←2006-04-29 2006-04-30 2006-05-01→ ↑2006 ↑all
00:00:05 * calamari wonders if graue is still here
00:00:59 <calamari> hmm, that checkbox captcha could also be strengthened by assigning the input a random name
00:01:09 <calamari> you might already do that
00:01:22 <calamari> or a name based on the date
00:01:40 <pgimeno> my advantage is that spammers probably don't speak spanish, hehe
00:01:46 <calamari> then they couldn't hardcode for it
00:01:51 <calamari> hehe
00:02:31 <pgimeno> isn't there a captcha plugin for Wikimedia?
00:02:52 <calamari> yes but that is bad to use because it is standard
00:03:25 <calamari> it is preferrable to be nonstandard so that they dont bother to fix it for oner wiki
00:04:36 <pgimeno> captcha images were beaten by means of free pr0n webs AFAIK
00:05:44 <calamari> lol
00:06:31 <calamari> I wonder if sites like driverguide are doing the same
00:12:22 <ihope> Wait... just how were they broken?
00:17:15 <calamari> ,,,,,,,,/ dzzzz fffffffffffffffffffff nb.bbbbbbbbbbn h t; n kyy
00:19:19 <calamari> d 4v 3vggg
00:19:44 <calamari> ggggggnkl
00:20:01 <ihope> Um...
00:20:26 <calamari> my son typed that :)
00:25:10 <ihope> :-)
00:25:22 <ihope> How old is he?
00:26:12 <pgimeno> ihope: "Enter the code in this image to access this free-porn site"; the image comes from the site to crack and the user input is redirected to the captcha field
00:26:36 <pgimeno> there are enough users as to spam as hell
00:27:03 <calamari> ihope: 3 months
00:27:32 <graue> calamari: your OISC interpreter is 10 bytes longer than the longest program I ever wrote directly in machine code
00:27:56 <graue> have you posted it somewhere?
00:28:14 <calamari> no, not yet. I need to test it first.
00:41:13 -!- tgwizard has joined.
00:42:22 <calamari> graue: what do you think of the checkbox idea? it can be very basic, then get more advanced if they ever bother to adapt
00:43:36 <ihope> "Check half these boxes: [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ]"
00:43:59 <calamari> ihope: lol, that's too much work
00:44:31 <calamari> ihope: easier is to implement the changing name code, then have them type a password, etc
00:44:34 <ihope> "Check half these boxes: [ ] [ ]"
00:45:11 <calamari> ihope: wouldn't that be backwards compatible with the original? :)
00:45:39 <lament> "Check half these boxes: [ ] [ ] [ ]"
00:45:52 <pgimeno> "How many boobs does an amazon woman have? [______]"
00:46:22 <lament> "Check half these boxes: [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ]"
00:46:28 <ihope> "What should we do with the drunken sailor? [____________]"
00:46:51 <ihope> lament: I could just check every other one.
00:47:11 <pgimeno> "How many bottles of beer are there in the wall? [___]"
00:47:14 <calamari> lament: pretty sure that would be effective to stop real posts :)
00:47:49 <lament> "Prove the Jordan curve theorem: [_________________]"
00:47:58 <pgimeno> heh
00:48:13 <ihope> "Enter the full text of the Bible: [_]"
00:49:18 <lament> "Enter word 89 on the page 323 of The Manual: [_____]"
00:49:27 <ihope> :-D
00:49:49 <ihope> Find the prime factorization of this number: 740375634795617128280467960974295731425931888892312890849362326389727650340282662768919964196251178439958943305021275853701189680982867331732731089309 00552505116877063299072396380786710086096962537934650563796359
00:49:54 <calamari> lol
00:50:04 <ihope> [___________________________________________________________________________________________________________________________________________________] [___________________________________________________________________________________________________________________________________________________]
00:50:34 <pgimeno> "[___________________________] 42"
00:51:30 <lament> "This is a position from the 1977 World Chess Championship. Kasparov just went Nxb6. Where should you play? [__]"
00:51:34 <pgimeno> which is, incidentally, the number of bytes of calamari's interpreter
00:51:59 <ihope> "Enter a string that matches the regexes /a.*/ and /b.*/: [_______]"
00:53:19 <lament> "Enter the number of virgins that await the faithful in heaven: [__]"
00:55:15 <ihope> "Enter the first digit of the last prime number: [_]"
00:55:41 <lament> "Enter the last digit of the first prime number: [_]"
00:56:03 <pgimeno> that reminds me of a recent news article in the spanish press
00:56:57 <pgimeno> it said something like "X and Y have found that the number Pi has MORE THAN 300,000,000 DIGITS"
00:57:13 <pgimeno> so, "Enter the last digit of pi: [_]" would do
00:57:18 -!- tgwizard has quit ("Leaving").
