LibreOffice to Compose, Calibre to Convert

[arjaybe]

Originally Posted by gmw

I've isolated that particular piece from your post for two reasons. First because I think we've covered everything else - at least once - but also because it highlights where I believe the idealist stance departs from reality.

I had to take a couple of days off because I'm exhausted by the immovable focus on the artificial conflict. I started this thread so I could find out if anyone had any tips on using LibreOffice to compose and Calibre to convert. There were some helpful comments. Then someone suggested that I could use another piece of software. I demurred, with my reasons. Then people began to quibble my reasons. I was glad to try to clarify them, but I don't think I can make them any more clear. And I don't care whether or not my reasons for doing what I do matter to anyone else. Nor do I care if they seem unrealistic.

I understand that other people have their own reasons for their choices. I've heard them all, many times. Good for them. I hope they're happy.-)

[gmw]

Ah, but had I been able to sway you over to a more flexible point of view then whole new vistas of opportunity would have opened up. There's a line I love out of Pride & Prejudice that fits here: Excuse my interference, it was kindly meant.

[arjaybe]

Originally Posted by gmw

Ah, but had I been able to sway you over to a more flexible point of view then whole new vistas of opportunity would have opened up. There's a line I love out of Pride & Prejudice that fits here: Excuse my interference, it was kindly meant.

Mine is the more flexible point of view, but I despair of making you see that.-)

[MarjaE]

I'm currently using NeoOffice, because of my strobe sensitivity, and LibreOffice's crashing on the Mac.

I'm having some trouble though.

I'm using tables, instead of text tables, so things with with different page widths. Otherwise I'm using simple formatting. I've had complex formatting go haywire and/or crash LibreOffice when converting between docx and odt.

I get erratic spacing when I convert to html.

I don't know anything about cleaning up the html.

I get more erratic spacing when I try to convert to epub in Calibre.

Any idea what to do about spacing and cleanup? Or how to replace the text table of contents with a hyperlinked one?

Thanks.

[Hitch]

Originally Posted by gmw

Ah, but had I been able to sway you over to a more flexible point of view then whole new vistas of opportunity would have opened up. There's a line I love out of Pride & Prejudice that fits here: Excuse my interference, it was kindly meant.

REGARDLESS:

A comment about YWriter:

Spoiler:

If there's anyone else here, that wants to find software that is excellent writing software, particularly for fiction projects that are scene-based, I can cheerfully vouch for YWriter--which I shan't categorize as "free" in any sense, as apparently, that's a trigger word here. If you wish to use YWriter, sans payment, you are FREE to do so. The developer will cheerfully accept a donation of $25, for the "licensed" version, which has exactly O additional features. I license it, because I like to support software that I think has value, much as I donated heavily to Sigil back when Valloric and then User_none, etc., were accepting donations. Why? Because 99%
of all "open source" software goes down the crapper, due to the fact that people don't want to work for nothing, and have zero interest in donating their time to software that won't reward them. Such is life.

A comment about other "writing" software:

Spoiler:

I've tried and used probably 15 different "writing" systems/software. Half because I was curious; half because it's somewhat my line of work. Most, IMHO, regardless of the price, spend entirely too much time trying to make the buyers feel like "real writers" rather than making the writing tasks themselves useful/efficient. It's like LSB XE, which has a character/scene timeline--that you have to fill out yourself. Gee, thanks, like I couldn't have done that with something else, like a mindmap? (Whereas, the inexpensive/no-cost YWriter does it FOR you. Now, to me, that's value.)

Back to the issue of sending LO/OO/whatever to HTML--I can see that if you are going to use Calibre to make an eBook, that naming thing gets to be an issue. I see the stylename-number thing in lots of eBooks that somehow, wend their way around the net, until they find us, to be "fixed," which Calibre-33 or whatever as their styles and naming conventions. Really, the Writer2ePUB thing doesn't work around this?

@arjaybe:

Presumably, OO/LO has something equivalent to Track Changes in Word, is that right? What you describe as your initial problem sounds to me like what I see when a Word user does not "accept all changes in Document" before exporting to HTML. Any chance that's what is happening with you?

