Extended <head> chapter: NOT necessary? - Page 3

Turtle91 · 02-15-2017, 12:05 PM

Quote:

Originally Posted by DaleDe

Anyone changing i and b are just wrong to do so. That is what em and strong are for. em and strong default to i and b but can be modified as desired. They are semantic statements.

Dale

You are quite right! I guess putting that section inside the spoiler wasn't the best choice - it might've caused someone confusion over what I meant. Thanks for the clarification!

chaot · 02-15-2017, 01:09 PM

Quote:

Originally Posted by DiapDealer

You figured out the missing title tags

Yes. That were my minimization measures. Your comments were very instructive!

Quote:

Originally Posted by Turtle91

I cringe every time I see how bloated with calibre classes some books get. Most of them are unnecessary.

I also think so. What do you do about it?

Quote:

- Please do not use a class to describe your predominant styling
- Please do not use a class to describe your and tags unless they are something other than the standard italics and bold

The code is a complete mess, I know. But so I got the book. I do that much smarter.

Next I show here!

Code:

 - Please do not use empty paragraph tags (<p class="calibre69"> </p>) to create a space between items/paragraphs

So far I use   for single empty lines. Or   to make empty lines traceable via Search: blanc.

 is editable via CSS p.space {whatever} - that's an advantage. I will follow you using for 'non-standard spaces' between paragraphs.

I studied code & CSS - I understand everything but this:

Code:

p  {whatever styling you want your standard (98% of the book) paragraph}

Quote:

Originally Posted by JSWolf

Some of the CSS is not needed. You do not need the i and the b styles as they are italic and bold as is.

That's correct!

Additional question: Is lang="de" superfluous?

Code:

<html xmlns="http://www.w3.org/1999/xhtml" lang="de"

Today it feels like a kind of Christmas and Easter, all in one. That was very friendly!

you all. A fruitful thread.

Turtle91 · 02-15-2017, 01:37 PM

Quote:

Originally Posted by chaot

[CODE]
I studied code & CSS - I understand everything but this:

Code:

p  {whatever styling you want your standard (98% of the book) paragraph}

Books are mostly just regular paragraphs with no special styling. I estimated about 98% of a book has no special formatting. Therefore you can have a normal paragraph defined in the CSS and then only define classes for those few paragraphs that need special formatting.

Instead of:

Code:

<p class="normalpara">This is a normal paragraph.</p>
<p class="normalpara">This is a normal paragraph.</p>
<p class="normalpara">This is a normal paragraph.</p>
<p class="normalpara">This is a normal paragraph.</p>
<p class="red">This is a special paragraph.</p>
<p class="normalpara">This is a normal paragraph.</p>
<p class="normalpara">This is a normal paragraph.</p>
<p class="normalpara">This is a normal paragraph.</p>

CSS:
p.normalpara {margin:0; text-indent:1.2em; font-size:1em}
p.red {color:red}

You can have:

Code:

<p>This is a normal paragraph.</p>
<p>This is a normal paragraph.</p>
<p>This is a normal paragraph.</p>
<p>This is a normal paragraph.</p>
<p class="red">This is a special paragraph.</p>
<p>This is a normal paragraph.</p>
<p>This is a normal paragraph.</p>
<p>This is a normal paragraph.</p>

CSS:
p {margin:0; text-indent:1.2em; font-size:1em}
p.red {color:red}

chaot · 02-15-2017, 01:51 PM

I was suspicious about (98% of the book). OK, now I got you.

Quote:

Originally Posted by Turtle91

To be sure...none of the CSS I posted was a requirement...it was a recommendation. I thought I was pretty clear in stating that in my post...

I didn't think otherwise!

BetterRed · 02-15-2017, 08:09 PM

Quote:

Originally Posted by Turtle91

I cringe every time I see how bloated with calibre classes some books get. Most of them are unnecessary.

Quote:

Originally Posted by chaot

I also think so. What do you do about it?