00:58:54 <pgimeno> someone should have said the reporter that the decimal expression of 1/3 has... more than 400,000,000!
00:59:09 <pgimeno> (even if ! is taken as a factorial sign)
01:01:30 <lament> dr. evil pinky-in-mouth gesture
01:01:48 <lament> three hundred MILLION digits!!!!!
01:02:30 <ihope> http://3.141592653589793238462643383279502884197169399375105820974944592.com/
01:04:03 <pgimeno> (just checked, sorry, it was actually 51 billion digits... not sure if the decimal expression of 1/3 beats that, I would have to check)
01:05:16 * pgimeno counts digits
01:11:44 <pgimeno> <lament> "Prove the Jordan curve theorem: [_________________]"
01:11:56 <pgimeno> nah, systems like Coq or Mizar could do that automatically
01:15:10 <pgimeno> how about "Enter a BF program that outputs the number <random number>: [_________]" and interpret user input?
01:16:36 <ihope> What if the user input "+[>+]"?
01:17:47 <pgimeno> impose a cycle limit, e.g. for number N never let it run for more than k*log(N) for suitable k
01:18:23 <lament> why logN?
01:18:23 <ihope> k = 1 >:-)
01:18:36 <lament> and not expN
01:18:41 <lament> or something equally bizarre
01:18:42 <ihope> lament: that's the number of digits, more or less
01:18:57 <ihope> Oh, well, um.
01:19:25 <ihope> Then giving that the number 9837491257 would probably still cause problems.
01:19:32 <lament> pgimeno: mizar never "proved" anything automatically
01:19:57 <lament> pgimeno: it just verified that proof made by people.
01:20:32 <pgimeno> I know, I was just kidding
01:21:54 <pgimeno> so does Coq IIRC
01:22:30 <ihope> "Enter a regex that matches a number of the letter 'a' a number of times that is coprime to 938475: [_________________]"
01:23:30 <lament> /aa/
01:23:35 <pgimeno> lament: anyway it's already proved there: http://www.mizar.org/ (announced in the index page itself)
01:23:41 <lament> yeah, i know
01:24:49 <ihope> lament: it would have to match any such string
01:25:04 <lament> oh.
01:27:01 <pgimeno> /a/
01:28:10 <pgimeno> /a/ matches "aa", "aaa", "aaaaaaa", etc. as required
01:29:04 <pgimeno> oops, "aaa" should not be there
01:29:58 <pgimeno> maybe you want that any other string is NOT matched?
01:31:04 <ihope> Right.
01:31:17 <pgimeno> ah, that's harder to accomplish then
01:31:19 <ihope> And I want it to consume the entire string, as well.
01:31:46 <ihope> Anyway, it still should be possible.
01:32:18 <ihope> If the number were 6, it would be /(a|aaaaa)(aaaaaa)*/.
01:33:38 <pgimeno> provided you add ^ and $, I guess
01:33:59 <ihope> Well, if we assume the regex must match the whole string, then...
01:34:32 <pgimeno> ah ok
01:34:35 <ihope> Harder: "Enter a regex that matches a number of the letter 'a' a prime number of times, assuming that 938475 is prime but no lower number is: [_________________]"
01:34:55 <pgimeno> I was just being picky because PHP is picky with me :)
01:36:07 <pgimeno> gtg now, it's 2:35 am here... good night
01:36:37 <ihope> 'Night
01:39:57 <calamari> found ab ug but realized a way to get it down to 40
01:42:10 <lament> Enter a prime number: [___]
01:42:18 <graue> but now its size is no longer equal to life, the universe, and everything :(
01:42:43 <graue> add some features to bring it back up 2 bytes, calamari
01:42:47 <calamari> lol
01:43:00 <graue> like maybe add a debugger -- you can do that in 2 bytes, right?
