00:00:21 <hppavilion[1]> If a magic space beam were to hit earth and flip everybody's gender
00:00:57 <hppavilion[1]> Would trans people be happy about it, or would they stay trans but have their identity flipped too?
00:01:08 <hppavilion[1]> (so a trans woman now has the body of a woman but identifies as a man)
00:02:01 <hppavilion[1]> Really, it's a question of whether the aliens that did this for whatever reason thought to flip identity too (so whether men still think of themselves as men, but are physical women)
00:03:02 <hppavilion[1]> Or if they made it set your identity to match your appearance (so men (who have become women) identify as women, women identify as men, trans women (who become women) identify as women (score), and trans men become men (and identify as such))
00:04:32 <prooftechnique> I assume people would have other things to worry about, like who's bombarding the planet with philosophical quandaries
00:09:02 -!- ais523 has joined.
00:09:41 <ais523> I run my own mailserver, and recently gmail started bouncing messages from it for a vaguely explained reason
00:10:04 <pikhq> hppavilion[1]: This is a question based on questions of both what the beam is *doing*, as well as as-yet open questions of what the physical and psychological bases of "being trans" *are*...
00:10:15 <ais523> so I started adding email auth features
00:10:40 <ais523> currently I'm messing around with dmarc, which asks mailservers who receive forged messages claiming to be from you to send you statistics about them
00:10:50 <hppavilion[1]> pikhq: Let's assume that identity is completely disjointed from (but still correlated with) physical gender
00:10:56 <ais523> (as a method of determining what impact adding more anti-forgery protections would have)
00:11:33 <hppavilion[1]> ais523: Is your personal mailserver running a software I've heard of?
00:11:44 <ais523> I just got my first report today; I reported a bug in Ubuntu, and it sent email about it with my email address forged as the from address
00:11:51 <ais523> hppavilion[1]: it runs a ton of software
00:11:59 <ais523> some of it you've probably heard of, like postfix
00:12:50 <ais523> @tell oerjan we could simply make 9 permit anon editing and add a separate anti-anon filter; better would be to disable anon editing in the settings
00:12:54 <pikhq> hppavilion[1]: Shit gets complicated, because physical sex is not just one phenotype that you can flip, and for some trans people that may well be super obvious and relevant.
00:12:57 <ais523> (or find some way to allow anons to edit in less spamy times)
00:13:12 <ais523> hppavilion[1]: a bunch of different programs are involved in saving and receiving emails
00:13:24 <zzo38> I have heard of Postfix; I use Exim since it has the features I require.
00:13:43 <hppavilion[1]> (Hm, hermaphrodit...edness is a matter of your body developing incorrectly (or awesomely, if that's what the reader is into), not anything genetic, right?)
00:14:01 <ais523> zzo38: which featuers do you use that postfix doesn't have?
00:14:18 <zzo38> I use Heirloom-Mailx as the front-end; it has the built-in ability to forge a from address easily (another feature I need, due to the way I have aliases set up)
00:14:47 <ais523> zzo38: most mail clients can forge a from address easily
00:14:56 <pikhq> There are some cases of being intersex that can be directly triggered genetically. Hell, there's even weird cases where you can end up with an unambiguously "normal" male or female phenotype without a normally matching genotype.
00:14:59 <ais523> not sure how many mail submission programs will forge a bounce (MailFrom) address
00:15:29 <zzo38> ais523: The way I have aliases set up is one of them. I don't know if Postfix can do that, but when looking at the possibilities I have found that Exim does support it so I used that.
00:16:19 <zzo38> (I have set it up so that it only receives messages send to aliases; not messages sent to UNIX usernames.)
00:16:51 <pikhq> You can have XX cis men and XY cis women. Both of these are *really* rare, but well-documented phenomena.
00:17:05 <ais523> pikhq: chimerism, perhaps?
00:17:08 <zzo38> pikhq: O, I did not know such thing would be possible
00:17:52 <zzo38> (You actually can send to a UNIX username if you connect directly to my SMTP server and do not use the local domain name instead of the internet domain name, but you shouldn't need to do that.)
00:18:10 <pikhq> ais523: No, in the XX case the father's gametes had a recombination with the X and Y chromosomes and the SRY gene landed on the X; for XY, SRY gets mutated or deleted such that it doesn't work.
00:18:50 <pikhq> And SRY is the major gene that triggers the masculinization of a fetus.
00:19:11 <mad> any cases of intersex that are fertile?
00:19:13 <zzo38> Can such people have children? If so, how are their children affected?
00:19:15 <ais523> so I guess the next mystery is, why canonical.com does not have an SPF record set up
00:19:33 <pikhq> (though, with the XY case *some* of the other genes doing that may still be around, so you get an incompletely masculinized human instead)
00:19:38 <zzo38> What does a SPF record mean?
00:20:39 <ais523> zzo38: it's something that you put in your DNS for a domain, that sets out the conditions on which an email connection with a HELO or MAIL FROM on that domain is legitimate
00:20:56 <pikhq> mad: To my knowledge, not all mechanisms result in infertility, but many of them do.
00:21:00 <ais523> (i.e. actually comes from that domain)
00:21:39 <ais523> one common one is to state "an email will only legitimately have this domain in its MAIL FROM or HELO if the connection is being made from the IP in the domain's A record"
00:21:47 <pikhq> The major relevant question here is if the ovaries or testicles are able to correctly form, and if so, if they're able to produce functioning gametes.
00:22:21 <ais523> the idea is so that you can authenticate the server that relays the message as actually being the server that relays the message, rather than a fake
00:22:41 <zzo38> In my case I am actually sending from the domain that the MAIL FROM is, but then the ISP will forward the message to the correct SMTP server, so it will also be legitimate if the ISP is forwarding it for me.