I mainly convert to EPUB from DOCX. In some cases I convert from PDF to DOCX and then to EPUB. I have lots of 'tools' in Word, some are my own, others are from 3rd party sources, and I've used some of them for decades.

Until the advent of the calibre editor I converted my DOCX's to EPUB via calibre's conversion facility. However when the calibre editor's ability to Import DOCXs was introduced (without CSS flattening that happens in calibre conversions) I switched using it. I tried several other conversion tools (not the online ones, I do too many to consider using them) and found them unsuitable for my purposes.

I have used Word Styles religiously since they were first introduced last century - e.g. I never format paragraphs individually. An exception is italicising words or phrases. Consequently I never got boatloads of CSS entries, even when I was calibre converting ex Word RTFs to EPUBs I didn't get a lot of the cruft some complain about.

Occasionally I would 'rename' the 'blockN' and 'calibreN' CSS entries to have names similar to the Style names in the Word Template, so 'block4' might become 'first_para'.

But more recently I have switched to using the Sigil DOCXImport plugin, this has a facility that allows me to create and store mappings between a Word Template file to an EPUB CSS file. This means I get close to a one for one mapping between Word's Styles and the stylesheet entries with similar names. As a bonus the underlying tool (Mammoth converter) marks italicised text with instead of etc.

So how do I avoid the calibre generated CSS entries - mainly by using Word as a word processor rather than a typewriter, and secondly by not using calibre as my final DOCX->EPUB conversion tool.

I sometimes do 'interim' conversions to get a feel for what the EPUB will look like, or to do multilingual spell checking, or to look at a Calibre or Sigil report - for this I use calibre's conversion facility, pressing 'C' and 'Shift+V' is too easy. I rarely look at the EPUB code at this stage.

BR

DiapDealer · 02-15-2017, 08:39 PM

Quote:

Originally Posted by BetterRed

But more recently I have switched to using the Sigil DOCXImport plugin, this has a facility that allows me to create and store mappings between a Word Template file to an EPUB CSS file. This means I get close to a one for one mapping between Word's Styles and the stylesheet entries with similar names. As a bonus the underlying tool (Mammoth converter) marks italicised text with instead of etc.

Not to derail things too badly; but just a note to mention that for those who may not prefer Mammoth's "em, strong" defaults, that behavior can easily be overridden in a custom style-map. There's commented-out directives at the end of the simplistic sample map provided with the plugin that explains how.

Glad to see the plugin is proving to be useful.

chaot · 02-20-2017, 11:44 AM

Quote:

Originally Posted by BetterRed

I have used Word Styles religiously since they were first introduced last century - e.g. I never format paragraphs individually. An exception is italicising words or phrases. Consequently I never got boatloads of CSS entries, even when I was calibre converting ex Word RTFs to EPUBs I didn't get a lot of the cruft some complain about.

Occasionally I would 'rename' the 'blockN' and 'calibreN' CSS entries to have names similar to the Style names in the Word Template, so 'block4' might become 'first_para'.

But more recently I have switched to using the Sigil DOCXImport plugin, this has a facility that allows me to create and store mappings between a Word Template file to an EPUB CSS file. This means I get close to a one for one mapping between Word's Styles and the stylesheet entries with similar names.

So how do I avoid the calibre generated CSS entries - mainly by using Word as a word processor rather than a typewriter, and secondly by not using calibre as my final DOCX->EPUB conversion tool.

I sometimes do 'interim' conversions to get a feel for what the EPUB will look like, or to do multilingual spell checking, or to look at a Calibre or Sigil report - for this I use calibre's conversion facility, pressing 'C' and 'Shift+V' is too easy. I rarely look at the EPUB code at this stage.

Very interessant, as usual! I read several times, still there are ambiguities. Understanding right it would be advisable to involve sigil for converting books. My OS is Linux Xubuntu, so Word (Microsoft) is not available for me.

In order to get my CSS clearer I name them somehow 'appropriately', e.g. for something what looks like header, but doesn't function as one.

