- gutvol-d - lists.pglaf.org

Re: [gutvol-d] book of james -- 014
by Bowerbird＠aol.com 18 Jan '12

18 Jan '12

james said: > I just wanted to be certain that > Bowerbird had some plan to do that. thanks for giving me a chance to prove i'm not stupid. -bowerbird p.s. i have studied e-books _intensely_ for 3 decades. if something is bloody obvious, you can probably just assume, safely, that i have stumbled across it before...

2 1

really?, so nobody is gonna tell him?
by Bowerbird＠aol.com 18 Jan '12

18 Jan '12

really?, so nobody is gonna tell him? because, you know, if he works on it for a time, when he finds out later -- as we all know he must, eventually -- he'll wonder why no friends told him. -bowerbird

1 0

Re: [gutvol-d] book of james -- 014
by Bowerbird＠aol.com 18 Jan '12

18 Jan '12

james said: > I agree that using them for things that are > not actually quotes could qualify as tag abuse, > it is one of the tags that the Kindle understands but that's not a matter we are deciding right now... down the line, we'll do customizing, for the kindle, and kf8, and ibooks, and nook, and .pdf, and so on, but for now, we're just nailing down the basic .html. don't get ahead of yourself. we'll arrive there soon. > I suppose some of these places where I'm using > blockquotes could be replaced with bulleted lists. hold on. i never said a "bulleted" list. if you check, you'll see i actually specified that no bullet be used. so the average person won't even know -- or care -- that those chunks are technically tagged as "a list"... and now i realize that i probably didn't remind you that the latest version of the .zml file is stored here: > http://zenmagiclove.com/bhaga/bhaga.zml i'm pretty good about updating it when i point you to a new version, so you can store that info-nugget. anyway, on page 11, for instance, you will see that the .zml file has an " o " at the start of every item, and the converter turned it into an unbulleted list. that's true of all of the lists that i tagged in this file. > So if I put asterisks at the beginning of each line > in these lists you could make them a bulleted list? i already did it. but it's " o ", and it's _unbulleted_... (i haven't made a firm rule about the exact .zml tag for an unbulleted list, but for your file, it'll be " o ".) > I don't think there is any poetry in this book so we won't even think about that. be in the moment. *** and then later... james said: > Forget what I said about bulleted lists. i decided to turn it into a teaching opportunity instead. > These lists are pretty clearly just indented paragraphs. i'm not clear exactly what you mean by that "just", or indeed with the entire sentence for that matter... there are three different things which "matter" here: 1. what the chunk of text _is_, ideally and structurally. 2. the _.html_mark-up_ we apply to the chunk of text. 3. how the chunk of text will appear to the end-user. it helps to be exact in your thinking, and communication, on which of these 3 things you have under consideration. > The Kindle and Nook indent the first line of > every paragraph by default, so we're good. we try not to depend on "defaults", since they can change. > a lot of words with backslashes in the middle of them. > I'm guessing the backslashes represent soft hyphens. you're "guessing"? seriously? don't "guess". _sleuth_. and once you look at the .zml file, it should be obvious. what you need to do a lot -- reverse-engineer the .zml, so you will gain a deeper understanding of the format -- is view the .zml input file in one browser-window and the .html output file in another, in a side-by-side comparison. (i'll have more on that particular pedagogical method later.) > In the family trees the names of women are italicized. > I've marked them up that way > but your HTML does not italicize them. i can't remember if i did the italics conversion in this file. maybe not, since it is unnecessary in this work-product... > When you remove the underscores > the alignment of the trees will be thrown out of whack too. the "alignment" you were looking at was purely temporary, framed for the particular purpose of this _work-product_... you need to burn that into your brain, james, right away, because i am _already_ tired and bored with repeating it. > Maybe we need a special rule that > when underscores occur in a <pre> section > they are replaced with spaces rather than removed. please do not be under the impression that we are only now just making up "the rules". most of 'em have been very firm for _many_years_ now, and the ones which are still "flexible" are in that state because it's better if i can _bend_ 'em a bit... from the perspective of .zml, there is nothing difficult about your book. _nothing_. so just sit back and relax, james, and answer the exact questions that i ask, and leave it at that. ok? > I'm still kind of puzzled about how we're going to deal with > page breaks (and numbers) in the final output. well, there will be many forms of "final output", some of which honor those pagebreaks -- and thus treat them appropriately... but for something like the kindle version, they'll be eliminated. i explained the precise rules for their deletion to john redmond a while back. you didn't respond to my suggestion that we do a friendly python coding competition, for the greater e-book good. but if you wanted, you could write a script based on those rules... > list of words and names have no accents or diacritical marks, > and so does the HTML. Even with global search and replace > that's not going to be fun to put back. temporary! work-product! answer the specific questions only! temporary! temporary! temporary! work-product! temporary! so sick and tired of repeating this! aaaaaarrrrrrgggggghhhhhh! aaaaaarrrrrrgggggghhhhhh and aaaaaaarrrrrrrggggggghhhhhhh! -bowerbird

