Bookish. it's like redactle but for books. beta testing / bugs etc

threw this together over the LONG jubilee weekend and it's now mainly working but some people trying it and spotting bugs would be useful.

http://www.koogy.clara.co.uk/bookish/

it's very much 1999 javascript, the only kind i know.

for instance, are there any more simple words that have to go in the list? this is an easy way of tweaking the difficulty.

A way to scroll to the words that have hits would be useful.

Also automatically clearing the previous guess from the text box.

Yay, this is a fun addition to the word game canon. I'm very much a sightseer (which is why I never put my redactle scores up) so it took me a lot of guesses, but still felt nice and simple to get.

My only big criticism is that the blank size and the word size don't seem to fit - I don't do letter-counting but it still seems a bit odd to have what looks like a big blank and then reveal that it was "is" all along. Might be a case of tweaking fonts or something? And I guess a minor thing would be to add some more words like "is" to the auto-fill, as you suggested. I don't mind having to put them in, tbh, but I bet other people would appreciate it a lot.

yeah, the standard full block character is wide so it tends to look larger. i'm trying to think of a way to make the blocks uncountable, maybe by making them images and maybe by making them word length +/- 1

i was worried it'd be too easy, '#### ####' = moby dick, but i'm not sure it doesn't tend itself more to strolling than gold

redactle common word list is here: https://www.redactle.com/commonWords.js but i don't like it. mine is currently 'a', 'and', 'in', 'be', 'are', 'so', 'the', 'this', 'that', 'then', which is ridiculous the other way perhaps.

> Also automatically clearing the previous guess from the text box

should be there now - https://github.com/acdean/bookish/commit/717bad2e24bcfe3674436ca0490bf771aea6c956

scrolling to words is less easy, because there are multiple hits. you can scroll to a tagged element, i think, but there may be many tagged elements. i can make it scroll to the first one, will have to think about the rest. (it is one of the redactle features i like)

i've been adding a few more texts for the future but i keep coming across things i've not thought about - foreign letters, poems, hex-encoded entities.

Got bookish #1 (I think?) in 3!

oh, the one thing i forgot is to say which game was being played...

(the files are dated, games/20220605.js etc and it injects the relevant one which in turn sets 3 variables. gets around CORS and works with local files and servers)

if you look at http://www.koogy.clara.co.uk/bookish/games.html you can play older games.

need to look into zipping those files, they are about 10x bigger than i was expecting.

I'm much more likely to play this than Redactle - good fun, even though it took me lots of guesses!

have added a game indicator.

given that this is all flat files, i wonder if i can get it to run from github pages, would save me having to ftp to my personal site.

This was fun - thanks and well made!

I felt totally at sea trying to guess the words, to be honest - I was picking away at common words but still not forming any idea of themes or useful avenues to take further guesses in. I got there from a lucky bolt of inspiration regarding the lengths of what looked like (and was) the author's name, so removing/fudging the lengths would have totally screwed me, which is possibly fair but might also leave me totally without a way in, so I'm not sure how I feel about that!

Seconding Leee's suggestion that the way you can click on a successfully guessed word in Redactle and be taken to successive occurrences of it is very useful. No, I have no idea how that works; sorry if it's a lot of work...

looked excitedly just now to see if it's changed over automatically but it's not gone midnight UTC so it's still the same puzzle

can think of several ways of doing the redacting, some of which make counting impossible, some of which make it easier (in fact looking just now on the phone you can see the individual blocks). maybe have this as a choice.

there's a scrollintoview method on html now which will make highlighting easier, i think. just need to give each highlight a separate id...

and the game files are ~20k each or 50 to a megabyte so I'm not sure why I'm worried. will probably run out of books before space

apostrophes, how are they? currently you have to guess them, so you have to type "don't" or it won't match. one of the puzzles has one in the title. the alternative is ?????'? everywhere and having to guess s or t or ve separately which was worse. Gutenberg uses rsquo (right single quote) for them all which i have to change to ' manually, which is a pain (especially when there are proper lsquo / rsquo pairs in the file too)

i didn't get yesterday's (the secret garden), i've read it at least twice but nearly 40 years ago! i guess the author names was probably a big hint but i couldn't think of any to match. struggling with today's too! i wonder what a good strategy is, i'm just typing in loads of common words...

yeah, the standard full block character is wide so it tends to look larger. i'm trying to think of a way to make the blocks uncountable, maybe by making them images and maybe by making them word length +/- 1

don't obfuscate word length, it's a helpful tactic! does redactle use smaller blocks? it's definitely confusing when a huge looking word turns out to be four letters long.

just got today's in 79 - saw what looked like a two letter character name and entered the first that came into my head, which was enough to give it away.

