Glowfic Ebooks

[latanius]

... or, if a comment has two children, just ignore the second one for a while and start a new "chain" for it later? We could go through the tree with a depth first search and insert links to jump between the branches?

That is, Barbecue Forks could look something like this:
<comments that start the chapter>
point A. "Hey, this is a point where we have two branches! one of them is coming up next, the other is here -> link to point B."
<comments from one branch>
<hr />
point B. "Hey, this is the start of a branch. If you want to go back to the other one -> link to point A."
<comments from the other branch>
(... maybe a link to point A here?)

(and something similar with three branches, but... do those even exist? Also, in what order would people want to read an Incandescence / sandbox style chapter by default, if it's different from the Effulgence style?)

[DanielH]

In the sandboxes, there’s one post which says

In Paraasilan, Esmaar, a pair of roommates are about to break a rule that, compared to the one about running in the corridors or even the one about unlicensed teleportation, is there for a good reason.

In unison, they complete their shared spell.

Then, each top-level comment is a completely different, unrelated story about the possible outcome of that shared spell. The spell is a summoning spell, and the possible outcomes are summoning Adarin, the Admin, Joy, Harry Dresden, etc. Each is a completely different sandbox with a shared premise, and can be read in any order. The main post, below the quote, has links to each of these scenarios. Similar sandboxes exist for summoning Cam, a few Milliways meetings, etc.

In contrast, the Incandescence thread in question has two groups of characters doing different things at the same time: Ice and Cypress are at home, while Renata and Zeviana (can never remember their nicknames) are flying to Washington. They are happening simultaneously, and again can be read in any order. This is also how Barbecue Forks works.

In both cases, the “first listed is read first” order works perfectly well.

-------------------------------------------------------------------------------------------

What I was thinking of with my comment would have Barbecue Forks as approximately:

<comments at start of thread>
Comment n: Two of them go upstairs, while two remain downstairs

There are other continuations from this point. (OPTIONAL: links to each of the continuations (including the people downstairs). I don’t think that’s actually a good idea, because it would be more trouble to implement than it’s worth and would be redundant in most sandbox cases and not useful in the Effulgence cases)

<some comments describing the people downstairs>
--------
This is a continuation from earlier (link to comment n).
<some comments describing the people upstairs>

This method means we don’t need to link from each thread to each other thread and easily allows you to get to the parent from anywhere.

[DanielH]

(Yes, I know this is a double post, but it seems relevant)

Since I haven’t seen any online activity from the original creator of the ebooks in a while, I will be releasing updated versions in a week. These versions will have all the chapters (the first ones skip “doors are complicated” due to a bug when processing a comment without a picture), include moity colors (a much bigger improvement to the reading experience than I expected), include symbellas for at least some readers (and most browsers), include external images and some external links, possibly have an appendix for coin shapes, Bellbook entries, etc, and maybe have other improvements not yet specified.

The majority of these changes are already implemented, some of them are merged into the original effulgence2epub github repository and some available as pull requests on separate branches on my fork of the repository (https://github.com/dhouck/effulgence2epub/).

[Alicorn]

Lemme know when I can collect things that should be hosted on my site.

[DanielH]

It’s been a bit more than a week; I’ll upload something tonight. It probably won’t have the external images or appendix, though there’s at least a reasonable chance for the images if nothing further goes wrong. In the meantime, I have a few questions:

Would punctuation improvements, like changing straight quotes to smart quotes or replacing certain hyphens with dashes, be OK? I’m doing that in my own version¹, but it’s an extra step and I can upload the version without that modification if you wanted. What about other improvements? The only other kind I’m likely to be able to get done tonight is replacing “&lt” with actual less-than signs; apparently forgetting the semicolon is a common typo with <animorph thoughtspeak> (oddly enough, in all of Effulgence, that typo is made a few times but no other character entity is mistyped similarly, not even the closing angle bracket).

I’m also thinking of bigger fixes, which I’m correspondingly less sure would be acceptable (for example, Micaiah appeared well before he got the account yovahs_kisses, and I thought about changing the username for the older comments from pythbox to yovahs_kisses). I won’t get to those tonight whether or not they’re wanted, but there’ll probably be updated versions either for program improvements or for Effulgence updates, and I might want to put this kind of fix in those versions.