00:22:59 <mad> pikhq : In what kind of case are the gametes functional?
00:23:00 <pikhq> Basically, biology is stupidly complicated.
00:23:03 <ais523> it's very easy to set up and makes anti-spamming technologies work much more effectively as it allows them to gain trust in a relay
00:23:13 <ais523> zzo38: well SPF is entirely about HELO and MAIL FROM, it says nothing about the From: line
00:23:23 <ais523> and an ISP relay will change the MAIL FROM I think?
00:23:54 <zzo38> I don't know if the ISP will change the MAIL FROM to its own address.
00:25:37 <ais523> zzo38: hmm, my mail logs show the HELO but not the MAIL FROM
00:25:43 <ais523> so I could see if it changes the HELO, at least
00:25:48 <ais523> let me try to see if I can turn on logging for MAIL FROM too
00:26:53 <zzo38> Sending the message first to ISP to have them forward it to the correct server is known as "smarthost" in the Exim configuration. (If I send a message directly to an external SMTP server, I normally get an error message that says that it is "graylisted", so I don't do that.)
00:27:05 <oerjan> ais523: that will work for anonymous ip, although i thought it might also be good to a separate error message if they do post logged in to the right page, but get the format wrong
00:27:22 <ais523> oerjan: then you can use two separate filters for the purpose
00:27:23 <oerjan> i suppose it remains to be seen if anyone actually has trouble with that.
00:28:22 <pikhq> mad: It Depends. One example is persistant Müllerian duct syndrome, where normal male sexual development happens, *but* the sexual organs which develop from the Müllerian duct (fallopian tubes, uterus, internal portions of the vagina) also end up developing.
00:29:21 <ais523> zzo38: so there are two ports on which emails are sent, 25 and 587
00:29:31 <pikhq> This can cause infertility if that ends up effecting development, but it doesn't necessarily.
00:29:48 <ais523> a message sent on 587 is assumed to be from an email sending user to a server, that typically requires authentication
00:29:59 <ais523> 25 is used for relays between servers
00:30:16 <ais523> (i.e. you use 587 when sending an email to your own mailserver, then your mailserver uses 25 to send to the recipient's mailserver)
00:30:29 <ais523> now, spambots normally pretend to be a mailserver, so they make a bunch of port 25 connections
00:30:40 <zzo38> I am sending to port 25 on the ISP, although that server is not accessible from outside of the ISP's network.
00:30:51 <pikhq> ais523: I assume most things relaying on port 25 try STARTTLS?
00:30:54 <ais523> but one of the jobs of a mailserver is to retry if there's a temporary error
00:31:08 <ais523> pikhq: quite probably; 587 is also plaintext + STARTTLS
00:31:26 <myname> so... i need ideas for stuff to add to pixel dungeon
00:31:29 <ais523> anyway, if an email server receives a connection on port 25 from somewhere it's never heard of
00:31:40 <ais523> it normally simulates a few errors first to see if the other side reacts like a mailserver would
00:31:48 <zzo38> (Although there is also a port 587 server for use if someone wants to send from the ISP's mail but is connecting from a cellular network or something else)
00:31:51 <ais523> many spambots don't bother reacting to them properly because it would slow their spamming down
00:31:54 <ais523> that's called greylisting
00:32:14 <ais523> myname: you're in charge of pixel dungeon?
00:32:35 <myname> i just thought if everybody forks it, i might as well do it
00:33:03 <ais523> zzo38: older mail clients use 25 for submission rather than 587, so most ISPs will treat 25 like 587 if the message comes with appropriate auth (in this case, from being sent from inside their subnet)
00:33:14 <myname> i buffed the blacksmith so far
00:33:32 <hppavilion[1]> (I wonder if there's a way to disable capslock entirely...)
00:34:09 <ais523> hppavilion[1]: I rebound my capslock to Compose
00:34:20 <ais523> a while back I rebound shift-shift to capslock, so I'd still have an easy way to type in allcaps
00:34:21 <zzo38> hppavilion[1]: If you can change that keybinding then yes
00:34:24 <ais523> but it doesn't come up that often
00:34:31 <myname> you can now foege weapon/armor + potion to give it an enchantment or inscription you desire
00:34:32 <ais523> and the shift-shift binding stopped working at some point
00:34:53 <hppavilion[1]> zzo38: I can change keybindings, but I think the official Microsoft keybinding maker doesn't let you rebind capslock
00:35:08 <hppavilion[1]> (you can only rebind character keys- letters, punctuation, symbols, numbers)
00:35:29 <hppavilion[1]> Enter/shift/ctrl/alt/caps/tab are all unrebindable
00:35:38 <myname> you also can dio weapons into potions now
00:35:40 <zzo38> hppavilion[1]: Yes, I have seen the keyboard layout making program for Windows and it does not allow rebinding capslock; many other keys also can't be rebound, including all number pad keys except for the . key
00:36:12 <hppavilion[1]> zzo38: I don't know if it's because Windows doesn't let you change those or because Microsoft thinks you shouldn't
00:36:32 <zzo38> But now I use Linux which you can rebind everything (at least in X, I think)
00:37:23 <hppavilion[1]> (I hate the manufacturer mentality of "a normal user probably won't want to do this thing that is potentially useful, so we can just make it impossible to ever do it no matter how smart you are without switching software or rewriting part of the OS")
00:38:34 <zzo38> I also hate that stuff
00:39:08 <zzo38> That is one thing why FOSS is better, and why UNIX is better.
00:39:16 <mad> is there a limit on what you can do with combinating accents on win32?