1 0

book of james -- 014
by Bowerbird＠aol.com 18 Jan '12

18 Jan '12

james- good job on the feedback. my replies are appended. you now have some new .html versions to preview... here's the full version: > http://zenmagiclove.com/bhaga/bhaga-preview2.html and a version which scrunches ordinary paragraphs: > http://zenmagiclove.com/bhaga/bhaga-preview3.html the latter version makes it much easier for you to check the paragraphs with some kind of "special" formatting... the formatting is now close enough for you to finish off. the next step will be to do a global correction of names. in that regard, you might wanna do a quick look at this: > http://zenmagiclove.com/bhaga/bhaga-spellcheck-fail.txt these are the bulk of the words that "failed" spellcheck. as is usually the case, the ones with an initial cap are almost exclusively names, and most of 'em are correct. i'll give you another version of this list later that tells us with far greater certainly which of them are likely right... the words that are listed in lowercase, at the bottom, are more likely to be scanning errors, as you can see, while there's also a mix of british spelling and words specific to this particular book, which is also typical... once we add the appropriate words to the dictionary, you'll do a focused correction of the remaining terms, after which there'll be very few words that get flagged while you're doing your page-by-page read-through, which will enable you to focus on punctuation glitches. i've got some other stuff brewing today, but i'll try to get to those names as soon as i can. in the meantime, if you're antsy to do some work, do more family trees... *** al said: > Can't help you with Bowerbird. To borrow a phrase, > "you've made your bed, now..." <g> i guess i must have missed the part where al offered to help you on this book, james. maybe he said that he'd re-do the o.c.r. for you, using abby finereader? yeah, i must have missed that... :+) -bowerbird > Page 1 color-coded text should be a single blockquote. who is being quoted? no, those are simple paragraphs. you can decide if you want them to have just an indent -- which means only the first line will be indented -- or whether you want the left-margin to be adjusted, in cases where reflow causes the first line to be split. but it would be "tag abuse" to call these "a blockquote". > Page 11 color-coded text could be a blockquote. again, who is being quoted? this probably should be a list. but i'd be inclined to mark it up as a plain old paragraph... ok, no, i take it back, let's just call it a list. that's what it is. > Page 13 blockquote for the first colored section nope again on the blockquote. once again, this is a list... > pre for the family tree below it. all the family trees will be tagged as "pre", no question... > Page 15 should be blockquotes. just indented paragraphs, as far as i can see... > Page 21 pre tag, I guess. I'd probably make > an HTML table for this myself, but I don't see > how you could do that automatically. i did this table for you. if you look at it in a monospaced font, you'll see it all lines up. however, it's not necessary to do that. a space in column 1 and the vertical-bar in column 2 signify a table in z.m.l. vertical-bars also demarcate your columns. these vertical-bars need to have a space on either side of 'em. (except the rightmost one does not need a space to its right.) right now, i'm just marking all the tables with a "pre" tag, but in a future iteration, the software will make real .html tables. > Page 25-6 block quotes. This is really a multi-level outline > so if I was doing it by hand I'd use nested block quotes. i'll have to dig out my sample code for multi-level outlines. but for now, i'll just leave all these as ordinary paragraphs. > Page 30 top one block quoted section. it's a list. > Page 30 bottom - 31 I'd probably reformat > (using words instead of ") so it worked as a blockquote. i did it as a table. > I'd rework the formulas so they could be read > in a proportional font and still make sense. Example: > 1 Manvantara = 12,000 x 1,000 / 14 = 857,142 6/7 Deva years. great idea. i'll let you do that. > Page 46 is a footnote. is that the only footnote in the book? if so, i'd just leave it as is. > You seem to be double-spacing my family tree tables. i probably forgot to take out the "br" tags in a "pre" paragraph. > We had also talked about doing both illustrations from > the book *and* ASCII art to represent family tree tables. > Does light markup work in that scenario? make the illustration, and then pull it in. see the example on page 11 to figure out how to have an illustration shown. > Page 82 will eventually be an HTML table. i did it as a table. i chopped it in half for you.