Click image for larger version

Name: no_h4_ital.png
Views: 248
Size: 62.0 KB
ID: 155152

Quote:

As a bonus the underlying tool (Mammoth converter) marks italicised text with instead of etc.

Why you speak of a bonus here? You know, and is semantically something slightly different. I wouldn't like to get all my changed to .

Some funny details!

Spoiler:

JSWolf · 02-20-2017, 12:10 PM

and are the same thing. Both italicize the text.

chaot · 02-20-2017, 12:17 PM

The result looks the same, both italic; semantically they are different.

JSWolf · 02-20-2017, 12:32 PM

Quote:

Originally Posted by chaot

The result looks the same, both italic, semantically they are different.

There is NO difference between & and & . I can change what each of them does via CSS. I can change so it look bold and I can change so it looks italicized. and can also be modified via CSS. Both pairs have the same default. So really, the pairs are no different.

DaleDe · 02-20-2017, 01:39 PM

Quote:

Originally Posted by JSWolf

There is NO difference between & and & . I can change what each of them does via CSS. I can change so it look bold and I can change so it looks italicized. and can also be modified via CSS. Both pairs have the same default. So really, the pairs are no different.

They are certainly different. You should never change an i or b. They are visual emblems and as such should remain as the author chose. em and strong have semantic meanings a these are much more subject to what the meaning for the user is. While eBook readers do not have support for designating the look I think this should be a feature of eBook reading devices. The fact that they can be changed in the CSS is not the criterion for the meaning of anything.

Dale

BetterRed · 02-20-2017, 05:07 PM

Quote:

Originally Posted by chaot

Very interessant, as usual! I read several times, still there are ambiguities. Understanding right it would be advisable to involve sigil for converting books. My OS is Linux Xubuntu, so Word (Microsoft) is not available for me.

Libre Office Writer is probably available for XUbuntu, it can read/write OOXML (DOCX), and because it was developed to the ECMA-376 standard, it's support for the format is 'better' than some older versions of MS Word

At its core Writer is reverse engineered Word '9?, many features like Templates and Styles are functionally similar to Word, an exception is macro's, Word is limited to VBA, Writer gives greater choice - e.g Python. I use LO Writer on the occasional ODT, normally I save as DOCX from Writer and work with that in Word. But sometimes I have to work with the ODT directly, but I save it as DOCX for the purposes of creating an EPUB via Sigil's DOCXImport PI.

Quote:

Originally Posted by chaot

Why you speak of a bonus here? You know, and is semantically something slightly different. I wouldn't like to get all my changed to .

As was previously mention by DiapDealer, the 'starter' mapping that his plugin ships has the wherewithal to turn it off, I wittingly chose to leave it as-is, viz:

Code:

# Uncomment these next two if you don't like the default <em> and <strong>
#b => b
#i => i

The fact that some people here don't like em and strong tags matters to me not one jot. The epubs I create are read by a closed group who wouldn't know a 'b or strong tag' from the price tag on the clothes they bought yesterday

BR

Tex2002ans · 02-21-2017, 05:09 AM

Quote:

Originally Posted by JSWolf

and are the same thing. Both italicize the text.

[...]

There is NO difference between & and & .

chaot is in the right here. HTML5 has gone back from saying and is "purely presentational" to giving them semantic meanings.

The article below goes into more detail with some examples:

http://html5doctor.com/i-b-em-strong-element/

You may also want to read W3C's page, "Using and elements":

https://www.w3.org/International/que...a-b-and-i-tags

Quote:

Originally Posted by JSWolf

I can change so it look bold and I can change so it looks italicized. and can also be modified via CSS. Both pairs have the same default. So really, the pairs are no different.

While + and + typically default to the same look (italics and bold), that doesn't mean they always do (think screen readers).