00:41:45 <mad> I think the win32 keyboard stuff is for doing all the italian/croatian/thai/etc keyboards, not doing fancy stuff like mapping ctrl to tab and tab to caps
00:43:22 <prooftechnique> I've had caps bound to Esc/Ctrl for as long as I can remember
00:44:28 <prooftechnique> Then again, I also don't mess about with keymaps, I just reprogram my keyboard.
00:45:05 <mad> yeah ok if you want to emulate the space cadet lisp machine keyboard in windows, you'll be disappointed yet
00:46:00 <ais523> prooftechnique: how often do you change between Esc and Ctrl?
00:46:21 * ais523 has a vision of a key in caps lock position that does ctrl if you hold it while pressing another key, and esc if you tap it
00:46:40 <mad> how often is Esc used these days?
00:47:02 <mad> It used to be used a ton but I think it's mostly out of use now
00:47:39 -!- augur has joined.
00:47:55 <prooftechnique> ais523: I guess technically it's not really a Caps Lock, it's just a key that I programmed to do that which is coincidentally in the same spot.
00:48:18 <prooftechnique> But, I have used software on occasion to get the same effect on built-in boards
00:48:25 <ais523> mad: the joke is that you bind caps lock to ctrl if you use emacs and esc if you use vim
00:48:40 <ais523> prooftechnique: almost all keyboards have "caps lock" printed on that key though
00:48:58 <ais523> because although many #esoteric denizens are the sort to rebind keyboards, fewer of them change the physical keycaps
00:48:59 <zzo38> I normally just use Ctrl+[ for escape in vim though
00:49:14 <ais523> my computer from my last job had an azerty keyboard
00:49:23 <ais523> sometimes it was set to English layout, sometimes to French layout
00:49:29 <prooftechnique> True, but how often do you pay attention to Caps Lock? :D
00:49:39 <ais523> I think mostly I just typed in English but with French keycaps
00:49:39 <mad> azerty is crazy
00:49:52 <prooftechnique> I just think of it as a more convenient Ctrl, and now I get weirded out when I use other people's computers
00:50:02 <mad> shift to type numbers? wut?
00:50:58 <ais523> I seem to remember using a computer that had both caps lock and shift lock
00:51:00 <pikhq> For me, an AZERTY keyboard would drive me crazy, even set to an English layout.
00:51:06 <ais523> I'm not sure why you'd use a shift lock
00:51:32 <ais523> one annoying thing is that with caps lock on, holding shift didn't let you type in lowercase
00:51:44 <ais523> which was annoying because I was using a programming language which required all keywords to be uppercase
00:51:49 <hppavilion[1]> Instead of dressing baby boys in one color and girls in another
00:51:56 <ais523> (this mostly just lead to everything being uppercase as mixing case was too tedious)
00:52:01 <pikhq> But I'm used to not just a QWERTY keyboard, I'm used to ANSI mechanical layout QWERTY keyboards.
00:52:05 <prooftechnique> mad: If you think shift to type numbers is weird, that's my layout above
00:52:24 <pikhq> While that AZERTY keyboard, just like a standard UK QWERTY keyboard, would be an ISO mechanical layout keyboard.
00:52:47 <ais523> prooftechnique: red on grey makes it quite hard to read, and reverse video doesn't help
00:52:57 <prooftechnique> hppavilion[1]: Or we could stop ascribing meaning to random colors
00:53:23 <mad> pikhq : is that the one with the straight enter key rather than the flipped L shape?
00:53:35 <ais523> also how do you get at the alternative combinations? holding modifier keys, or are there seven keyboards next to each other?
00:53:35 <prooftechnique> ais523: I might have a link to the layout editor. Hang on
00:53:40 <pikhq> mad: Yes, that's the ANSI layout.
00:54:02 <prooftechnique> ais523: The red keys are held down to get to specific layers, and some of them are modal
00:54:33 <prooftechnique> The layout on the top right is a "dead" layer. I hit what's basically comma to get to numbers and special symbols
00:54:46 <prooftechnique> And the qwerty and steno layouts have their own switches
00:55:13 <prooftechnique> ais523: http://www.keyboard-layout-editor.com/#/gists/2b875f7d5d76fe4408c0a5b3bd76ddae
00:56:31 -!- jaboja has quit (Remote host closed the connection).