4 6

the book of james -- 013
by Bowerbird＠aol.com 17 Jan '12

17 Jan '12

al said: > from what I've read here the past few weeks, > I think you've taken on a project that's > more than your current skills/experience > will let you handle easily. well, al, i underestimated james based on that, too. so i can't say i blame you. but i found i was wrong. his book showed that he possesses lots of expertise. james has done a ton of work to learn his way around. he is kind of "google-smart", rather than experienced, but google-smart's a good way to get yourself started. i mentioned his strengths in a recent post, and forgot to include that he obviously has a love for digitizing, and -- deeper still -- both love and respect for books. and that's a great cornerstone on which you can build. i also forgot he built himself his own scanning station. so, as for his "shortcomings", i think it boils down to 2. the first is that he's "frugal", as in cheap. the second is an irrationality about open-source, which could probably be related to being cheap, except that irrationality can't be spun as a virtue. (but, of course, such irrationality is fairly common with open-source advocates, which unfortunately mocks the intelligence bedrocking the philosophy.) the combination of these two things leads james to value his time at a very low worth, and that is sad... the proof of this is that, even after he was told -- point-blank, by the d.p./p.g. infrastructure, as he reports in his book -- that tesseract is worthless, he _continued_ to use it for his following projects. as he put it in his book: > since my original page images were of > poor quality (due to the age of the book > and inadequate lighting) it was difficult to > convince anyone that ABBYY Fine Reader > would not have done a better job on the OCR > than Tesseract. It is possible it would have. read that again: > It is possible it would have. there are two ways to interpret that. the first is that he hasn't done the research to verify it. well, hey, james, if that's the case, research it! (or else take the word of the people who did.) the second, more troubling, interpretation is "they say it is so, but i question their wisdom." and _that_ boils down to a sheer irrationality bordering on pure stupidity, akin to "they say the earth is round, but i don't know about that." there's no question about the fact of the matter, james, and you need to face that fact head-on: abbyy finereader does better o.c.r. than tesseract. out of 100 books, finereader will do 95-99 better. i wish it wasn't so lopsided. but that's what it is... *** so, al... those are the two "shortcomings" i see in james -- hey, i'm sure he can see a lot more in me! -- :+) but neither of them affects his ability to pull off a digitization of this book. he has the tenacity... i agree that his inability to see the difficulty that he was causing himself by rewrapping the text _before_ he did the proofing was quite glaring, but i think he now sees the error of the practice. plus i hope he now knows the truth of tesseract. so i think we need to give him full credit now... (as if my "credit" will buy him anything!) ;+) *** james, the other "problem" i had with your book has nothing to do with the book per se, but with your co-author -- rebecca "webchick" malamud. back when i was trying to communicate with the internet archive people, she was working there... at first, she was all _nice_ and everything to me, but once she learned that i wasn't going to just "go away" after lodging my complaints, she got rather mean and even did some attacks on me... so until she apologizes to me, i don't like her... that, of course, doesn't have to have any bearing on _our_ relationship... unless you want it to... but it sounds like she might not be a close friend. (if she is, you should ask her if she has finereader, and if she'll be willing to do some o.c.r. for you.) *** more to come shortly, as we get back to the work... -bowerbird