> in 79

it does seem harder, except when the titles and authors are obvious - #### #### = moby dick, obv. maybe a middle way would be to obfuscate the titles, leave the rest alone.

i checked the block width and 20 blocks was as long as
123456780123456789012345
so 25% wider. i think i can change the letter spacing or use black on black letters rather than blocks.

just got bitten by the apostrophe as separate & redacted character for 20220601 - also it hasn't given me the win after guessing the full title.

20220602 was quite fun, working out the style of narration and getting it from there - still took about 70.

plans for storing state? i've lost progress a couple of times by accidentally pressing 'back'.

Couple of thoughts. Be able to click on words you’ve successfully guessed rather than having to scroll through the page.

Secondly don’t know what I did but it cleared and reset my guesses. Fairly sure I didn’t hit back.

I didn’t solve it, but is it just the title or title and author?

> plans for storing state?

save points are for wimps. (i will try and add a warning about navigating away, that might be easiest. cookies probably require sending something and currrently it just updates the current page, doesn't re-request anything)

dan, jump to highlights is tonight's coding task, i think i have a way.

yeah, just the title. but i may've messed up and included a reserved word in the title which it won't let you guess and if you can't guess it it then you can't win, as ledge has pointed out. (but i thought i'd tested all the games?)

ha, yes, 20220601, it's not the apostrophe, it's the 'in'. i have plans to sort that (and save myself having to specify the title separately from the text - by parsing the h1 contents)

have fixed 20220601, although getting your browser to pick up the new file is surprisingly hard

Got this one in two - the two it took to type in the title. Mainly because of guessing the author's name - a three-letter middle name isn't that common, so Louisa May Alcott came to mind right away, and then it was just a matter of counting letters.

oops sorry

was obviously supposed to be spoiler tagged, sorry about that

it's now mostly scrolling to highlight.

it needs to highlight both the current highlight and the related guess but that's harder than it sounds - because it means de-highlighting the previous value.

also realised that you can guess empty strings... 8(

0610 has 'façade' in it. do i dumb it down? would be unguessable otherwise, unless you type in the cedilla...

Fixed your spoiler tags, LD.

- stopped the empty guesses.
- changed the redacted squares.
- now warns when you try and navigate away.
- title is now calculated from the contents of the h1 tag, which saves me a job and lessens the mistakes.
- added ghost text in input box
- introduced new harder-to-fix problems...

I didn't get today's despite getting about half the words including the first word of the title. I like the numbering redacted squares on highlight though!

pure accident. i was hoping they wouldn't show (they are less clear on mac firefox i noticed)

yes, hard one today, not something i've read myself. and not one of the author's best known works. (or maybe it is, it's in the PG top 75 downloads at least.)

i noticed that there's no ?s in the text...

ok, not so much time tonight but... some changes:

entering 'answer please' as a guess will reveal the answer

now scrolls to the top when you win (whereas it used to scroll to the first instance of the new guess which was always the title, cutting off the WINNER message) ((i need to turn that off generally, it's not correct))

sets focus to textbox on load

#20220608 in 4, my first perfect :)

Are the guesses case sensitive? I assume so as title only accepted it when capitalised.

not for me.

no, everything goes through tolowercase before matching to catch words in titles and at the starts of lines

btw yesterday's was The Happy Prince by Oscar Wilde which is a short story - that was the entire thing iirc

not so obvious today. and again something i haven't read.

spent yesterday evening trying to get the layout better but despite my efforts it's still about 105% the size of the screen and things move about in bad ways. i wonder if it's whitespace after the divs?

also fixed ampersands and converted a couple more books into games. getting to the point where it needs a refactor though.

#20220609 372 guesses! The era and location were pretty clear long before that so if I'd given more thought to possible authors I might have got it a lot sooner.

it's kinda of annoying that i don't get to try and solve them myself.

maybe when i've added a few more i'll be able to go back and do random backissues and not remember them.

more damage today. oh well.

did a noddy, dynamic Previous page, which should update daily - http://www.koogy.clara.co.uk/bookish/previous.html

surprisingly it wasn't much code because the data class is clever enough to know that date - 10 is the previous month (although date - 0 is yesterday, somehow). uses a *string* compare to make sure it doesn't go back beyond 20220601, which is lol.

Fun one today, got it in 70.

