February 2012 - gutvol-d - lists.pglaf.org

Crowdsourcing
by don kretz 19 Feb '12

19 Feb '12

Very interesting. It appears that Wikicommons has implemented a more refined publicly accessible text proofing process, and at least one of PG's texts is part of their subject corpus. Take, for example, PG #32607; particularly the article about diazo compounds. http://www.gutenberg.org/files/32607/32607-h/32607-h.htm#ar23 Notice the embedded images; for example http://www.gutenberg.org/files/32607/32607-h/images/img174b.jpg Here are two versions of the same text on eb.tbicl.org. First, a clone of the PG ebook: http://eb.tbicl.org/vol08/4/#ar23 and then after processing it into an article: http://eb.tbicl.org/diazo-compounds As we have discussed previously, I have added to the page numbers by making them linkable to the page image on TIA. ------------------------------------------------------------- Here is the same article at Wikisource. http://en.wikisource.org/wiki/1911_Encyclop%C3%A6dia_Britannica/Diazo_Compo… You can see that the cropping in the embedded image is identical to PG's. Further, they have included a linkable page number (including an additional one at the heginning of the article.) But it doesn't link the same place. Instead it links to http://en.wikisource.org/wiki/Page%3AEB1911_-_Volume_08.djvu/191 which is their own editing interface for that page. So there it is. The public can proof a text derived almost directly from PG, using an interface that includes the page image and text text editor for matching. (Note to self: the image they use is from a source I wasn't aware of and the quality is pretty good - I need to check http://upload.wikimedia.org/wikipedia/commons/thumb/0/02/EB1911_-_Volume_08… using a well-crafted page url template.) One might choose to repudiate the term "crowdsourcing", but it would be just semantics at that point.

1 0

Re: [gutvol-d] update on zero markup
by Bowerbird＠aol.com 18 Feb '12

18 Feb '12

roger said: > Technology that a developer won't release there's an unfair and inaccurate diss right there. *** i'm busy right now. we'll discuss this next week. -bowerbird

3 2

Re: [gutvol-d] update on zero markup
by Bowerbird＠aol.com 18 Feb '12

18 Feb '12

roger said: > Here's an update on > my "zero markup" experiment. oh roger. roger, roger, roger. what a baffling thread this has been. i don't even know where to begin... i do need to say that i'm very sorry that none of your friends here told you that your "zero markup" is one and the same with my "zen markup"... indeed, for the first years, i even called it "zero markup". (i renamed it "zen" because it's cooler; but "z" always was a mere placeholder, so i can say i'm "two steps past" x.m.l.) so when i read your initial post on this, i wondered if you were just punking me. (i suspect that you do indeed have some friends here on this list, but they did not say anything to wise you up because they were as perplexed by your post as i was.) the language you used was a very close paraphrasing of language _i_ have used, and been using here for 8+ years now... you can see this for yourself if you google "zero markup language", where you'll find lots of cases of that exact same language. but it didn't seem like you were punking. it seemed like you were serious about it. which was even _more_ baffling, honestly. for years, you have dumped on me, roger. not constantly, or extreme, but you _did_. is it really true that -- for all that time -- you had not one clue about the major idea i promulgated on this list that whole time? that's a _shocking_ admission, roger. it says you judged me without considering the entirety of the evidence in front of you. that shows you don't have one bit of class. and, as bad, you don't do your homework. and i don't say any of that in anger, either. it's just a straightforward assessment of it. but it's a damning one. it woulda been better to say you were just punking me than to admit you're ignorant, especially as the ignorance is self-inflicted. and to be ignorant about what i have done _and_ diss me in addition? inexcusable... like i said, i'm sorry i had to be the one to tell you this. i know it woulda been better coming from a friend, and much earlier, when you could have "explained it away" by saying "i thought what i was saying is different from what bowerbird says, but upon further reflection i see that it's not", or something like that... but now you are just plain stuck in the middle of the road, with cars whizzing all about, to and fro... plus you've made yourself look very bad. it's sad. *** as for your success, or lack of it, i'm uncertain. perhaps you haven't worked on it long enough? because i'm getting my z.m.l. to work just fine. and the things that you consider to be obstacles are problems that i solved a long, long time ago. either that or they aren't really problems at all... of course, i'm not navigating marcello's workflow, because -- frankly -- his approach will not work, at least not easily, or in a mutually satisfying way. (marcello doesn't care about mutual satisfaction.) so i'd suggest that you overthrow that workflow. and once you've removed that political obstruction, you'll discover the technical challenges melt away... > I don't know where I'm going with this from here, > if anywhere. that's ok. i can direct z.m.l.'s progress just fine. so i'll take it from here, roger. -bowerbird