2 1

Re: [gutvol-d] Producing epub ready HTML
by Bowerbird＠aol.com 17 Jan '12

17 Jan '12

james said: > Bowerbird has been suggesting using light markup in text files > so we can derive other formats automatically and reliably. > He will doubtless have something to say on the subject. thanks, but no thanks; i've stepped off that merry-go-round. if hunter wants to know how he could get better results from marcello's converters, hunter should discuss it with marcello. i'm more interested in having discussions with people who want to get better results from _my_ conversion routines... so, james, what do you think of this .html file of your book? > http://zenmagiclove.com/bhaga/bhaga-preview1.html i left the pagebreaks in for your convenience. you'll notice i've colorized the tables, family trees, and some other stuff. all the styling is inline, but it's a tightly-constricted subset. take a look at that formatting, comparing it to the p-book, and let me know where i've gotten any wrong, so i can fix it. once that's right, we can take a look at the .mobi and .epub. (those formats are based on the .html, of course, but i also generate different x/html files for the two different targets. i'll also break the targets down to even more specific levels, such as .epubs specialized for the nook, adobe, or ibooks... there's also a slew of customization options for end-users, including stuff like paragraph indentation, curlyquotes, etc.) once you approve, the book is ready to be turned back to you. thank you for being so patient while it was under my control... -bowerbird

3 4

Amazon releases new KF8 and MOBI tools
by James Adcock 17 Jan '12

17 Jan '12

Amazon has finally released their "new improved less crappy" tools which support as output format KF8 and MOBI. As input format they accept HTML5 and CSS3, not that that will help us submitters because PG refuses to accept HTML5. Hopefully PG will at least pick up the new version of kindlegen so they can turn back off some of the Marcello file rewriting. The new version of kindlegen: www.amazon.com/kindleformat/KindleGen Improved simulator software: <http://www.amazon.com/kindleformat/KindlePreviewer> www.amazon.com/kindleformat/KindlePreviewer

5 7

Re: [gutvol-d] Producing epub ready HTML
by Bowerbird＠aol.com 17 Jan '12

17 Jan '12

james said: > First, you moved all the front matter like title page, > TOC, preface, etc. to the back of the book. that's of no consequence. this is a temporary work-product. (i did it that way so the vast majority of pagenumbers worked, because i was too lazy to build in the offsetting mechanism.) > Second, you have put <pre> tags around some content > that should be just blockquoted text. which paragraphs, specifically, as a general class? and yeah, i realize that question is self-contradicting, but... the decision on formatting was made by a crude algorithm, and i can probably improve it significantly, if i merely know which paragraphs it formatted incorrectly. so please tell me. i would look myself, but i've grown tired of this book, sorry. all of this formatting can be changed by you, at a later time, but i'm trying to do as much as i can for you, automatically, since correcting the text will cause you enough pain as it is. > The family trees should be pre-formatted, but much of > what you made pre-formatted text doesn't really need to be. right. that's what i was asking you to check. > italicized words also have underscores before and after. no consequence. > Fourth, I noticed that this book is incredibly long > and I was crazy to work on it. 880k. it's not _that_ long. it does have a lot of names that are not that familiar to westerners, but they can be handled in a bulk manner, so that doesn't have to be so hard either. diacritics can be hard, but that's just one global change too, once you have a correct version of how each word should be. (you'll laugh at yourself for having done 'em one-at-a-time.) once you boil it down this way, it's not any harder than doing two 225-page books. in fact, it's a little bit simpler than that. > I prefer having the contents of the book be > in the same order as the original you don't really think i would move frontmatter to the back in the final product, do you? no one would do that, would they? > and I think PG prefers TOCs with no page numbers. you must learn not to confuse what p.g. wants with what should be done. but yes, those pagenumbers can be eliminated... but no, you shouldn't do that... and finally, no, p.g. doesn't really care these days... > I'm not really sure what I'm supposed to do with this HTML. you aren't supposed to do _anything_ with this .html... except what i said -- look at the colorized paragraphs, and tell me which of them, if any, is improperly coded... > Is this the output of a light markup file you will give me? it's output from the current version of the file, found at: > http://zenmagiclove.com/bhaga/bhaga.zml but the book is still in my hands, and i'm not done yet... i know that the idea you (and most people) now have is that .html is something that you create "at the end", when you're "all done" with the book. but i use it as _a_tool_ that visualizes the work as it is in-progress. try to rewire your brain to think that way. > Is there some sort of web interface coming for > working on this? I could use some clarification. yes. -bowerbird