Got it in the end, 217 guesses. Unlike Redactle, where you can try topics (book, country, animal etc) finding it a lot more difficult to narrow these down. Would like to see a lot of the two letter words (it, to, by….) filled in already.

yeah my first 20 or more guesses are usually pretty much the same - all two letter words, pronouns, is/can/have & conjugations, there, then, say, said, why where who etc...

i was worried they would make the titles too obvious too often. but i could easily add some more. might do some analysis this weekend. (plan is to define enough games to fill out the rest of the month and then i can use that set as input)

227 today, cheated in the end when I got a half completed sentence. Would never have got it otherwise as I didn’t know the book or the author.

these are all still from the top 75 downloads at project Gutenberg. yes, one i haven't read, but, oddly it was mentioned in the podcast i was listening to this morning.

I got the authors name straight away, struggling to remember the title.

I remembered. definitely an odd entry for the top 75 and not one you could work out if you didn't know it.

the weekly top 100 or so can be a bit faddish, i find. and there are some obvious things that aren't in it.

anyway, i have another 20 games ready to go, another 20 or so downloaded and awaiting editing. another 20 things on my want-to-read list 8(

     1	   6373 the
     2	   3757 and
     3	   3144 of
     4	   2730 to
     5	   2633 a
     6	   1876 in
     7	   1773 i
     8	   1523 he
     9	   1483 was
    10	   1358 it
    11	   1320 that
    12	    881 his
    13	    874 with
    14	    871 for
    15	    853 as
    16	    832 is
    17	    781 had
    18	    772 you
    19	    764 not
    20	    698 at
    21	    683 but
    22	    664 on
    23	    577 be
    24	    569 all
    25	    564 we
    26	    548 her
    27	    536 have
    28	    520 my
    29	    491 they
    30	    468 by
    31	    457 there
    32	    453 she
    33	    450 him
    34	    442 so
    35	    437 which
    36	    427 this
    37	    426 me
    38	    397 one
    39	    392 from
    40	    388 were
    41	    383 said
    42	    359 no
    43	    352 are
    44	    344 or
    45	    326 when
    46	    325 out
    47	    323 would
    48	    310 what
    49	    305 who
    50	    301 them

again i'm thinking about difficulty levels... top 45 for easy mode, 30 for medium, 15 for hard...

yeah Carmilla, i guess the easy way in was the author's name, which is quite odd given the 4 words.

today's is easier

I got today's in 7 after rattling off some two letter words without thinking, should've been a bit sharper.

> today's is easier

especially with the 30 free words now. 8(

(i think ledge got in before the update)

Honestly, as a player who just plays for the fun of it, rather than getting the best score, I'd actually prefer *no* words filled in at the start. I don't mind typing in filler words, but trying to work out which ones are needed and which ones aren't is slightly annoying to me.

ha, that's like the opposite of what other people want...

i will add options (eventually). need to learn how to store options first though...

(i don't disagree, i mean today's had 60% of the title in the free words)

Yeah, I figured I'd be in the minority in that opinion. It's just a bit weird if "he" is already filled in but you still have to put in "his" etc. One option is to fill in more of the words, but another option is to fill in way fewer. Maybe just the articles, the basic conjunctions ('and'/'but') and participles of 'to be'. That would still risk giving away some titles ('The _____ is a ______ ______' for instance) but would be an understandable format for why those particular filler words are given.

And I will say that 'The ____ of the ____' is fairly nondescript and could apply to several titles! Obviously the lower-score seekers will pinpoint those possibilities and do the letter-counting thing to narrow it down, but I played happily for a good while before even attempting the title (it was fairly simple to get, true).

I wonder if you could leave the title fully redacted even if it has free filler words in, or if that would be too confusing.

yeah, i did think of a fully redacted title but it'd require you to enter reserved words to reveal them, which is currently trapped.

(um, could also remove reserved words from the internal title array, you won't be able to enter them and won't ever see them until it's solved, which seems a bit off)

today has mostly been trying to stop the unwanted scrolling. turns out the page width isn't 100% because of the vertical scroll bars. and using that space adds a horizontal scroll bar... i am getting there...

(and adding games up to the end of the month, and beyond)

new version at new url - https://acdean.github.io/bookish/ - let github do the hosting, also easier for me to update - no ftp...

cracked the layout, i think. boxes are the right size and scroll as they should (nice new purple scroll markers too). some of the text needs resizing (the guesses list) and the about and previous buttons are placeholders. html still not great for this layout stuff.

this version also has the hard redacting. no counting characters for you!

and all of a sudden it was a quarter to midnight...

https://acdean.github.io/bookish/

now with options - add as many words as you like, choose the redacting block type (3 different styles). choose from the previous games. marvel at the finally centred list of guesses.

https://acdean.github.io/bookish/index.html <- same link not hindered by ilx hyperlink bug!