01:44:28 <calamari> oops, 38
01:44:40 <calamari> guess I have plenty of room for that debugger now ;P
01:46:22 <calamari> I could strip 2 more bytes, but then the source files couldn't be ascii
01:47:44 <ihope> :-)
01:51:27 <lament> We should hold a contest for the coolest program of length 42
01:52:43 <lament> calamari would probably win, though
01:53:11 <calamari> nah, 4 nops would detract from coolness
01:53:55 <ihope> "[-]+[]!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!"
01:55:02 <lament> ihope: but that's not very cool.
01:55:17 <ihope> True.
01:56:19 <lament> "QQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQ"
01:56:32 <lament> prints a pretty-looking square in HQ9+
01:57:07 <lament> (actually, it probably doesn't.)
01:57:29 <lament> change that last Q to a newline.
02:31:33 <calamari> 36 bytes
02:32:32 <calamari> reading the program is just under half the code now :)
02:38:37 <calamari> yay, the equivalent of "[-]" seems to be working
02:38:44 <lament> haha
02:39:04 <calamari> well, it's a major pain to write better programs.. working on it :)
02:41:38 <calamari> next test is: [-]+. (exit)
02:42:56 <graue> what good is an OISC interpreter?
02:43:33 <calamari> what is your complaint
02:44:51 <graue> well, it doesn't do input or output
02:45:01 <calamari> mine does
02:45:20 <calamari> (well, still writing the program to verify that :)
02:45:22 <graue> how? memory-mapped?
02:45:33 <calamari> no, it uses ms-dos function calls
02:45:46 <graue> what? how do you do MS-DOS function calls in OISC?
02:46:07 <calamari> it isn't oisc, its based on oisc
02:46:13 <graue> oh
02:46:27 <graue> well, it's not a very portable language if you need MS-DOS function calls to use it
02:46:30 <calamari> basically, it was my inspiration, then I optimized for size
02:46:43 <calamari> graue: this isn't about portability, it's about raw size
02:47:15 <graue> I know the implementation isn't portable, but it would be nice if the implemented language was
02:47:29 <calamari> well, the turing complete part of the language is portable
02:47:45 <calamari> and the i/o part can also be coded
02:47:54 <calamari> but it won't be 36 bytes
02:48:21 <calamari> esp since Linux ELF executables can't be less than 45 bytes
02:51:02 <calamari> the language works like this: aaaa vv jjjj
02:51:34 <calamari> first it adds your offset aaaa to the memory pointer (adding a neg is like subtracting)
02:51:48 <calamari> then, it checks vv.. if 0, then it performs a system call
02:52:03 -!- ihope has quit (Read error: 110 (Connection timed out)).
02:52:09 <calamari> otherwise, it adds vv to the byte at the memory pointer (again with the negs)
02:52:50 <calamari> if the byte at the memory pointer is 0, then it adds jjjj to the ip, otherwise it continues to the next instruction
02:57:08 <calamari> typo.. pointer ISN'T 0,
03:19:48 <graue> cool
03:29:18 * SimonRC lols at the above captchas
03:33:21 <SimonRC> calamari: you could get a lot more speed by doing a less-than, rather than an equality test
03:33:25 <SimonRC> or maybe LE
03:33:42 <SimonRC> 'cause a<=b and b<=a means a=b
03:33:48 * SimonRC goes to bed
03:39:52 <calamari> got it :) nponnnnmimnnonnnlonnnpnnnnnnnn
04:07:00 -!- graue has quit ("Leaving").
04:26:02 -!- GregorR_ has joined.
04:26:30 -!- GregorR has quit (Remote closed the connection).
04:27:58 -!- GregorR_ has changed nick to GregorR.
04:51:36 <calamari> http://kidsquid.com/files/bf/miniscul.asm
04:51:39 <calamari> http://kidsquid.com/files/bf/miniscul.com
05:39:42 <calamari> http://esolangs.org/wiki/Miniscule
05:44:36 <calamari> changed caps to http://esolangs.org/wiki/MinISCule
06:53:06 -!- calamari has quit ("Leaving").
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:50:18 -!- nooga has joined.
08:50:28 <nooga> hi
11:06:57 -!- nooga has quit ("Lost terminal").
12:08:39 -!- pgimeno has quit (Read error: 104 (Connection reset by peer)).