00:57:08 <prooftechnique> http://www.keyboard-layout-editor.com/##@_backcolor=%23292727&name=Workman%20Dead%20Planck%20with%20KWM%20and%20Plover&author=Jack%20Henahan&switchMount=cherry&pcb:true%3B&@_c=%23ffffff&a:7%3B&=Tab&=Q&=D&=R&=W&=B&=J&=F&=U&=P&=%2F%3B&=Back%20space&_x:0.5%3B&=&=%25&=%2F&&=%3F&=+&=%2F@&=$&=%2F_&=%5B&=%5D&=!&=%3B&@=Esc&=A&=S&=H&=T&=G&=Y&=N&=E&=O&=I&='&_x:0.5%3B&=&=%2F%2F&=(&=%2F=&=0&=%7B&=%7D&=1&=*&=)
00:57:14 <prooftechnique> &=-&='%3B&@=Shift&=Z&=X&=M&=C&=V&=K&=L&=Dead&=.&=%2F%2F&=Enter&_x:0.5%3B&=&=6&=7&=8&=9&=%7C&=%5C&=2&=3&=4&=5&=%3B&@=Ctrl&=RAlt&=LAlt&=⌘&=Fn&=&=&=KWM&=%2F&larr%2F%3B&=%2F&darr%2F%3B&=%2F&uarr%2F%3B&=%2F&rarr%2F%3B&_x:0.5%3B&=&=&=%3C&=~&=%60&=,&=,&=%23&=%5E&=%3E&=&=%3B&@_y:0.5%3B&=~&=!&=%2F@&=%23&=$&=%25&=%5E&=%2F&&=*&=(&=)&=Back%20space&_x:0.5%3B&=Tab&=Q&=W&=E&=R&=T&=Y&=U&=I&=O&=P&=Back%20space%3
00:57:20 <prooftechnique> B&@=Del&=F1&=F2&=F3&=F4&=F5&=F6&=%2F_&=+&=%7B&=%7D&=%7C&_x:0.5%3B&=Esc&=A&=S&=D&=F&=G&=H&=J&=K&=L&=%2F%3B&='%3B&@=Shift&=F7&=F8&=F9&=F10&=F11&=F12&=%7C&=±&=.&=Vol+&=Enter&_x:0.5%3B&=Shift&=Z&=X&=C&=V&=B&=N&=M&=,&=.&=%2F%2F&=Enter%3B&@=Ctrl&=RAlt&=LAlt&=⌘&_c=%23b05858%3B&=Fn&_c=%23ffffff%3B&=&=&=Adjust&=⏮&=⏭&=Vol-&=⏯&_x:0.5%3B&=Ctrl&=RAlt&=LAlt&=⌘&=Fn&=&=&=KWM&=%2F&larr%2F%3B&=%2F&darr%2
00:57:26 <prooftechnique> F%3B&=%2F&uarr%2F%3B&=%2F&rarr%2F%3B%3B&@_y:0.5&f:2%3B&=Reset%20KWM&=&=Toggle%20Split&=Float&=Fill%20Cont-ainer&=Full%20Screen&=Insert%20Marked&=Insert%20West&=Insert%20South&=Insert%20North&=Insert%20East&=&_x:0.5&f:3%3B&=%23&=%23&=%23&=%23&=%23&=%23&=%23&=%23&=%23&=%23&=%23&=%23%3B&@_f:2%3B&=&=Mark&=Rotate&=Prefix&=Term&=BSP&=Swap%20Marked&=Swap%20West&=Swap%20South&=Swap%20North&=Swap%20East&=&
00:57:32 <prooftechnique> _x:0.5&f:3%3B&=&=S&=T&=P&=H&=*&=*&=F&=P&=L&=T&=D%3B&@_f:2%3B&=Shift&=Mark%20West&=Mark%20South&=Mark%20North&=Mark%20East&=Mon-ocle&=Float&=Focus%20West&=Focus%20South&=Focus%20North&=Focus%20East&=Enter&_x:0.5&f:3%3B&=&=S&=K&=W&=R&=*&=*&=R&=B&=G&=S&=Z%3B&@_f:2%3B&=Ctrl&=RAlt&=LAlt&=⌘&=Adjust&=&=&_c=%23b05858%3B&=KWM&_c=%23ffffff%3B&=%2F&larr%2F%3B&=%2F&darr%2F%3B&=%2F&uarr%2F%3B&=%2F&rarr%2F%3B
00:57:38 <prooftechnique> &_x:0.5&f:3%3B&=Exit%20Plover&=&=&=A&=O&=&=&=E&=U&=&=&=%3B&@_y:0.5&x:6.25&f:2%3B&=&=DFU%20Mode&=&=&=&=&=&=&=&=&=&=Del%3B&@_x:6.25%3B&=&=&=&=Audio%20On&=Audio%20Off&=OS%20X%20Keys&=Win%20Keys&=QWER-TY&=Work-man%20Dead&=Plover&=&=%3B&@_x:6.25%3B&=&=Voice-&=Voice+&=Music%20On&=Music%20Off&=MIDI%20On&=MIDI%20Off&=&=&=&=&=%3B&@_x:6.25%3B&=&=&=&=&_c=%23b05858%3B&=Adjust&_c=%23ffffff%3B&=&=&_c=%23b05858%
00:57:52 <prooftechnique> https://gist.github.com/jhenahan/ef372d1b77b2214ba645a7d324609bdf
00:57:53 <ais523> oh, you've stopped fakelagging
00:58:00 <ais523> I was just starting to type the commands to kick you
00:58:06 -!- moonythedwarf_ has quit (Ping timeout: 265 seconds).
00:58:16 <ais523> (not as a punishment, it's just the typical way to handle people with out-of-control clients)
00:58:38 <prooftechnique> Had no idea it was going to give me that mess instead of a nice gist URL :/
00:59:19 <ais523> I'm actually fairly impressed that it managed to wordwrap it
00:59:50 <ais523> most clients would have given up trying to fit an URL that long onto IRC long ago
01:00:17 <ais523> (I would have typed the previous sentence sooner if I hadn't spent so long trying to figure out which indefinite article "URL" uses)
01:00:20 <mad> auto split has.... some advantages
01:01:06 <prooftechnique> I should probably see about altering the splitting strategy to just give up on URLs in general
01:18:16 -!- MDude has quit (Remote host closed the connection).
01:24:09 -!- hppavilion[1] has quit (Ping timeout: 276 seconds).
01:26:05 <prooftechnique> I like that Andrew Topping was so reclusive as to not even have a Wikipedia page
02:26:20 -!- oerjan has quit (Quit: Nite).
02:49:11 -!- ais523 has quit (Remote host closed the connection).
02:49:36 -!- MDude has joined.
02:49:56 -!- ais523 has joined.
03:00:33 -!- AlexR42 has joined.
03:08:04 -!- yorick_ has joined.
03:08:04 -!- yorick_ has quit (Changing host).
03:08:04 -!- yorick_ has joined.
03:08:35 -!- Phantom_Hoover has quit (Read error: Connection reset by peer).
03:09:52 <zzo38> I have seen both kind of indefinite article "URL" uses and I think it does not matter much
03:10:50 -!- yorick has quit (Ping timeout: 250 seconds).
03:25:30 -!- AlexR42 has quit (Quit: My Mac has gone to sleep. ZZZzzz…).