Hitch

[arjaybe]

Originally Posted by Hitch

@arjaybe:

Presumably, OO/LO has something equivalent to Track Changes in Word, is that right? What you describe as your initial problem sounds to me like what I see when a Word user does not "accept all changes in Document" before exporting to HTML. Any chance that's what is happening with you?

Hitch

LO has Track Changes, which I have never knowingly used, and I've never been asked to accept any. I did manage to get rid of the CSS litter by (I think) turning off some automatic feature. Too close to bedtime to remember what it was, though. Maybe it'll come to me in my dreams.-)

Edit: Sleep is a wonderful thing. It was Word Completion. I turned that off and got rid of those pesky little bits of useless stuff in the HTML.

[gmw]

Originally Posted by MarjaE

I'm currently using NeoOffice, because of my strobe sensitivity, and LibreOffice's crashing on the Mac.

I'm having some trouble though.

I'm using tables, instead of text tables, so things with with different page widths. Otherwise I'm using simple formatting. I've had complex formatting go haywire and/or crash LibreOffice when converting between docx and odt.

I get erratic spacing when I convert to html.

I don't know anything about cleaning up the html.

I get more erratic spacing when I try to convert to epub in Calibre.

Any idea what to do about spacing and cleanup? Or how to replace the text table of contents with a hyperlinked one?

Thanks.

You probably should have started a new thread rather than activating such an old one ... but never mind, we're here now

Both LibreOffice and NeoOffice derive from OpenOffice - but that doesn't mean they are the same product, only that they have lots of similarities. I use LibreOffice (I'm still on v5.0 at the moment, I only update when I get around to it), not NeoOffice, so what I have to say may or may not apply to NeoOffice.

I am not entirely certain what you mean "I'm using tables, instead of text tables", so I can't really comment on that other than: anything beyond simple styled paragraphs and headings is very likely to cause issues when converting to HTML. Not saying it cannot be done, just saying these areas are likely to need adjustment in the HTML itself.

As for conversions between docx and odt, LibreOffice seems to do very well. Its not perfect, but I get away with using LibreOffice here while all my clients use MS-Office and about the only documents we have any difficulty exchanging are complex spreadsheets. I have had LibreOffice lock up a few times when opening doc/docx files, but it doesn't seem to happen too often (not often enough to have made me update my copy of LO).

Erratic spacing after conversion to HTML might have various causes. The first conclusion I would jump to is inconsistent using of styles. For a clean conversion to HTML you must learn how to use styles consistently - this is true whether you use MS-Office, LibreOffice, OpenOffice. Any inconsistency in your use in styles in the source document is bound to lead to problems in the HTML (and therefore the epub). There was a lengthy discussion here a few years ago (see this thread, but you may have to page around a bit to find the relevant posts).

That thread mentions that I use an OpenOffice/LibreOffice plugin called writer2xhtml to produce an epub directy, which I then clean-up in Sigil. This plugin does produce a hyperlinked table of contents from the source table of contents. However, support for this plugin (part of a package called writer2latex see here) has faded out about 2 years ago and there have been issues with later versions of LibreOffice (which I have been able to fix for myself, so I'm still using it). You may have better luck with writer2epub (see sub-forum here in MR), but I've never used it so cannot speak for how well it works.


As to how to clean-up HTML ... that's a big topic. I use Sigil, but I've read that the Calibre epub editor is coming along okay, so you may prefer to try that. A tool like this lets you switch back forth between the raw HTML and what it will look like. ... But how useful this is will partly depend on how much you know, or are will to learn, about HTML and CSS.

It's not really that difficult to learn enough for what you need in most epubs, but it will take a lot of time if you are starting from scratch - so you may need to get help. I could direct you to some useful HTML/CSS tutorial sites, but these may be overwhelming because they talk about a lot of stuff you will never (or should never) try to use in a epub. Also, you can open existing epubs you already have to get an idea of how things are done.

[gmw]

It occurred to me that perhaps I should clarify "inconsistent use of styles" (save you having to delve into that other thread I linked). The biggest surprise that most people hit when they export to HTML is that their spacing or layout has gone awry. To avoid that the rule is pretty simple:

All document whitespace and layout should be managed using styles - not with the spacebar, return or tab keys.

Use your spacebar only to put a single space between words. Use your return key only to start a new paragraph or heading. Do not use your tab key.

