February 2011 - gutvol-d - lists.pglaf.org

Re: cryptic star-trek reference again
by Bowerbird＠aol.com 21 Feb '11

21 Feb '11

michael said: > However, if it is NOT "all over," then perhaps I should say > that BB took my advice and "let days seem like weeks." > And therefore, because day seemed like weeks, > the absence a day might have made was not fatal. um, let me apply my own windex. if one day passes without some feedback on this app that i have just put up, then the whole thing is over... yes, even one measly day. i'm not gonna sit around for a week to get the message. i've already been sitting around waiting for 7 years now. the rest of the world caught up, so i can't wait any more. they are interested, even if _you_people_ can't care less. so it's _every_ day, or nothing. there are 300 people on this listserve. if each one takes responsibility for just one day, you can have me doing work on this app for the next 10 months, which would generate one heckuva tool, for sure... but if no one responds within _each_ calendar day, _every_day_, the whole thing comes crashing down. i will still build my tool -- i'll build it for the world -- but i will no longer be entertaining your suggestions. i hope this makes things totally clear. -bowerbird

1 0

Re: time after time
by Bowerbird＠aol.com 21 Feb '11

21 Feb '11

bastien said: > Can you share some usable code? you've been around long enough to know the answer to that. so, am i going to just hand over all my code to my enemies -- the mean people who called me names for many years, and said that what i was suggesting was "impossible" -- so they can use my code against me, for their own agenda, once they decided -- as indeed, they have now decided -- that i was totally correct all along, and that the only viable workflow for project gutenberg is to use light markup? no. no. a thousand times no. and throw in a "screw you" too. am i willing to help guide people who are willing to work one-tenth as hard as i have? well, yes, i most certainly am. will i allow these people to be stuck for any period of time, follow an unproductive path, or otherwise waste their time? no, absolutely not. that's what a guide is for -- to _guide_. will i provide code fragments to these people so that they can see how it was done, even if they cannot use the code directly, because it's in basic, and they use something else? yes i positively will, without any hesitation at all. even deeper, will i explain carefully just exactly _why_ that approach to the code is the very best one to take, and why other perspectives are flawed in one way or another? yes, again, i will indeed provide such careful explanations. am i going to give you a bucket of fish? no, i'm not. am i willing to teach you how to go fishing? yes, i am. -bowerbird

1 0

Re: time after time
by Bowerbird＠aol.com 20 Feb '11

20 Feb '11

jana said: > Now, I get stuff like this: oops. ok, i'll fix that. good eye. > Oh ick, are you serious about that background image?... well, i like it. plus it's funny when people squirm. but i will change it for you... :+) > What are those @@@ lines? > Whatever they might be, they don't belong > into the output, so get rid of them. well, they are scene-breaks, so yes, they _do_ belong in the output. but you don't like "@@@"? ok, i'll change it to the standard "***" for you... > Or, how about letting the user choose > background and foreground colours? i'm all in favor of user-choice. and isn't it very interesting, folks, that once you start _asking_ people what they actually prefer, you learn that they have all kinds of preferences lurking there. once you give them control, they want to control the style of the quotes, and the style of dashes, and the background and foreground colors, and even the way that they choose to _spell_ "colours", all kinds of things. and yes, of course, if you just feed 'em something -- heck, just about anything -- they will eat it, and not complain much, and maybe you _think_ you are "making them happy" with that, but it's only because you offered them zero choice... if they are allowed to seize the power to control the display to their own preferences, they _will_ do so... there is an important lesson to be learned here... *** > weren't you going to allow us to > load in our own text to the app? yes, and a limited version of that is now there. in the search-field, type "doload140" -- one word, no spaces, no quotes -- and then press [enter]... give it a minute, because it's loading in a copy of "the jungle" directly from project gutenberg, pg#140. then type "dosplit" in the search-field, and press [enter] to have the file split up into its sections... again, give it couple minutes, because it has to upload the file from you, and then split it, and this is a rather big book (800k), so it takes time. when it has finished, it will display the first section. if you examine the pop-up menu, you will discover something strange. it skips from chapter 21 to chapter 24. also, it ends at chapter 26, which might not seem strange unless you know that this book actually contains _31_ chapters. the problem here is that pg#140 wasn't consistent in following the rule that chapter-headers are preceded by 4 blank lines. you can verify this by looking at pg#140 yourself: > http://www.gutenberg.org/files/140/140.txt indeed, what you will find is that only chapter 1 and chapter 21 have 4 blank lines preceding them. (and chapter 1 is inconsistent too, because it has 5.) my routine was still able to find most of the chapters in spite of this because it required only 3 blank lines. nonetheless, it still missed several chapters. so let's do this. reload the text from project gutenberg: type "doload140" in the search-field and press [enter]. now, instead of typing "dosplit", type "docorrection" in the search-field and press [enter]. this routine will use a different rule for splitting the file into sections. it will look for the word "chapter" at the start of a line. because of these inconsistencies in the p.g. library, it's hard to make rules to handle all the edge cases. it's better to find the inconsistencies and correct them, so the rules we compose can be as simple as possible. especially for fundamental necessities, like finding the sections of the book, we want the rules to be simple... it was because project gutenberg absolutely refused to ensure consistency in its files that i gave up on it. -bowerbird