I like the options! Yesterday's was a gimme, today's... not so much. I guessed le and la at 230 but forgot about les (and didn't linger on the title or author) for another 250 guesses.

these aren't obscure books by any means and it still takes hundreds of guesses, unless the title's obvious.

(although this one starts a bit atypically iirc. the fact there are 3 levels of numbering is maybe a clue that it's a long book, but, yes, it is a bit sneaky)

the recent penguin classics version translates the title, at least in the intro, but the cover is the usual. would 'The Wretched' have helped though?

bookish 20220615
total guesses: 36
total hits: 34
total misses: 2
percentage correct: 94

the 30 words i added at the start really up the percentage because they are pretty much guaranteed to be hits (but not always). in redactle they wouldn't count towards the total or the percentage. maybe i should do the same.

Took me 337, and I got it only after I decided to guess possible authors.

i had to remove the accent 8(

You usually get a bit of a "vibe" off the work after a while (is it spooky? Stuffy? Romantic?) but it took me ages to get anything off this one - took me guessing until I'd revealed "I beheld his black eyes" and then I had it straight away.

I definitely play pretty differently to other people but I confess I did get a great satisfaction out of getting Crime and Punishment in only two guesses.

updates at https://acdean.github.io/bookish/index.html - be sure to hit ctrl-f5 because its a bit stubborn about using the old version otherwise.

no visual changes (oh, i really need to add a version number) but better behaviour of the guess list - click on a previous guess to re-highlight that word. re-entering a word does the same. redactle has this nailed, i am slowly getting there.

20220616 - after nearly 300 guesses I don't know what the hell this is but I like it. (Hope I haven't read it before and forgotten!)

> I don't know what the hell this is but I like it

i predict: you will regret saying this

Just gonna hide my entire response.

fucking lol! I also thought "I don't know what it is but I like it", because it seemed particularly dark and dystopian, and I enjoy dark and dystopian. But then my eyebrow started to raise when I was uncovering that the Bad Big Brother element seemed to be communal life, and I wondered if it was some old right-wingy perspective... took me a long time to work out that it was VERY much a right-wing perspective, and oh the shame of being intrigued by it. The shame, the shame.

Nooooooooooooooooooooooooooooooooooooooo!

I had a feeling that since it was so distinct but ringing no bells that it might be bad. But not so bad. So bad and wrong.

Oh well at least it wasn't mein kampf.

particularly hard because non-english name - ??? ???? - looks like it should be easy to guess but...

i think mein kampf is also in the list

Didn't get todays, I haven't read it in 30+ years; title/author formatting threw me, I thought the title was the author and the subtitle/author the title.

i did toy with making the (little known) subtitle part of the title, so you had to guess it, but am kinda pleased it threw people off anyway (sorry)

i went to test tomorrow's and found i couldn't remember it and got about 200 words in before i did remember. wouldn't've got it otherwise tbh.

(fwiw i always put a br between (title (and subtitle) and author) and (the rest of the text). but not always between book/volume/part/chapter+their names and the next thing)

<h1>title
(<h2>subtitle)
<h2>author
<br/>
<h2>part x
<h2>whatever
<p>it was the best of times, it was the blurst of times...

I don't mind really, it's more fun trying to guess without using title/author - I might start not looking at them at all at the beginning.

have up after 200 today, worried that there was a bug. turns out i just can't remember famous book titles i spent 10 minutes editing about a week ago.

That took me far longer than a book I know and like should have done!

such me as difficult because the title was a name - you weren't going to chance upon it whilst entering English words

struck me as...

I went for 250 guesses without looking at the author or title. Too hard. I guessed don at 296 but got no hits - maybe there was a space at the end, do you trim the input? it worked 30 guesses later, after I tried la and then mancha.

> maybe there was a space at the end

yeah, don't do that!

(fixed. now restricts input to a-z 0-9 and apostrophe)

also fixed something i noticed yesterday - it should scroll to the word in the guess list if you've entered it before

today's obviously by some kook, got it from the title in 68 (I never studied him in philosophy, didn't even recognise the title as one of his)

today's in 287 without ref to title or author, man riding in carriage to strange place with someone warning him of untoward happenings 'to-night'. could've been a bit quicker, i find it usually takes at least 100 boilerplate words before i can start narrowing it down though.

i know the phrase Beyond Good And Evil from the Pop Group song more than the book.

https://www.youtube.com/watch?v=JySQE4axlJY