Any time you think you want to hit the spacebar or return key more than once (or a tab key even once) to insert extra space then know that you are not using styles properly (are are likely to run into problems with HTML export*).

Use explicitly named styles for your part and chapter headings - typically these would be level 1 and level 2 headings, respectively. Front matter should not use heading level styles, just explicitly named paragraph styles (to avoid having them accidentally show up in the TOC).

That is the objective. The only interesting part after that is learning how to get the styles to do what you want.**

Note: It's only difficult when you first start. Once you get a collection of named styles that work as you want, you just use that as a template for your new documents and keep writing.


There is a second rule: keep your formatting as simple as possible. The more complex your format requirements the more difficulties you will experience when you migrate between formats.


* Certain export tools can try to use various tricks to simulate your intentions in HTML, or you can do lots of post processing on your HTML files to clean up. I suggest it is better to learn to use styles correctly in the first place - it's good practice for what you need to understand to work with HTML, because what I explain above is how HTML/CSS works - and therefore - epub too.

** Disclaimer: There are some things your *Office styles will do that are not easy/obvious in HTML, and vice versa. However, by using named styles (and assuming your export tool uses those names - some don't) adjusting things inside the epub is much easier because you only edit the CSS file (HTML's version of the styles) and everything will adjust to match.

[MarjaE]

Thanks.

I have generally used tab and return, because styles would make everything go haywire.

I'd have an asterisked list, and a developer in would turn it into a numbered or bulleted list in Word, and I'd edit it in LibreOffice-- if LibreOffice didn't crash due to incompatible bullet points-- I'd end up with part numbered or bulleted, part not, and part numbered or bulleted again without continuity.

I created three styles and switched to "proper" tables instead of tabbed text tables, because I know text tables won't convert as well, but I absolutely need tables and need the items to line up, which also requires blank lines between some entries in the same cells.

P.S. If I change one of the styles to add a space after each paragraph, how do I update oher documents using the older versions of these styles?

I switched to NeoOffice because of my strobe sensitivity, but right now, neither LibreOffice nor OpenOffice will run due to a "Mutex" bug.

[gmw]

Originally Posted by MarjaE

Thanks.

I have generally used tab and return, because styles would make everything go haywire. [...]

Cross reference this with what you first said about "I get erratic spacing when I convert to html." It's basically the same problem.

Fixing the problems after-the-fact (after putting all those manual spaces, tabs and returns into the document) is going to take time - whether you do it inside the source document (which I recommend) or whether you do it in the HTML. If you have been consistent enough you can sometimes fix many things with a replace-all, but even that takes practice to get right.

If you want to get a good epub result the problems have to be fixed, it is up to you to come up with the process you find easiest whether that is to fix in the source or fix in the HTML ... or something else, see below.

Originally Posted by MarjaE

[...]I'd have an asterisked list, and a developer in would turn it into a numbered or bulleted list in Word, and I'd edit it in LibreOffice-- if LibreOffice didn't crash due to incompatible bullet points-- I'd end up with part numbered or bulleted, part not, and part numbered or bulleted again without continuity.[...]

And here we're getting into the realm of how consistently the first person created their list styles - leaving the question of whether the software is doing anything wrong or whether it's just a question of inconsistent usage. I suspect that latter. With Word and *Office it is very easy to get things to "look right", to cover over inconsistencies, and you only uncover the hidden mistakes when you go to covert to another format.

Originally Posted by MarjaE

[...]P.S. If I change one of the styles to add a space after each paragraph, how do I update oher documents using the older versions of these styles?

You can load styles from other documents. Look in the help under "Loading" or under "styles", there is (in LibreOffice anyway) a section explaining how to do this.

Originally Posted by MarjaE

[...]I created three styles and switched to "proper" tables instead of tabbed text tables, because I know text tables won't convert as well, but I absolutely need tables and need the items to line up, which also requires blank lines between some entries in the same cells. [...]

Between the use of bullet lists and tables, it seems a safe assumption that you are creating technical documents rather than narrative text. This will always present much greater challenges. Many technical writers, preparing documents for publication, prefer to use other tools. Many like LaTex because it offers more explicit control over some layout issues. See also DocBook.

Getting a smooth result with some complicated format requirements in an epub - that works well across most e-readers - can be very difficult. You may find it easier to pass the problem onto a professional in this area.