1 0

Re: time after time
by Morasch＠aol.com 20 Feb '11

20 Feb '11

weren't you going to allow us to load in our own text to the app?

1 0

Re: time after time
by Bowerbird＠aol.com 20 Feb '11

20 Feb '11

new version up at same location: > http://76.74.254.239/~bowerbird/a.fcgi/ proper italics now. font changed from comic sans. em-dashes utilized. curly-doublequotes utilized. found search-terms colorized. some cosmetic changes. most of that stuff was requested by jana. some of you might not know her, because she hasn't posted much here, but she is a hard-working volunteer over at d.p. who helps lots of mac users with their problems. -bowerbird

2 1

Re: time after time
by Travis Siegel 20 Feb '11

20 Feb '11

> >> >>> On Feb 19, 2011, at 5:16 PM, Bowerbird(a)aol.com wrote: >>> >>>> travis said: >>>>> Only your page does not conform >>>>> to the html guidelines for accessibility >>>> >>>> good point, travis. i'm more worried about the >>>> functionality now, but catch me on that later, ok? >>>> >>>> but if you think it could, or should, do something more, >>>> i invite you to contribute some suggestions to that effect. >>>> >>>> Which is exactly what this is. My experience is that if >>>> accessibility isn't designed from the start, often times it's >>>> overlooked entirely, or deemed too expensive/time consuming to >>>> engineer into the product later. I'm not saying labeling buttons >>>> couldn't be done later, but I am saying that if you ignore possible >>>> accessibility issues now, you'll wind up with a product that is (at >>>> best) a pain to use for those who need accessible applications, and >>>> (at worst) a completely useless >>> piece of software, because their screen reader won't work with it, >>> or >>> they can't resize text, or something else that is dead simple to >>> incorporate at the start, but nearly impossible to reverse engineer >>> into the product when it's complete. Just pointing out possible >>> pitfalls at the getgo so you aren't swamped with requests later to >>> "fix" things later that you didn't consider in the beginning. >>> >>>> actually, it's more like a demonstration that z.m.l. format >>>> can be used as a "master" that can churn out other formats, >>>> which is an issue that p.g. and d.p. have been struggling with. >>> >>> Which (I thought) was exactly what xml was supposed to be used for. >>> >>> (not complaining about the use of zml, only pointing out that other >>> options already exist. >> > > >