12:09:37 -!- pgimeno has joined.
12:42:30 -!- tgwizard has joined.
12:43:02 -!- kate` has left (?).
14:01:47 -!- kipple has joined.
14:29:28 -!- ihope__ has joined.
14:29:30 -!- ihope__ has changed nick to ihope.
14:30:03 <ihope> ChatZilla's nice, but it does too many alert thingies.
15:01:39 -!- ihope has quit (Read error: 110 (Connection timed out)).
15:29:03 -!- ihope__ has joined.
15:29:05 -!- ihope__ has changed nick to ihope.
16:09:59 <SimonRC> RMS has finally lost it: http://software.newsforge.com/software/06/04/28/1648203.shtml?tid=150
16:16:30 <SimonRC> Slashdotter #1: "Autographs are only the beginning. I hear a Richard Stallman nude calendar is in the pipeline!" Slashdotter #2: "He threatens to display it at the next LinuxWorld conference unless a donation of $100,000 to the FSF is made."
16:20:06 <ihope> :-)
16:55:21 <ihope> I forgot the question, but the answer is NIN No MD No 18WPI.
18:31:10 <SimonRC> ihope: ?
18:31:34 <ihope> The answer is NIN No MD No 18WPI.
20:20:03 <ihope> It's asy to undrstand popl vn without th bnfit of 's.
21:38:04 -!- int-e has joined.
21:39:07 -!- sekhmet has quit ("a reboot! how novel.").
21:40:09 <SimonRC> ihope: whate does int-e mean?
21:41:59 -!- sekhmet has joined.
21:42:26 <ihope> Integrals eating?
21:44:54 <GregorR> Ensigned Integral ^^
21:45:05 <GregorR> It's an integral in the army ... with no sign bit.
21:46:49 <ihope> :-)
21:47:36 <int-e> SimonRC: a play on my initials and intel opcodes. the nick is old.
21:49:28 <SimonRC> int-e: do you know any ASM lang other than intel?
21:50:35 <int-e> not really. I've seen MIPS (and coded a bit on the strange simulated machine that SPIM provides), and I've done a bit of Z80 assembly ages ago.
21:51:27 <ihope> MIPS and SPIM, eh?
21:52:02 <SimonRC> I was hoping you would be really experienced and be able to tell me that one was clearly the best asm lang ever.
21:52:56 <ihope> Well, so what's the opcode?
21:53:14 <SimonRC> the int opcode, of course
21:53:23 <SimonRC> raise INTerupt
21:53:24 <ihope> Aha.
21:53:47 <SimonRC> used to ask the OS to do something, usually
21:54:15 <SimonRC> What is wrong with a plain function call I do not know.
21:54:21 <int-e> not so much anymore with the advent of the sysenter instruction (at least on newer x86)
21:54:35 <int-e> and never on mips, they have a trap instruction I think
21:54:51 * SimonRC does know x86 well enought to understand that
21:54:55 <ihope> So when do we get functional machine code?
21:55:00 * SimonRC does *not* know x86 well enought to understand that
21:55:01 <int-e> SimonRC: simple. a plain function call can not escape the memory protection (provided by the mmu)
21:55:12 <SimonRC> ok
21:55:25 <SimonRC> ihope: I believe there is one already
21:55:31 <ihope> What is it?
21:55:38 <int-e> in earlier days, code size also played a role - an int instruction is 2 bytes, a far call is 5 ...
21:55:50 <int-e> (talking 16 bit x86 code now)
21:57:34 <SimonRC> heh
21:57:45 <SimonRC> ihope: sodomy non sapiens
21:57:55 <SimonRC> (Buggered if I know.)
22:03:02 -!- sekhmet has quit ("next stop: 800MHz Duron (zomg)").
22:28:38 * int-e chuckles
22:28:46 <int-e> try /join #realworld ;)
22:37:31 <ihope> :-)
23:03:30 -!- ihope has quit (Connection timed out).
23:18:02 -!- calamari has joined.
23:26:45 <calamari> hi
23:27:45 <kipple> hi
23:37:02 <SimonRC> hi
23:42:39 -!- tgwizard has quit ("Leaving").
23:46:07 -!- sekhmet has joined.
23:50:33 <SimonRC> hi
←2006-04-29 2006-04-30 2006-05-01→ ↑2006 ↑all