2 1

RST/PGTEI/etc
by hmonroe.pglaf＠huntermonroe.com 17 Feb '12

17 Feb '12

Sorry to repeat myself but you can get the RST and PGTEI tools sets (Epubmaker and Gnutenberg Press) installed easily on Windows systems as bundled with guiguts-win here http://sourceforge.net/projects/guiguts/files/guiguts/. There are many interesting development opportunities here: fleshing out a GUI interface to these tools, making it easier for users to generate Epub friendly HTML while provideing warnings when they have not, tweaking the code used to create HTML so it creates PGTEI. Contact me with a sourceforge ID and I can add you as a developer. On the last idea, the first step would be to modify lib\Guiguts\HTMLConvert.pm to replace a line like [code]$textwindow->ntinsert( "$step.0", '<h1>' );[/code] with [code]$textwindow->ntinsert( "$step.0", $chapterheader);[/code] then create a file html.rc with lines like [code]$chapterheader='<h1>';[/code] which gets read in at the start of htmlautoconvert(). Then you create pgtei.rc [code]$chapterheader='<header>';[/code] You would then have a routine to produce TEI that is already far better than the one on the dp web site. A minor change to another routine (is_paragraph_open) could ensure that the <div> for the previous chapter is closed. Hunter

14 95

rewrapping p.g. to an existing scan-set
by Bowerbird＠aol.com 17 Feb '12

17 Feb '12

don said: > I don't understand. c'mon, don. i know you have the chops for good thought. so why don't you demonstrate the capability? p.g. has the scans for the books that d.p. did. but those books do not need to be corrected. for the most part, they are all "good enough". and because p.g. has the scans, those books can point to those scans as their provenance. so there's almost no need to work with them. p.g. does not have scans for the early books, many of 'em classics, which _do_ need fixing. indeed, even those which are "relatively clean" cannot mount credible proof of that accuracy. and because they cannot point to a scan-set, they're completely lacking in any provenance. so these are the books which need attention... specifically, they need to (1) find a scan-set to be their parent, and (2) get in step with it. for (2), their lines will need to be rewrapped. now, don, what is it, about all of that, which you couldn't have figured out all by yourself? *** carlo said: > Do nothing. that's your best advice? seriously? while you're sitting on your hands, someone will be eating your lunch. -bowerbird

3 2

rewrapping p.g. to an existing scan-set
by Bowerbird＠aol.com 17 Feb '12

17 Feb '12

does every conversation here _have_ to be tedious? *** ok, let me 'splain you... first, if you're gonna wait for users to report errors, instead of proactively seeking them and fixing them, you might as well _give_ your lunch to someone else and then trot off to your mat so you can take a nap. if the tech world has taught us anything at all lately, it's that you'd better improve your product offerings _yourself_, before any of your competitors do it first. project gutenberg e-texts have one powerful benefit -- they are cleaner than many other texts out there. (but don't look back, because google's coming fast.) the e-texts also have 2 significant related liabilities -- they have no provenance, so the accuracy isn't verifiable. this means a competitor can knock you out quite easily, and even use your own text to do it. they simply take it, hook it up to a scan-set, and do a proof via comparison. voila. they have _cleaner_ text. which _has_ provenance. then they point to your absence of any provenance, and draw attention to your errors -- which they've fixed! -- and boom, they've done a number on your reputation... do that a couple dozen times, with high-profile classics, in a way that goes viral, and the blow would be crushing. crushing. that's if they beat you to the punch. they eat your lunch. there's your top-down seminar... and now for the bottom-up part. if you're going to build a system capable of pulling up an arbitrary page from an arbitrary book in response to an error-report regarding some of the text on that page, then you might as well blossom it into a system that will serve to correct all the text on every page of every book. you dig? -bowerbird

1 0

rewrapping p.g. to an existing scan-set
by Bowerbird＠aol.com 17 Feb '12

17 Feb '12

carlo, perhaps the answer was hidden in your last post, but let me present the situation, and ask the question directly... the project gutenberg e-text for "pride and prejudice" is relatively accurate, but completely lacking provenance... the internet archive scan-set for "pride and prejudice" is self-documenting, but the o.c.r. text from it is abysmal... even though, as you point out, they are different editions, it seems we could use these problems to offset each other. so... how would you suggest that we go about doing that? we can, of course, just send the internet archive stuff through d.p., but that wouldn't take advantage of our already-proofed-and-relatively-accurate p.g. e-text. so, carlo, what would you recommend, specifically? -bowerbird

2 1

rewrapping p.g. to an existing scan-set
by Bowerbird＠aol.com 17 Feb '12

17 Feb '12

don said: > You can link to images at TIA and Google by page number. so what? don, we're not doing our first round of brainstorming. and we are certainly not grasping at straws, hoping for an idea -- any idea -- no matter how vague it might be. we've got an actual text-file we're using as an example, and an actual scan-set representing that same book, so the best way to play this little game is to mount a demo. -bowerbird

2 1

rewrapping p.g. to an existing scan-set
by Bowerbird＠aol.com 17 Feb '12

17 Feb '12

don said: > how about can you be more cryptic? -bowerbird

2 1

rewrapping p.g. to an existing scan-set
by Bowerbird＠aol.com 17 Feb '12

17 Feb '12

don said: > Assume PG has the images. if p.g. has the images, odds are that the text doesn't need to be corrected. (it certainly wouldn't be cost-efficient; few of us object to the final 50 errors.) plus you already have the provenance. so that's a _totally_ different situation. i mean, you might care about such cases. but i can't imagine that many people will. so there's no volunteers to work with you. -bowerbird

2 1