1 0

Re: time after time
by Travis Siegel 20 Feb '11

20 Feb '11

>>> >>> On Feb 19, 2011, at 5:58 PM, Bowerbird(a)aol.com wrote: >>> >>>> >>>> >>>>> How about real em-dashes instead of double-hyphens? >>>> >>>> h=replaceall(h,"--",chr(209)) >>>> >>>> >>>>> How about curly quotes? >>>> >>>> h=replaceall(h," "+chr(34)," "+chr(210)) >>>> h=replaceall(h,chr(10)+chr(34),chr(10)+chr(210)) >>>> >>>> h=replaceall(h,chr(34)+" ",chr(211)+" ") >>>> h=replaceall(h,chr(34)+chr(10),chr(210)+chr(10)) >>>> >>>> Is there really a need to do all that? Not complaining, only >>>> wondering myself. Screen readers don't generally take well to >>>> upper >>>> ascii characters. Sure, they'll (generally) tell what the >>>> characters are, but often times, it changes the pronounciation of >>>> the word, because it tries to incorporate the symbol into the word >>>> it's surrounding. Honestly, double hyphens are fine with me :) >>>> (and, for some reason, the rss specifications forbid such >>>> characters), though I'm not sure why that is, perhaps for widest >>>> compatibility with all rss readers, including those that can >>>> display >>>> text only. >>> >>> >>>> >>>>> It doesn't help me much that I've found some text >>>>> when I still have to read through the whole chapter >>>>> to actually find the passage I'm looking for. >>>> >>>> i would say you seem to have forgotten the fact that >>>> you can just use the search function in your browser. >>>> >>>> the app finds a page which contains the search-term, >>>> and the browser tells you where on the page it exists. >>>> it is a bit clumsy, but not as bad as you made it sound. >>>> >>>> >>> Yeah, but it's not difficult to spit out a modified page containing >>> the link to the searched for text if it comes to that, (which, you >>> did >>> say on the next line, so ...) >>> >>>> p.p.s. here is the current text-to-html conversion routine... >>>> >>>> h=session.gc(x) >>>> h=replacelineendings(h,chr(10)) >>>> h=trim(h) >>>> h="<h2>"+h >>>> h=replaceall(h,chr(10)+" ",chr(10)+" ") >>>> h=replaceall(h,chr(10)+chr(10),chr(10)+""+chr(10)) >>>> h=replace(h,chr(10)+""+chr(10),"</h2>"+chr(10)+" ") >>>> h=replaceall(h," "," ") >>>> h=replaceall(h,chr(10)+" >>>> ",chr(10)+"________________") >>>> h=replaceall(h,chr(10)+" ",chr(10)+"_______________") >>>> h=replaceall(h,chr(10)+" ",chr(10)+"______________") >>>> h=replaceall(h,chr(10)+" ",chr(10)+"_____________") >>>> h=replaceall(h,chr(10)+" ",chr(10)+"____________") >>>> h=replaceall(h,chr(10)+" ",chr(10)+"___________") >>>> h=replaceall(h,chr(10)+" ",chr(10)+"__________") >>>> h=replaceall(h,chr(10)+" ",chr(10)+"_________") >>>> h=replaceall(h,chr(10)+" ",chr(10)+"________") >>>> h=replaceall(h,chr(10)+" ",chr(10)+"_______") >>>> h=replaceall(h,chr(10)+" ",chr(10)+"______") >>>> h=replaceall(h,chr(10)+" ",chr(10)+"_____") >>>> h=replaceall(h,chr(10)+" ",chr(10)+"____") >>>> h=replaceall(h,chr(10)+" ",chr(10)+"___") >>>> h=replaceall(h,chr(10)+" ",chr(10)+"__") >>>> h=topmatter.text+h+chr(10)+chr(10)+"</body></html>" >>>> _______________________________________________ >>> >>> Which a simple nl2br would solve nicely if this were php. :) >>> Again, not finding fault, simply pointing out another method. >>> I've never tackled something of this complexity, but perhaps a >>> second >>> app written in php would be a good way to compare and contrast >>> different ways of doing things. >>> Of course, it could wind up being an exercise in futility, but heck, >>> I'm always willing to help, and php is what I know, so ... >> > > >

1 0

Re: time after time
by Travis Siegel 20 Feb '11

20 Feb '11

> >> >>> On Feb 19, 2011, at 5:16 PM, Bowerbird(a)aol.com wrote: >>> >>>> travis said: >>>>> Only your page does not conform >>>>> to the html guidelines for accessibility >>>> >>>> good point, travis. i'm more worried about the >>>> functionality now, but catch me on that later, ok? >>>> >>>> but if you think it could, or should, do something more, >>>> i invite you to contribute some suggestions to that effect. >>>> >>>> Which is exactly what this is. My experience is that if >>>> accessibility isn't designed from the start, often times it's >>>> overlooked entirely, or deemed too expensive/time consuming to >>>> engineer into the product later. I'm not saying labeling buttons >>>> couldn't be done later, but I am saying that if you ignore possible >>>> accessibility issues now, you'll wind up with a product that is (at >>>> best) a pain to use for those who need accessible applications, and >>>> (at worst) a completely useless >>> piece of software, because their screen reader won't work with it, >>> or >>> they can't resize text, or something else that is dead simple to >>> incorporate at the start, but nearly impossible to reverse engineer >>> into the product when it's complete. Just pointing out possible >>> pitfalls at the getgo so you aren't swamped with requests later to >>> "fix" things later that you didn't consider in the beginning. >>> >>>> actually, it's more like a demonstration that z.m.l. format >>>> can be used as a "master" that can churn out other formats, >>>> which is an issue that p.g. and d.p. have been struggling with. >>> >>> Which (I thought) was exactly what xml was supposed to be used for. >>> >>> (not complaining about the use of zml, only pointing out that other >>> options already exist. >> > > >