03:41:02 <zzo38> How well do you think it might work in a computer program if the text of Magic: the Gathering cards is written using RDF and then is compiled into the code needed to execute? Examples might be [:counter [:target :spell]] and [:destroy [:target [:and [:not :black], :creature]]]
04:24:04 -!- Zarutian has quit (Quit: Zarutian).
05:45:21 -!- carado has joined.
05:52:09 -!- Jafet has quit (Ping timeout: 258 seconds).
05:58:25 -!- Jafet has joined.
06:09:55 -!- Jafet has quit (Quit: Jafet).
06:30:24 -!- carado has quit (Read error: Connection reset by peer).
06:32:00 -!- carado has joined.
06:40:09 -!- hppavilion[1] has joined.
06:51:27 -!- carado has quit (Read error: Connection reset by peer).
06:51:50 -!- carado has joined.
06:56:22 <izabera> http://www.cs.princeton.edu/~rs/talks/QuicksortIsOptimal.pdf in the examples, are the arrays 1-indexed?
07:08:22 * pikhq is unimpressed by O(n^2) worst case behavior
07:09:15 <izabera> in his example in page 9, k is not declared
07:10:14 <pikhq> Though, it is possible to devise a variant of quicksort with O(n log n) behavior...
07:16:48 -!- Jafet has joined.
07:26:31 * FreeFull is unimpressed by pdf pages being sideways
07:27:07 <FreeFull> pikhq: You can make O(nlogn) guaranteed quicksort with the right pivot choosing algorithm, but it has worse constant factors
07:27:35 <shachaf> is that even quicksort at that point? sounds more like a cheap ripoff
07:27:42 <shachaf> those people don't have all their ducks in a row
07:28:00 <pikhq> Sure, but you can also know if you're in a pathological case and switch pivot choosing algorithms appropriately. :)
07:28:32 <shachaf> actually just aiming for the pun there
07:29:03 <shachaf> you can tell i'm not serious because the sentence isn't capitalized
07:29:19 <pikhq> (I don't know if this actually performs well, I've just thought of it, not tested it)
07:30:13 <FreeFull> https://en.wikipedia.org/wiki/Median_of_medians
07:30:39 <FreeFull> You can also choose random pivots, which will hit O(n^2) behaviour if you're really, really unlucky
07:31:08 <pikhq> The thing I'm not sure of is if quicksort-and-use-median-of-medians-when-you're-in-worst-case-behavior-land is actually worth it.
07:31:16 <FreeFull> Really really unlucky would involve an adversary that knows the state of your random number generator, essentially
07:31:32 <FreeFull> pikhq: Random pivots are faster
07:31:58 <pikhq> Tempting, use a CSPRNG for qsort. :)
07:32:25 <pikhq> (*cough*arc4random or get out)
07:34:31 <pikhq> The BSD API for a CSPRNG that's nice, quick, and easy, and glibc really should have it.
07:35:22 <shachaf> There's the argument that userspace CSPRNGs are rarely justified compared to just using urandom.
07:38:01 <pikhq> There's the rather bigger argument that having CSPRNGs be hard to use means nobody uses 'em, and instead implements their own, shit implementation that often doesn't even work correctly.
07:42:42 -!- mad has quit (Quit: Rouringu de hajikunda!).
08:20:02 -!- AnotherTest has joined.
08:32:20 -!- Caesura has joined.
08:32:44 -!- augur has quit (Remote host closed the connection).
08:34:33 -!- Kaynato has quit (Ping timeout: 240 seconds).
08:42:15 -!- hppavilion[1] has quit (Ping timeout: 276 seconds).
08:46:41 -!- carado has quit (Remote host closed the connection).
08:48:16 -!- carado has joined.
08:53:33 -!- Caesura has quit (Ping timeout: 240 seconds).
09:02:02 <b_jonas> zzo38: the new card is Spy Kit
09:02:24 <b_jonas> zzo38: and it's not really as powerful as the un-card, because the un-card works anywhere, this only on the battlefield
09:03:32 -!- ais523 has quit (Ping timeout: 244 seconds).
09:05:14 <Taneb> b_jonas, I've got a friend planning a spy kit deck
09:06:03 <izabera> pi seconds ~ 1 nanocentury
09:07:40 <zzo38> b_jonas: Yes I have seen that, and my question (still not answered) is, which layer does the name-changing effect apply?
09:09:14 <zzo38> (In my own collection of custom cards I have added the name-changing layer directly after the text-changing layer; I have done this far before Spy Kit card was revealed.)
09:11:44 <zzo38> The notes do not seem to answer that question, although they do answer the other questions I had (about monarch and about the new keyword actions and keyword abilities), as well as a lot of questions that I did not have because existing rules explain it clearly enough.
09:14:03 -!- augur has joined.
09:15:12 <izabera> why does recaptcha ask me to copy and paste a random string that's valid for 2 minutes?
09:15:18 <izabera> what security does it add?
09:15:28 <izabera> bots can't copy and paste? o.O
09:17:04 <zzo38> It adds security against XSS I think?
09:21:35 <b_jonas> zzo38: which layer => we'll probably find that out from the set FAQ only. there are usually some rules questions open until that
09:22:55 <b_jonas> zzo38: my guess is that it applies later than ability changes and earlier than p/t changes, but it rarely matters.
09:26:29 <b_jonas> zzo38: Arcane Savant also needs a new rule in the comp rules, near the Leyline rules, to clarify how it works and whether you get to reapply it for each mulligan. The set FAQ will clear that up too.
09:38:57 <b_jonas> zzo38: and I'm not even sure what the Spy Kit ability means in the sense of what names it chooses, eg. whether it takes only the names of cards not banned in the current format, but it probably also doesn't matter.
