originally Posted by Quoth

What I meant was does anyone know what devices don't support Prime marks?

I would suspect that it would be a matter of whether a font supports prime marks rather than a device. On a Kobo using a kepub, the renderer will look in other fonts for a non-supported glyph but an epub is limited to the font in use.

[phillipgessert]

I could see underlining a book's title if it appears within an italicized blockquote. I'm not dead sure whether the usual approach of reverting to roman text is suitable in that context.

ETA, my vote is to use proper primes as it seems impossible to account for every font. By which I mean a user using a very flexible e-reader app could hypothetically choose a
homemade font that doesn’t even include proper curly quotes. So, being impossible to entirely avoid an ugly fallback, better to just code the intended.

[Quoth]

Finally I got round to testing.
The proper prime marks, rather than italic ' and " giving ' and " do work on old mobi only Kindles. Not a problem on epubs or azw, of course. I've not tested a pre-epub Sony.
To enter the the real Prime and Double Prime characters on Windows if you have a separate number pad, then press Alt+8242 (press and hold the Alt key while you type 8242) for prime, or Alt+8243 for double prime. On MacOS either use the Character viewer or add the International Unicode Hex Keyboard and switch to it to enter 2032 for prime or 2033 for double prime in Unicode Hex. Windows and Linux have similar utilities.
In MS Word go Insert > Symbol. Enter 2032 for prime or 2033 for double prime in Unicode Hex.
In LO Writer, on any platform, go Insert > Special Character and put 2032 for ′ or 2033 for ″ in the Hexadecimal U box. Click on Add to Favourites and then the symbol.
The prime characters can only be inserted if they are actually in the character set. Actual symbols: ′ and ″. The relative weight will depend on the font family. They are a little bolder than italic ' and " in Georgia and quite light on the normal serif font on the Kindle DXG. I didn't try them in Courier, on the DXG, which does seem to render as a font.

I'm sure I've seen Sans on an unmodified DXG or older Kindle, but currently I've forgotten the recipe.

[Tex2002ans]

Originally Posted by Quoth

But if you don't embed a font or it's an old ereader? What happens if you you use real prime marks?

I wrote in detail about Prime + Double Prime characters just a few months ago:

My response in Diap's "PunctuationSmarten Sigil plugin" thread.

Also see mine + Jellby's thorough discussion on Primes (and various other typographical questions) in:

"Punctuation rules of thumb?"

Some old readers, like my original Nook, had issues with the actual prime characters missing in fonts... but honestly, I currently recommend using the proper primes no matter what (for TTS and Accessibility reasons).

Originally Posted by Jellby

No, we should use the proper characters and blame the reader (the device, not the person) if they don't work.

I agree completely (as usual).

Originally Posted by Quoth

A throw back to typewriters as actually is the underline. IMO real ebooks and paper books should never use the underline except for URLs, they meant "typeset this in italic".

Agreed. The only reason underline was used was because typewriters didn't have italics fonts.

In modern documents/typography, underline should be relegated to URLs only. (And as Jellby said, in Print, no underline needed.)

In the case of books/documents that originally were typeset using underline, but are now digitized/modernized, this is what I've done:

Original:

<p>This is an <u>Example Book Title</u> from a typewriter.</p>

Alternate #1 (preferred)

You digitize what the original author intended to write:

<p>This is an <i>Example Book Title</i> from a typewriter.</p>

Alternate #2

You mark the "original visuals", but override it with modern typography:

<p>This is an <span class="underline">Example Book Title</span> from a typewriter.</p>

span.underline {
	font-style: italic;
}

If for some reason you needed to duplicate the original look, this allows you to easily change to:

span.underline {
	text-decoration: underline;
}

(Another alternate could be <i class="underline"> with a "font-style: normal".)

Side Note: This reminds me of a book I digitized back in 2014. It was written on a typewriter, and the author used bold instead of italics:



Clearly, you can tell the bold text was meant to be italics, and the (50) markup was meant to be footnotes.

Modernizing and retypesetting the book, I updated to this:



The new print/ebook doesn't have an underline in sight.

(The original also had ASCII art diagrams/equations... which I recreated + vectorized.)

[Quoth]

I've decided to replace ' and " with prime marks on measurements as it seems very few ereaders don't support it and unless you switch off Smart Quotes it can mess them up. I'm considering turning off SmartQuotes as I can easily do AltGr-v, AltGr-b, AltGr-V, and AltGr-B for “ ” ‘ ’ and have the primes saved in Insert Custom Character. I may add them later to Compose, though compose ' and compose " are in use.

