- gutvol-d - lists.pglaf.org

General question on download totals
by James Simmons 19 Jan '12

19 Jan '12

I don't know who to ask this but I'm puzzled about how download totals work on the PG site. Are they cumulative, that is every download since the book was posted? Or are they something else? I ask this because I check the totals for books I've donated and the current numbers are MUCH lower than I remember them being. Ancient Manners, for instance, was more popular than War And Peace the first week it was available. Now it shows 125 downloads. It got many more than that the first day! I'm just puzzled. I use the download counts as motivation and now I don't know if I should trust them. James Simmons

3 2

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

18 Jan '12

james said: > if I seem a little dense if you seemed "dense", i wouldn't bother... > it's because you reveal what you're doing > in bits and pieces, hints, and old email threads it's a big ocean that i'm boiling here, james... be patient, and drink one cup of tea at a time. > I assumed that the page breaks were going away > until I saw your work product > where the TOC actually had page numbers > and linked to those pages. > That made me wonder. ok, let's use this as a teachable moment. (it's wonderful that you're a good learner, since i always end up being pedagogical.) first of all, nothing in a _work-product_ will necessarily be in the _final_ product. so that "wonder" shouldn't even come up. indeed, what you _should_ be thinking is "why is this aspect in this work-product?" and "what need is it serving in this case?" second, and of an enduring importance, is the fact that _navigation_ is one of the essential qualities of an electronic-book. by "navigation", i mean 3 specific things: 1. knowing where you are in the e-book. 2. knowing exactly where you want to go. 3. being able to go there quickly and easily. pagenumbers help to keep you grounded; they make it easy to know where you are... they also often make it easy to know where you want to go. they serve as sign-posts... in the electronic text environment, _links_ are the things that make it easy to _jump_ -- i.e., go somewhere quickly and easily... so navigation is why we keep pagenumbers in our "master" file -- because they are _so_ essential in our editing -- and often even in an output-format created from that master. (but only if they won't interfere with usage. still, we keep 'em up until the last moment.) navigation is crucial in a work-product too. which is why i included the links in that file. they are tremendously useful, at times, and you can just ignore 'em the rest of the time. another thing that you can note about this is that pagenumbers help us to formulate _an_intelligent_file-naming_scheme_to_use. notice that i use the prefix of "bhaga" on all of the files associated with this book... (that's what defines it as a coherent entity within my system, which has many books; that 5-letter label also names its _folder_.) after "bhaga" is "f" or "p", depending on if it's a forwardmatter page or a regular page -- i.e., either roman numerals or arabic -- _then_ the pagenumber (padded to 3 digits). so the scan for page 123 is "bhagap123.jpg". and the .html filename is "bhagap123.html". the text filename would be "bhagap123.txt". and the html anchor is likewise "bhagap123". this intelligent naming-scheme means that the book has a dependable predictability, so it's easy to create .html anchors, and links, and to call images, and so on... thus, those pagenumbers in a p-book table-of-contents can be _effortlessly_ turned into .html links, with a minimum of fuss or muss or coding. it's like falling off a log. so it's something that's simultaneously powerful and simple, great navigability at practically zero cost... so the question is: "why _wouldn't_ i do it?" and the answer: "i'd be stupid not to do it." all the links in that huge table-of-contents? i accomplished them with two lines of code: the first one created each of the anchors... the second made the links to those anchors. and did you catch the results of a third line? each of the page-separators is a _link_ that will load the image-scan for that page into a _named_ window, which also means that subsequent clicks on other page-separators load their scan in that same named window. so if you wanted to refer to the page-image to resolve any of those paragraph questions, it would have been extremely easy to do so. this is a little trick which you and i will use later, james, so make sure you play with it. and, again, all of that with one line of code. so these are great examples of the way that it can be fun and easy to code books which are powerful, fast, easy, functional, and fun. and _that's_ what you should be noticing, james, not complaining about "these links which i do not want". there's all kinds of gold injected into the samples i give you, and you should really be panning for it... not all of the nuggets will be as big and obvious and easy to explain as these, but ounce for ounce, they're just as valuable... *** that's page _numbers_. page _breaks_ are another animal. they are useful less often, so most of the time we will not retain them. (but again, we keep 'em as long as we can.) it's probably also important to note that we have a way to "rejoin" paragraphs even while retaining the pagebreaks and pagenumbers. (for instance, that's important when we do the check for balanced doublequote-marks, where the paragraphing exerts an impact.) > A design is beginning to emerge > from everything you've told me. > Looking at your ZML file helped. good. glad to hear that. :+) you have to also realize that there are hundreds of subscribers to this listserve, and you're just a vehicle for this thread, where i'm actually talking to all of them, not just to you individually... so i have to make sure this is still interesting to them, even though they've heard the whole story. (well, at least they _believe_ that they have.) *** the other thing is, i love life. i laugh a lot. i'm always smiling when i write these posts. i have a good time. i enjoy the interaction. so i'm having fun. even when i pretend to complain... (although it does get tiresome when i must repeat myself too frequently!) you don't have to fret over offending me... -bowerbird

1 0

book of james -- 015
by Bowerbird＠aol.com 18 Jan '12

18 Jan '12

james- i wanna make sure that you and i don't get sidetracked by this discussion that's going on about guiguts and .epub... i think it's important to stay focused because -- for one -- you recommended guiguts in your "enlightenment" book, and -- for another -- you've said here that _you_ use it... let me say, right at the outset, that i do respect guiguts... thundergnat -- the original programmer -- did a service for distributed proofreaders that very few will ever match. people who've worked on it since steve, including hunter, have helped as well; guiguts has been invaluable to many. and i mean all of that praise quite sincerely, i really do... at the same time, however, guiguts is long in the tooth... and it never was too user-friendly; it's a _bear_ to install. but the worst thing about it is that it was molded out of the d.p. workflow, which is _not_ something to emulate. so unless you're working inside d.p. -- and maybe even not then -- you definitely do _not_ want to use guiguts. still... as far as the current _discussion_ goes, join in! hunter is _asking_ for advice, so give him some, and he might just be able to improve the resultant output. i'm not too hopeful about that, for reasons i'll append, but i'm not opposed to change, and it could happen... but even if guiguts does "improve", in terms of input for the .epub converter, you still won't want to use it. all of the useful things it does -- and it does many -- are things that you can write your own scripts to do. and scripts you write will work much better, for you. this _specifically_ includes the text-to-html routine... you might or might not decide to adopt my .zml, but even if you don't, i can _guarantee_ that you will have seen enough of the light to realize you can d.i.y. this. so join the fun of the current discussion, and see if you can learn anything from it, but understand that -- in the long run -- guiguts will not matter to you. -bowerbird p.s. so why i am not optimistic about this discussion? 1. we discussed it a million times; it never mattered. 2. marcello doesn't care a whit what anybody thinks. 3. the d.p. postprocessors don't care what you think. 4. postprocessors got recruited as snowflake makers. 5. everyone wants the e-books made to _their_ taste. 6. too many cooks spoil the soup; so then it's spoiled. 7. 3 d.p. committees worked on this; they all gave up. 8. .epub is a moving target; tomorrow, apple moves it. 9. it's impossible for a single .epub to please everybody. 10. end-users will soon demand an ability to customize. so... have fun. but i've got better things to do... :+)

2 1

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