Brand New KK

[DNSB]

Originally Posted by Sirtel

Never seen it with well-formatted books. There must be something wrong with the original coding, then. Calibre doesn't invent any styles that weren't there originally, nor does it duplicate them.

And then we have the folks who don't like CSS and every flipping style is done using inline ( style= ) just to bloat the size and make it hard to make changes. calibre converts those to CSS much like using Sigil's Remove Inline Styles plugin.

[Hardboiled]

Originally Posted by Sirtel

Never seen it with well-formatted books. There must be something wrong with the original coding, then. Calibre doesn't invent any styles that weren't there originally, nor does it duplicate them.

Again you're misunderstanding me or I'm not explaining myself well, probably the latter. I'm not saying Calibre invents styles. I'm saying the template has some 20 odd styles, regardless of input. So, therefore if you were take a well formed eBook, with far less styles, the template still has twenty odd entries in the stylesheet upon conversion, some of which are duplicates and can be in the same identical body text, of the same section. Not an issue for a reader conversion at all and makes sense for the core purpose of Calibre but it's not always as convenient as other, cleaner methods, if you do edit and reformat your books regularly. At least that's how I find it and so I try and use other methods. At the end of the day, CSS/HTML is pretty simple and it's not the end of the world and I do use the Calibre Editor too but I wouldn't and didn't ever pass an eBook I was building professionally through it.

[Sirtel]

Originally Posted by DNSB

And then we have the folks who don't like CSS and every flipping style is done using inline ( style= ) just to bloat the size and make it hard to make changes. calibre converts those to CSS much like using Sigil's Remove Inline Styles plugin.

Yes, that's a huge PITA. And often you can't even merge them, because there are 50 different sizes of text-indents or something.

[Hardboiled]

Originally Posted by DNSB

And then we have the folks who don't like CSS and every flipping style is done using inline ( style= ) just to bloat the size and make it hard to make changes. calibre converts those to CSS much like using Sigil's Remove Inline Styles plugin.

That's what I mentioned earlier - you see it particularly with a lot of public domain early books, such as those badly OCRed and converted on Internet Archive and such forth, if it's still being done now for other eBooks - very bad. Even harder for something like Calibre to deal with, I dare say, to be fair. And adding earlier to what I said for clarification, Calibre certainly does duplicate styles because you will see Calibre_3, Calibre_4, Calibre_5 spanning paragraphs in the same identical body text even when the source originally had the one body CSS style, after conversion.

[Sirtel]

Originally Posted by Hardboiled

That's what I mentioned earlier - you see it particularly with a lot of public domain early books, such as those badly OCRed and converted on Internet Archive and such forth, if it's still being done now for other eBooks - very bad. Even harder for something like Calibre to deal with, I dare say, to be fair. And adding earlier to what I said for clarification, Calibre certainly does duplicate styles because you will see Calibre_3, Calibre_4, Calibre_5 spanning paragraphs in the same identical body text even when the source originally had the one body CSS style, after conversion.

Well, I've never seen such a thing, unless the source formatting was a mess. With proper formatting, never.

[Sirtel]

Originally Posted by Hardboiled

Again you're misunderstanding me or I'm not explaining myself well, probably the latter. I'm not saying Calibre invents styles. I'm saying the template has some 20 odd styles, regardless of input. So, therefore if you were take a well formed eBook, with far less styles, the template still has twenty odd entries in the stylesheet upon conversion, some of which are duplicates and can be in the same identical body text, of the same section. Not an issue for a reader conversion at all and makes sense for the core purpose of Calibre but it's not always as convenient as other, cleaner methods, if you do edit and reformat your books regularly. At least that's how I find it and so I try and use other methods. At the end of the day, CSS/HTML is pretty simple and it's not the end of the world and I do use the Calibre Editor too but I wouldn't and didn't ever pass an eBook I was building professionally through it.

Your experience with Calibre and mine seem to differ considerably, then. Perhaps we read different kinds of books.

Of course, I merge identical styles and delete the unused ones before conversion. Perhaps that's why I almost never see the mess you describe.

And Calibre is certainly not meant for professional book-building. It's meant for personal use.

[Hardboiled]

Originally Posted by Sirtel

Well, I've never seen such a thing, unless the source formatting was a mess. With proper formatting, never.

Well we're looking at two different programmes then, because I jsut checked it on mine. On an old book that has always bugged me with it's false margins, not good on a 6" reader, I opened the ePub and then AZW3 in the editor. Sure enough the AZW3 has added duplicates on the same body text. And those annoying margins are, for example margin: 3pt 0 0 2em; They could be 0 0 0, or simply 0, to get the same marginless effect.