SmartQuotes is actually dumb and broken. '89 gives ‘89 instead of ’89. Also on LO it does “ instead of ” after — or …!


I wonder does the old Nook without them work if the epub has an embedded font like Georgia that does have them?

[hobnail]

Originally Posted by Quoth

The Underline is certainly a typewriter thing and should be rarely if ever used in fiction. There might be some use for it, but I can't think of it.

In old novels, like the ones on Project Gutenberg, there are often letters that the characters write to each other. Sometimes the book even says that the letter was handwritten but in the printed book they'll change fonts within the letter; e.g., using small caps for the writer's signature. Whenever I redo the formatting for a PG book I set the letters in italics since it's vaguely sort of handwritten looking (and I don't embed fonts). Whenever the PG version has italics in the body of the letter (not the address or whatever) I switch to underline.

[Tex2002ans]

Originally Posted by Quoth

I wonder does the old Nook without them work if the epub has an embedded font like Georgia that does have them?

Don't see why not.

The only reason why the Nook couldn't display many "foreign"/"rare" characters was because of poor character support across the device's fonts.

(If I remember correctly, the Nook also didn't have a fallback font like Kindles with Code2000, so if the selected font didn't have it... you got the '?'.)

Even many Eastern European characters were missing (like 'č' or 'Š').

But in 2020, I wouldn't worry about it. Just use the proper characters.

Originally Posted by Quoth

SmartQuotes is actually dumb and broken. '89 gives ‘89 instead of ’89. Also on LO it does “ instead of ” after — or …!

I ranted about that back in 2014 too... lol, I even ranted about primes:

Read Post #19+ in "[old thread] non breaking spaces (* and *) automatically removed".

Not too much has changed since then, but there have been some enhancements:

• Prime/double-prime detection
  
  • Calibre smartens these fine now.
• Quote-after-em-dash
  
  • Calibre also smartens this better.
• Exception lists
  
  • Diap's plugin lets you manually add words where it won't put a LEFT QUOTE.
    
  • ’em + ’til + ’tis
    
• Toxaris's "Check Dialogue" in his EPUB Tools
  
  • Points out all the mismatching opening/closing quotation marks.
    
  • Greatest thing since sliced bread.
    
  • Long monologues in Fiction have multiple successive paragraphs with an opening quote and NO closing quote. Toxaris has an option for that, so you don't waste time on false detections.

At the very bottom of my 2016 post, I also gave 2 Regexes I use to catch the shortened years + common ones like the "exception list" above:

"Proof reading help please - General, How to"

So that's mostly how I've been handling this the past many years:

• Run a smartening tool
• Regex a few edge cases
• Use Toxaris's Check Dialogue to catch any strays

[JSWolf]

Originally Posted by Quoth

I've decided to replace ' and " with prime marks on measurements as it seems very few ereaders don't support it and unless you switch off Smart Quotes it can mess them up. I'm considering turning off SmartQuotes as I can easily do AltGr-v, AltGr-b, AltGr-V, and AltGr-B for “ ” ‘ ’ and have the primes saved in Insert Custom Character. I may add them later to Compose, though compose ' and compose " are in use.

SmartQuotes is actually dumb and broken. '89 gives ‘89 instead of ’89. Also on LO it does “ instead of ” after — or …!


I wonder does the old Nook without them work if the epub has an embedded font like Georgia that does have them?

If the embedded font has the characters needed, then they will display on a nook as nook uses some older version RMSDK (but new enough for hyphens so it's at least equivalent to 2.0.1)

[Quoth]

Originally Posted by JSWolf

If the embedded font has the characters needed, then they will display on a nook as nook uses some older version RMSDK (but new enough for hyphens so it's at least equivalent to 2.0.1)

Great.

Pity about 1/2 width spaces for numbers and units, or thin spaces either side of dashes (not hyphens). Still, it's one less difference between eink and paper publishing sorted.

I was thinking this morning that everyone that designed ebook formats too much concentrated on existing displays, fonts in PDAs etc and current HTML at the time rather than putting a complete set of typography support and specifying how the renderers should fall back to use the local device or display limitations.
Another strange limitation lacking that could have been in ebooks from the beginning was client side image map HTML. It's dead simple. Rather than a single link with an image you have rectangles defined for each link in the image. No javascript, CSS or complex processing needed other than subtracting location selected from image location on page.