09:48:48 <zzo38> I would assume meaning cards legal in the current format (which seem the most logical way to me), but I am not sure???
09:51:50 -!- carado has quit (Read error: Connection reset by peer).
09:52:02 -!- carado has joined.
09:52:49 <b_jonas> zzo38: it seems it probably doesn't matter, because Spy Kit is legal only in eternal formats (vintage, legacy, commander), and it seems there are no silver-bordered or banned creature cards for those formats whose name is a type combination or the name of a token.
09:56:43 <zzo38> But then, there can also be custom formats (which might or might not include unofficial cards, as well as other arbitrary selections of official cards)? For these reason, and also I don't like that names of implicitly named tokens can compare equal to card names, so I have written some of my own rules.
09:58:51 <zzo38> My own (unofficial) rules specify that the name of an implicitly named token is a (possibly empty) (unordered) set of subtypes, rather than a "canonical name" which is the name of a card or explicitly named token.
09:59:01 <zzo38> To me, these rule seem more sense mathematically.
09:59:44 <zzo38> (So the possibility that a card name and subtype may be spelled the same way in English is irrelevant.)
09:59:45 <b_jonas> zzo38: about named tokens, if I have a Llanowar Mentor in play and artificially evolve it to replace Elf to Viashino in its text, then what token will it put into play? do you dislike that kind of named token too?
10:00:50 <zzo38> It is still named Llanowar Elves, since that is an explicit name and not a subtype.
10:01:04 <zzo38> However, the subtype of the token will now be Viashino Druid.
10:01:30 <b_jonas> zzo38: right, and the Mentor itself will also be a Viashino Spellshaper
10:03:24 <zzo38> (And the token will have the same name as the card "Llanowar Elves". I believe all of these things are the same both in official and unofficial rules though, anyways.)
10:04:37 <b_jonas> zzo38: I'm not sure how it works officially. artificial evolution still has to act on plurals like "Elves" so as to work on cards like Wirewood Pride correctly
10:05:11 <b_jonas> but it probably still doesn't act on the "named Llanowar Elves" part
10:05:55 <zzo38> I think that the word "Elves" in the text there is still not a subtype so it still doesn't act on it. But, if it says "Elves" where it is the plural of a subtype, then I would think it is still the same word; the way you write it in any language should be irrelevant.
10:07:42 <b_jonas> sure, but in the official rules how the card name is written in English still matters for cards like Giant Slug or Viashino Skeleton or Elder Druid, sadly
10:08:17 <zzo38> Yes, and that is too bad it is why I made better rules for unofficial. Hopefully they can fix the official rules too
10:08:28 <b_jonas> let me check what your exact fix is
10:09:59 <zzo38> It is exactly what I wrote above, although I give more examples in my file
10:12:09 <b_jonas> zzo38: if I understand right, then in your rules, (1) tokens that are given a name only implicitly from its subtype has a name that never compares equal to a name that's given explicitly or copied from a card name; and text-changing effects like Artificial Evolution don't change card names, whether they're on the card name line or mentioned in the rule text as a card name. is that right?
10:13:22 <zzo38> Yes, although with my rule also Artificial Evolution does change subtypes listed on the card name line (although only tokens can ever have subtypes on the card name line).
10:15:01 <zzo38> Also, if a name is the empty set of subtypes, that is not the same as having no name at all.
10:15:29 <b_jonas> can you ever have a name that's the empty set of subtypes?
10:15:33 <b_jonas> I don't think that's ever possible
10:16:13 <zzo38> I don't know if with official cards, but with unofficial cards it seem that it clearly should be.
10:19:55 <b_jonas> zzo38: and also, both in the official rules and your rules, if you change color words (eg. with Spectral Shift or Sleight of Mind), that doesn't change color words that happen to be in a card name (whether it's on a name line or rules text)
10:20:33 <zzo38> Yes, that part is the same, too.
10:21:44 <zzo38> I have updated my file to write the rule better, since it previously did not actually explain much.
10:22:52 <b_jonas> I wonder if there's a named counter that has a name that has a subtype or color word in it, but probably no.
10:24:00 <zzo38> I don't think so, but even if it is, it is not a subtype word or color word, I think.
10:26:44 <b_jonas> It's hard to get a name of all named counters, with "counter" meaning two things
10:27:40 <zzo38> I did find a list once somewhere, although I think they ought to add it to the Comprehensive Rules too for reference
10:31:46 <zzo38> Do you like my variant rule for implicit names of tokens? Do you know who else like it or hate it?
10:34:06 <b_jonas> oh, there's "mine counter" technically, and "Mine" is a land type, but that doesn't come up in rules problems, because the text-changing effects only work with basic-land-types, and the cards like Shimmer and Traveler's Cloak don't interact with the counter name in any way.
10:34:27 <b_jonas> Vision Charm lets you change text from land types
10:34:38 <zzo38> But if "mine counter" isn't capitalized anyways, then it doesn't suggest a subtype.
10:35:37 <b_jonas> so Vision Charm doesn't let me change Mine Layer to use swamp counters instead of mine counters
10:42:58 <zzo38> I did make the unofficial card to change counter names in text. You can't change "+1/+1 counter" into anything else, but you can change "experience counter" into "poison counter".
10:44:49 <b_jonas> ah, http://mtgsalvation.gamepedia.com/Counter_(marker)/List_of_Counters might be useful
10:47:13 <zzo38> Yes that is what I have seen before.
10:49:15 -!- Reece` has joined.
10:51:34 * b_jonas checks the lists -- "Orgg", what the heck is that? apparently some crazy old goblins. maybe it's what they call the Uruk-Hai.