You'll have to trust me, that my own eBooks I made, are not badly formatted - I took great pains to make them well formed. I certainly didn't want someone opening them up, which they can do easily and seeing Calibre_ all over the CSS. And I was meticulous about the whole meta data too. So I understand your inference without offence but they too get mangled in Calibre.

[Hardboiled]

Originally Posted by Sirtel

And Calibre is certainly not meant for professional book-building. It's meant for personal use.

Exactly! Which kind of supports what I'm saying no? i.e. the conversions are not ideal. Great for a casual reader and mostly faithful conversion but if the book was bad to begin with and you want to try and make it a lot better, it might be easier, i.e less time consuming ultimately, to use other methods, which I do, when I can be bothered

[Sirtel]

Originally Posted by Hardboiled

Exactly! Which kind of supports what I'm saying no? i.e. the conversions are not ideal. Great for a casual reader and mostly faithful conversion but if the book was bad to begin with and you want to try and make it a lot better, it might be easier, i.e less time consuming ultimately, to use other methods, which I do, when I can be bothered

I do edit the css before converting, always. But as I've repeatedly said, I don't find your statement about Calibre adding useless duplicate styles to be true. If there are duplicate styles after converting, they were there before converting, either in the css or in the html files.

[Hardboiled]

Originally Posted by Sirtel

I do edit the css before converting, always. But as I've repeatedly said, I don't find your statement about Calibre adding useless duplicate styles to be true. If there are duplicate styles after converting, they were there before converting, either in the css or in the html files.

My posts are sometimes getting through and some go to approval, I understand I'm new here. But it's hard to make sense even within my limited capability I did address in a post that has not yet appeared. To whit, it certainly happens in my conversions, so it's true for me.

[Sirtel]

Originally Posted by Hardboiled

My posts are sometimes getting through and some go to approval, I understand I'm new here. But it's hard to make sense even within my limited capability I did address in a post that has not yet appeared. To whit, it certainly happens in my conversions, so it's true for me.

I saw your reply to a previous post now. I do find your experience of "mangled" conversions to be very strange. My experience is exactly the opposite, namely that a calibre conversion can sometimes improve a messy css. Not always, of course. But it almost never makes the css worse or "mangles" it, not in my experience. If there are lots of inline html styles in the original, then yes, the resulting css will usually be bloated, as Calibre will transfer those inline styles to the css. Also poorly formatted MS Word files can result in some really messy css. But I've never seen anything like you describe in well formatted ebooks.

[Hardboiled]

Originally Posted by Sirtel

I saw your reply to a previous post now. I do find your experience of "mangled" conversions to be very strange... But I've never seen anything like you describe in well formatted ebooks.

Yes it is strange. Any road, I didn't come here for that, so I don't want to get on anyone's nerves by coming over a know it all. We all get them working one way or another Back to topic, I just bought two batteries for both the K3s for 30 quid, although they are both pretty good still, as they are. I just want to put them both in as new condition. One is 1900 mAH and one 1750 mAH, like the original, so I hope the bigger capacity one works OK, if so it should hopefully give even bigger runtime.

[DNSB]

Originally Posted by Hardboiled

Well we're looking at two different programmes then, because I jsut checked it on mine. On an old book that has always bugged me with it's false margins, not good on a 6" reader, I opened the ePub and then AZW3 in the editor. Sure enough the AZW3 has added duplicates on the same body text. And those annoying margins are, for example margin: 3pt 0 0 2em; They could be 0 0 0, or simply 0, to get the same marginless effect.

I have trouble seeing your point when you use an example where while setting the top margin to 3pt (a very bad idea using absolute measurements) is close to 0.25em and close enough to 0 for some people while right and bottom are set to 0 but the left margin is set to 2em which is not close to 0.

I have seen several ebooks from Tor where there is a stylesheet for every text file but they are all virtually identical and contain little more than an @page setting the right and left margins to 5px. The remainder of the styling was done with inline styles. In one of those abominations, removing the inline styles, gave me 82 indented paragraphs styles with the indent expressed in pts, px and em, font sizes in multiple measurements and top/bottom margins and paddings again of various sizes and units. The conversion does not do approximations and a style with the bottom margin set to 4 px is not the same as a style with the bottom margin set to 3pt.

[Hardboiled]

Originally Posted by DNSB

I have trouble seeing your point when you use an example where while setting the top margin to 3pt (a very bad idea using absolute measurements) is close to 0.25em and close enough to 0 for some people while right and bottom are set to 0 but the left margin is set to 2em which is not close to 0.

Because in these books, I refer to (not mine), top and bottom margins were already set, in a 2nd style sheet, I should have made that clearer, in the example. Often seen it and is bad practice IMV. You can also set them independently in the same stylesheet. Trust me I know what I'm doing, making eBooks