1 0

Re: time after time
by Bowerbird＠aol.com 19 Feb '11

19 Feb '11

jana said: > How about italics in the rendered text > instead of the ugly underscores? well, yes, i did say the .html conversion was still crude... i've appended the complete conversion routine currently. to do the italics would take this: h=replaceall(h," _"," ") h=replaceall(h,chr(10)+"_",chr(10)+"") h=replaceall(h,"_ "," ") h=replaceall(h,"_"+chr(10),""+chr(10)) however, this only works because i know that i've used whitespace in a certain way around the italics markers, and it might not be robust with respect to other usages. also, this might conflict with other parts of the converter, so we need to do close examination to see if that happens. in the end, these reg-ex type changes are _not_ the best way to do the conversion. it takes a serious line-by-line analysis. but in order to get something up quickly, i'm doing this now. > How about using _any other_ font for the body text? > Please? Why not leave it at the user's default rather than > force something as ugly as Comic Sans? personally, i love comic sans. :+) plus it's entertaining to see other people squirm at it. but here's the line from the stylesheet. > font-family: "Comic Sans MS","Lucida Sans Unicode", > "Lucida Grande","Verdana",sans-serif; tell me how you want me to change it, and i will. and eventually, we'll let the user decide the font, using any font on their machine. > How about real em-dashes instead of double-hyphens? h=replaceall(h,"--",chr(209)) > How about curly quotes? h=replaceall(h," "+chr(34)," "+chr(210)) h=replaceall(h,chr(10)+chr(34),chr(10)+chr(210)) h=replaceall(h,chr(34)+" ",chr(211)+" ") h=replaceall(h,chr(34)+chr(10),chr(210)+chr(10)) > How about links from the table of contents to the chapters? that's a little bit more complex. but doable. still, let's save that for sometime next week... > It doesn't help me much that I've found some text > when I still have to read through the whole chapter > to actually find the passage I'm looking for. i would say you seem to have forgotten the fact that you can just use the search function in your browser. the app finds a page which contains the search-term, and the browser tells you where on the page it exists. it is a bit clumsy, but not as bad as you made it sound. > How about jumping to the right spot > in the chapter after the search? that would be very difficult. we'd have to put an anchor in the text. ok, i guess that's not really difficult at all... sometime next week for that. > Or highlighting the text somehow? that's very doable. we'll do that first. also next week. > Why does it only sometimes save my changes? I had to > edit the dedication twice to have it pick up my changes. i dunno. did you hit "save" both times? did you get the message confirming that the file had actually been saved? as i said, sometimes the response-time can be sluggish... perhaps you simply didn't wait long enough the first time? > Jana thanks for your input. collaboration isn't hard at all, is it? -bowerbird p.s. might wanna save multiple points for multiple days; no harm going slow, but we don't want to get burned out. p.p.s. here is the current text-to-html conversion routine... h=session.gc(x) h=replacelineendings(h,chr(10)) h=trim(h) h="<h2>"+h h=replaceall(h,chr(10)+" ",chr(10)+" ") h=replaceall(h,chr(10)+chr(10),chr(10)+""+chr(10)) h=replace(h,chr(10)+""+chr(10),"</h2>"+chr(10)+" ") h=replaceall(h," "," ") h=replaceall(h,chr(10)+" ",chr(10)+"________________") h=replaceall(h,chr(10)+" ",chr(10)+"_______________") h=replaceall(h,chr(10)+" ",chr(10)+"______________") h=replaceall(h,chr(10)+" ",chr(10)+"_____________") h=replaceall(h,chr(10)+" ",chr(10)+"____________") h=replaceall(h,chr(10)+" ",chr(10)+"___________") h=replaceall(h,chr(10)+" ",chr(10)+"__________") h=replaceall(h,chr(10)+" ",chr(10)+"_________") h=replaceall(h,chr(10)+" ",chr(10)+"________") h=replaceall(h,chr(10)+" ",chr(10)+"_______") h=replaceall(h,chr(10)+" ",chr(10)+"______") h=replaceall(h,chr(10)+" ",chr(10)+"_____") h=replaceall(h,chr(10)+" ",chr(10)+"____") h=replaceall(h,chr(10)+" ",chr(10)+"___") h=replaceall(h,chr(10)+" ",chr(10)+"__") h=topmatter.text+h+chr(10)+chr(10)+"</body></html>"

1 0

Re: time after time
by Bowerbird＠aol.com 19 Feb '11

19 Feb '11

travis said: > Only your page does not conform > to the html guidelines for accessibility good point, travis. i'm more worried about the functionality now, but catch me on that later, ok? > You'll need to place text labels > on the rest of the buttons if you want this > to be a truly universal replacement for the plain > text that all project gutenberg files already contain. well, we haven't really decided exactly what role this app will play -- because that's part of the whole process here -- but it won't be as a "replacement" for p.g. plain-text files. initially, i'd say the goal is to take an input-file and then churn out the .html version, and an .epub, and a .mobi... so this app is just an intermediary, a temporary-use tool, which is why i think it's ok to put off accessibility issues... but if you think it could, or should, do something more, i invite you to contribute some suggestions to that effect. > a replacement for the current method of doing things actually, it's more like a demonstration that z.m.l. format can be used as a "master" that can churn out other formats, which is an issue that p.g. and d.p. have been struggling with. -bowerbird

1 0