Also, there could potentially be differences in Text-to-Speech engines. might be spoken with more emphasis on the words, while could be spoken normally.

jackie_w · 02-21-2017, 07:08 AM

Quote:

Originally Posted by Tex2002ans

Also, there could potentially be differences in Text-to-Speech engines. might be spoken with more emphasis on the words, while could be spoken normally.

There are a lot of things which *could* be done with TTS engines but do you know of any mass market TTS app (or reading app) which actually does differentiate between and ?

In fact I'd go so far as to say that all the TTS/Voice combos I've ever tried (Windows and Android) appear to ignore all i/em/b/strong tags completely. I wish they did aurally emphasise italic and/or bold text, but they don't. Maybe one day ...

BetterRed · 02-21-2017, 11:52 AM

@jackie_w - I think I've sampled most 'voices' from Acapela etc and to me they all sound like Milton Keynes looks - pleasantly boring. TTS seems to have barely progressed since I used DecTalk boxes in call centre voice response units in the 1980's, the sound quality is a lot better but the rest...

The problem is that the intonation one might use on italicised text in an ironical statement would be different to that one might use in an adversarial challenge. TTS needs to do deep linguistic analysis (syntax + semantics + idiom, cant, figurative speech, metaphor etc) of the text and get the right base intonation, before it has any chance of using an appropriate tone for italics.

If that were available then the studios wouldn't be using regular actors to do the voice overs for animated movies. And radio broadcasters could get rid of their newsreaders.

If the BBC used what's available now, we wouldn't have to listen to the sneers that drip from their newsreaders tongues whenever they have to mention people of whom they don't approve.

BR

02-20-2017, 12:10 PM	#38
JSWolf Resident Curmudgeon Posts: 73,897 Karma: 128597114 Join Date: Nov 2006 Location: Roslindale, Massachusetts Device: Kobo Libra 2, Kobo Aura H2O, PRS-650, PRS-T1, nook STR, PW3	<em> and <i> are the same thing. Both italicize the text.

02-20-2017, 12:17 PM	#39
chaot Head of lunatic asylum Posts: 349 Karma: 77620 Join Date: Jun 2012 Location: UTC +1 Device: Tolino Vision 3HD	The result looks the same, both italic; semantically they are different. Last edited by chaot; 02-21-2017 at 12:10 PM. Reason: ,→;

02-21-2017, 11:52 AM	#45
BetterRed null operator (he/him) Posts: 20,553 Karma: 26954694 Join Date: Mar 2012 Location: Sydney Australia Device: none	@jackie_w - I think I've sampled most 'voices' from Acapela etc and to me they all sound like Milton Keynes looks - pleasantly boring. TTS seems to have barely progressed since I used DecTalk boxes in call centre voice response units in the 1980's, the sound quality is a lot better but the rest... The problem is that the intonation one might use on italicised text in an ironical statement would be different to that one might use in an adversarial challenge. TTS needs to do deep linguistic analysis (syntax + semantics + idiom, cant, figurative speech, metaphor etc) of the text and get the right base intonation, before it has any chance of using an appropriate tone for italics. If that were available then the studios wouldn't be using regular actors to do the voice overs for animated movies. And radio broadcasters could get rid of their newsreaders. If the BBC used what's available now, we wouldn't have to listen to the sneers that drip from their newsreaders tongues whenever they have to mention people of whom they don't approve. BR Last edited by BetterRed; 02-21-2017 at 11:55 AM.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
xpath to insert chapter breaks - but chapter name cut off ?	Rob557	Conversion	2	03-06-2014 06:59 AM
Troubleshooting Kindle K3 (keyboard) losing chapter head spacing.	timecheck00	Amazon Kindle	5	11-30-2012 08:57 PM
Pocket eDGe and Full-size eDGe: Head to Head Comparison	alefor	enTourage Archive	28	12-01-2010 07:44 PM
Chapter Head Text Over Image	theducks	ePub	10	09-03-2010 03:32 PM
Video Head to Head of Kindle and Reader	Kingston	Which one should I buy?	30	01-24-2008 08:03 PM

Advert

Advert