10:53:12 <b_jonas> "Pest" -- apparently a card from Mirrodin makes Pest tokens
10:55:17 <int-e> huh. 0/1... I guess they can grow by other effects and become more than a nuisance
10:56:25 <b_jonas> int-e: there are also Prism creature tokens from an old card, Reflection creature tokens from two old cards
10:57:09 <b_jonas> and of course the famous Sand creature tokens
10:57:51 <b_jonas> actually two cards put Sand creature tokens into play
11:00:25 <b_jonas> Ok, now for the counter list. There are gold counters (and tokens *named* Gold, which is irrelevant), currency counters, and credit counters. (Also arrow counters and arrowhead counters.)
11:03:44 <b_jonas> There are devotion counters and echo counters which have no relation to those keywords. There are level counters and experience counters, and they're not the same.
11:04:06 <b_jonas> There are fungus counter and different Fungus creatures
11:04:38 <int-e> . o O ( exparation counters )
11:07:10 <b_jonas> Some of the Fungus creatures have Thallid or Pallid or Thelon in their name, they use spore counters to produce Saprolings. Whereas some Fungus creatures and an instant have Spor- in their name and put fungus counters on a permanent. Not confusing at all.
11:09:12 <b_jonas> (Oh well, it's not really much worse than some of the strangely named recently printed cards.)
11:09:36 <b_jonas> Some of these counters have strange names.
11:10:29 <b_jonas> Ah, there are fate counters and omen counters.
11:10:47 <b_jonas> petrification counters and paralyzation counters.
11:13:16 <b_jonas> there are infection counters, poison counters, plague counters, but also healing counters and vitality counters. tide counters (Homarid) and time counters.
11:14:19 <b_jonas> age counters for cumulative upkeep, time counters for suspend and vanishing, fade counters for fading.
11:16:04 <zzo38> That allow you to see some of the thing that can be interacted with by my "Aether Hack" card.
11:16:21 <zzo38> As well as what cards might interact even without that.
11:18:05 <b_jonas> zzo38: hmm, how does Aether Hack interact with leveling cards? I don't recall how those work.
11:35:29 -!- boily has joined.
11:55:26 -!- Phantom_Hoover has joined.
11:58:57 <zzo38> No effect, since the text of the card does not say "level counter".
11:59:32 <zzo38> (However, it can affect other cards to cause them to place level counters on the leveler card.)
12:05:10 <b_jonas> zzo38: Ok. So the text-changing effects don't see inside the definitions of the level up keyword (or the fear keyword)? So the keyword abilities "represent" abilities (whatever that means), don't put on text, even if the ability represented by level up is an activated ability that can be stolen by an Experiment Kraj, right?
12:07:36 <zzo38> Level up is an activated ability, so I would think so.
12:09:26 <b_jonas> What I'd like to say is that when the rules say that a keyword ability (possibly parametrized) represents a set of abilities, that probably means that the keyword ability adds those other abilities to the card, but that happens in the layer for abilities, not the layer for text change, so those other abilities aren't part of the card text and aren't affected by text-changing abilities, right?
12:11:47 <zzo38> Yes those other abilities aren't part of the card text I believe, but I am not sure about the layers
12:17:11 <b_jonas> Whereas there are still a LOT of things that act on text level or below: at the copy layer: (creating a token, creating copy of spell, permanent becoming a copy of or entering tb as a copy of something, Primal, planeswalker's extra ability, morph and other face-down, double-faced, meld, flip, split and fuse), at text layer: (change word in text abilities, splice onto, entwine, overload),
12:17:17 <b_jonas> and there's probably more I forgot about.
12:17:32 <b_jonas> It's complicated and I don't understand all the interactions.
12:19:29 -!- AlexR42 has joined.
12:21:03 <zzo38> I am guessing that if a keyword represents two abilities, then that text causes it to have those two abilities instead of just one ability.
12:22:14 <zzo38> Text is not a continuous effect so the layer system does not apply, I think.
12:23:18 <b_jonas> the comp rules also say that "Level up is an activated ability." so maybe the Kraj or Quicksilver Elemental actually steals the level up itself too?
12:23:44 <b_jonas> for something like level up it doesn't matter if something ends up having two copies
12:23:56 <zzo38> It takes the level up ability, but the level counters do nothing for Kraj
12:25:19 <b_jonas> zzo38: forget the kraj then, just have something temporarily cytoshaped to a Quicksilver Elemental, make it steal all activated abilities that way, then let the cytoshape time out, the stolen ability stays
12:25:31 <b_jonas> I just assume that any permanent can have any activated ability stolen on it
12:26:30 <b_jonas> and I wonder how that interacts with keywords that are activated abilities or represent an activated ability (and possibly other abilities)
12:26:54 <b_jonas> let me see if there's an ugly case
12:28:16 -!- boily has quit (Quit: ÆON CHICKEN).
12:30:13 <b_jonas> It can steal an "equip {1}" ability from a sword onto a knight that's not an equipment or aura or forti, and I think then you can attach the knight to an elephant you control but the attachment doesn't do anything until the knight becomes the copy of something later.
12:30:23 <b_jonas> that's not really what I'm after though
12:32:02 <zzo38> You can activate the ability but since it is not Equipment or Aura, it doesn't do anything, unless it somehow becomes such a thing before it resolves.
12:32:25 <b_jonas> zzo38: what, why? is there anything that says that only equipments or auras can be attached?
12:32:49 <b_jonas> I thought the rules only say that creature auras can't be attached, and the same about creature equipment and creature forti.
12:33:36 <zzo38> Yes, only Aura, Equipment, Fortification can be attached to anything, and only if it is not also a creature (although I don't like this latter part and made a rule variant that discards it).
12:34:27 <zzo38> Rule 701.3b says (in part): If an effect tries to attach an object that isn't an Aura, Equipment, or Fortification to another object or player, the effect does nothing and the first object doesn't move.