¹ I feed it to a program to do this for me, so the algorithm is unlikely to handle all edge cases as a human would. I think the only systematic problem I know about is it replacing apostrophes at the beginnings of words with opening single quotation marks, which is great if that’s what they are but not so great if they really are apostrophes like in the word 'twas.

[Kappa]

My personal opinion is that smart quotes are a blight upon this earth but dashes are fine and correcting typos like "&lt" is good.

[Alicorn]

I don't like smart quotes much either - they seem so much more trouble than they're worth. Go ahead and do dashes if you want; my only request is that you never use any "--" that look broken in the middle. I hate those, I don't care if they're conventional where proper em-dashes are unavailable. (And I'm too lazy to type proper em-dash codes.)

[DanielH]

I am curious about the blight-upon-the-earth opinion, but am unable to separate dashes from the rest. Since it’s all-or-nothing, I think the best option is nothing. And I completely agree with you on dashes; I used hyphens before I found the way to turn the Caps Lock key into the Compose key on Linux and make proper dashes (among other things).

EDIT: The all-or-nothing was for the set of smart quote, dash, and ellipsis (which I forgot to mention in the first post). I can do the &lt fixing without touching the others, but those are all bundled into a single module I don’t want to take apart tonight.

[DanielH]

I will quite soon upload an epub and a mobi of Effulgence¹. The epub file is suitable for reading on most ereaders, though not all features will work on all readers; the mobi file should be suitable for kindle².

By complete coincidence⁴, the epub file is also a zip document. If you open it as such (you might need to rename it effulgence.zip) and extract all the files in the EPUB folder (including the subdirectories, but excluding content.opf and toc.ncx) into a place your web server can see, you will have an HTML mirror (with the start page being toc.xhtml, and no problems if you rename this index.html or anything else). This should have more features the ebook, but only because practically any browser will support more features than practically any individual ebook reader.

EDIT: The Kindle version will be delayed for an unknown amount of time as I try to figure out why the heck it isn’t working with profile images when previous versions did. I have a guess, but won’t be able to try it out and experiment until a more reasonable hour of day.

EDIT 2: I fixed the mobi problem, though I can't upload it now because my Internet connection broke and I don't have the file on my phone. The change that broke things was reorganizing into subdirectories; for some reason KindleGen can't handle that but can handle everything being in the EPUB directory.

------------------------------------------------------------------------------------

¹ I am generating the final (for now; I still have other improvements planned in addition to updating with new Effulgence updates) version now. They will probably be in this folder by the time you click the link, given the time of night at which I’m uploading these.

² I have not yet gotten out my Windows PC and Kindle Previewer isn’t available for Linux, so I haven’t tested on most available Kindles. If anybody notices any problems, please tell me. Additionally, the file size is too large for some methods of transferring to Kindle; USB still works, though. This problem is theoretically fixable³, but I haven’t found a way to do so in practice.

³ The file’s a combined format; splitting out the two different mobi versions should produce two files both individually small enough to transfer to the kind of kindle which supports that file, but I haven’t found how to split them and the official Amazon tools (apparently better than the Calibre ones) don’t seem to support generating anything but the combined kind.

⁴ The definition of what an epub file is

[DanielH]

I hate to triple-post, but I still feel it is appropriate.

I just now uploaded the mobi file; the epub/zipped HTML file has been up for hours. There were some odd warnings when building the mobi¹, but they don’t seem to cause any problems when I test it. I haven’t tried it on all kindles, though, so tell me if you catch anything.

If somebody knows how to split it into separate mobi6 and kf8/azw3 files, I would appreciate instructions so that I can upload a file for old kindles and one for new ones; the two individual files will each be small enough to send via GMailing to the kindle address or other send-to-kindle functionality². If nobody else knows, I’ll try to figure this out myself and post the split files if I find anything.

-------------------------

¹ KindleGen claimed the HTML was malformed in some chapters, where the specific tag it pointed to was very simple and clearly well-formed. The more odd part, though, is that it only made this complaint about some chapters even though the HTML is pretty much the same there for all chapters, and I can’t detect any pattern to the ones it complains about. It also warned that it couldn’t find one of the chapters, despite that chapter showing up in the file it generated.

² According to KindleGen, which tells the size of the split file but doesn’t seem to actually split the file itself, the one for old kindles would be about 14MB, and the one for new kindles would be about 20MB. This seems smaller that reasonable, especially since the combined file is far more than 34MB; maybe those are bad estimates or maybe there’s a lot of junk in the combined file I can take out.