2 1

kf8 and mobi_unpack
by James Adcock 15 Jan '12

15 Jan '12

Mobileread has a good discussion of kf8 issues, and a python tool mobi_unpack.py (suitable for unpacking and understanding public domain kf8 files) at: http://www.mobileread.com/forums/showthread.php?t=61986 <http://www.mobileread.com/forums/showthread.php?t=61986&page=17> &page=17 It runs correctly and without any modifications or installation hassles on my windows CPython installation. I haven't been able to get it to run under IronPython, but then again I don't have much experience with IronPython (or with CPython, for that matter) If anyone else is interested in playing with KF8 (and presumably has a Kindle Fire and/or Kindle Previewer software) or in reading pleasingly formatted PG books on their Kindle Fire, I'm having good luck running PG html files directly through the new kf8 kindlegen, [as opposed to compiling a PG epub for example, or reading the PG version of the .mobi file] - especially if I take a minute or two to fix obvious html coding errors in the PG html files -- such as specifying side margins in the body.

1 0

book of james -- 012
by Bowerbird＠aol.com 15 Jan '12

15 Jan '12

james said: > I don't pass myself off as an expert. well, be that as it may, now i know you _do_ have expertise, in many things, which was being obscured from my view by the glare of a few "inadequacies", for want of a better word. i'll deal with the shortcomings in a later post, but for now, what's important is moving the lessons to a different point on the learning curve, one that will leverage your strengths. what was unclear before, which i now know from your book -- and will detail outright, so the others here know it too -- is that you have a fairly solid grasp of the d.p. software, and scan tailor (which is a big plus), and o.p.d.s cataloging stuff. heck, you are even comfortable setting up your own server... in addition, you've used stuff like calibre, sigil, and booki, which means you know how to make an .epub, and a .mobi, so what you seek is a _superior_ solution for those formats. (which is what i have. thus, it makes it a lot easier, for me, if we both know that that's exactly what you're looking for.) another admirable quality is you are willing to do research. you don't sit and wait for someone to hand you something; you go out and perform an active search to find solutions... but your most vital strength is that you are a programmer. i knew that, in the back of my mind, since you said you had done stuff for the o.l.p.c. machines, but if i would've known that you code in python, we coulda cut to the chase earlier... because that means that you're probably a new subscriber to this list, and were not here last september (and earlier). when i have given numerous e-book app-coding tutorials. i mention september in particular because it was then -- for the first time -- i was coding in python _and_ actually sharing my source... i suggest you go back and read that series, running under the title of "these grapes are sweet". the posts started in september, and some ran into october: > http://lists.pglaf.org/mailman/private/gutvol-d/2011-September/date.html > http://lists.pglaf.org/mailman/private/gutvol-d/2011-October/date.html again, that's the first time i coded _anything_ in python, so my scripts will be silly and crude to you, but you will start to see how easy it can be to write code to help you. and i invite you to riff on my work. let's get it going on. i've been long waiting for a programmer to come along who would spur me into a positive collaborative spiral... and you, james, just might be that guy. -bowerbird

1 0