12:34:59 <b_jonas> ok, then you can only attach those
12:36:13 <b_jonas> can a permanent remain attached to something even when that permanent is no longer an aura or equipment or forti?
12:37:46 <zzo38> It becomes unattached as a state-based action (rule 704.5q).
12:38:25 <b_jonas> ok, that should cover all the holes
12:40:02 <b_jonas> And I don't think there's any keyword ability where it causes any problem with stealing all activated abilities, because there are very few keyword abilities that create activated abilities that work from the battlefield.
12:57:12 <zzo38> Possibly with some help, I may be able to form custom card sets, that may have some of the new rules, new keywords (abilities/actions), new subtypes, new cards (including ones I haven't yet listed in my file), and/or reprints.
13:08:31 -!- augur has quit (Remote host closed the connection).
13:08:37 -!- AlexR42 has quit (Quit: My Mac has gone to sleep. ZZZzzz…).
13:09:34 -!- AlexR42 has joined.
13:39:11 -!- AlexR42 has quit (Quit: My Mac has gone to sleep. ZZZzzz…).
14:02:16 -!- AlexR42 has joined.
14:22:00 -!- oerjan has joined.
14:22:12 -!- Sgeo_ has quit (Ping timeout: 240 seconds).
14:43:19 -!- `^_^v has joined.
14:46:05 -!- AlexR42 has quit (Quit: My Mac has gone to sleep. ZZZzzz…).
14:47:17 -!- Jafet has quit (Ping timeout: 244 seconds).
15:04:32 <b_jonas> I should learn how to use greasemonkey or something like that
15:05:04 <myname> easy, userscripts.org, look stuff up, change parts of code
15:06:50 -!- idris-bot has quit (Ping timeout: 258 seconds).
15:08:11 -!- AlexR42 has joined.
15:08:25 <oerjan> the new filter is eerily effective, but disturbingly it's the only one matching any of the spam.
15:24:06 <b_jonas> prooftechnique: wait, you have a keyboard layout that involves that classic stenography thing layout as one of the shift states? that looks crazy
15:24:17 * b_jonas is reading the scrollback of #esoteric... weird
15:25:27 <oerjan> @tell ais523 after pondering why rule 3 didn't catch one of the spams, i've concluded it actually checks for at least _two_ links hth. (also one of DatCodingGuy's edits would have got caught if it only checked for one.)
15:26:20 -!- MoALTz has joined.
15:26:22 <HackEgo> [wiki] [[Special:Log/abusefilter]] modify * Oerjan * modified [[Special:AbuseFilter/3]] ([[Special:AbuseFilter/history/3/diff/prev/53]])
15:33:27 <oerjan> today's long log just disappeared in a puff of magic(TG)
15:37:40 -!- Phantom_Hoover has quit (Read error: Connection reset by peer).
15:41:25 -!- Phantom_Hoover has joined.
15:46:37 -!- AlexR42 has quit (Quit: My Mac has gone to sleep. ZZZzzz…).
15:49:58 <int-e> hmm, seems good enough
15:52:07 -!- carado has quit (Read error: Connection reset by peer).
15:52:09 <int-e> wait, why is my academic account getting metaspam?
15:52:24 <int-e> "Would you be interested in *Bentley Systems Users List* for your sales and marketing campaigns?"
15:52:24 -!- carado has joined.
15:53:03 <int-e> (where in this instance, "metaspam" is spam that sells addresses to spam... physically, it appears)
15:53:36 <oerjan> it's because you're such an enterprising guy
15:54:09 <int-e> Am I now... why am I always the last person to learn about these things.
15:54:33 <int-e> I should turn lambdabot into a profitable business
15:54:40 <oerjan> AAAAAAAAAAAAAAAAAAAAAAAAAAA
15:54:57 <oerjan> maybe you could add a dow jones plugin
15:55:56 <int-e> ("all four letter acronyms look the same to me"... wasn't there an xkcd with approximately that title?)
15:57:10 <b_jonas> int-e: wasn't that all series of seven?
15:58:08 <int-e> b_jonas: ah, you're good!
15:58:30 <int-e> Yes, that's what I meant.
15:58:39 <oerjan> however, https://xkcd.com/1460/ is relevant to this channel.
16:00:37 * oerjan goes to explainxkcd for two of those
16:00:39 <int-e> has randall ever turned up here?
16:01:24 <lambdabot> LOWI 231450Z 08004KT 030V150 9999 FEW070 26/13 Q1026 NOSIG
16:01:31 <b_jonas> int-e: not on #esoteric I believe, but I think he was on irc at some point... or not? I dunno
16:01:43 <int-e> (surprisingly pleasant... I guess it's the low humidity)
16:03:01 <lambdabot> ENVA 231450Z 26004KT 9999 BKN035 16/10 Q1025 NOSIG RMK WIND 670FT 31003KT
16:10:40 -!- int-e_ has joined.
16:10:41 -!- int-e has quit (Disconnected by services).
16:10:48 -!- int-e_ has changed nick to int-e.
16:16:19 <quintopia> i dont think i ever realized until looking at that just now that smfw is "see my face when"
16:22:35 <int-e> hmm I wonder how many PhD theses will arise out of the The DAO debacle and subsequent Ethereum fork... this must be fascinating to study in detail.
16:23:50 <int-e> (from an economic and sociological perspective; technically I think it's comparatively boring, though apparently since the two chains are quite similar there are now replay attacks using transactions intended on one chain on the other)
16:28:02 <izabera> http://www.cs.dartmouth.edu/~doug/aqsort.c 3. Partitioning is a contiguous phase of n-O(1) comparisons, all against the same pivot value.
16:28:09 <izabera> what does this requirement mean?
16:30:00 <int-e> that you select a single pivot to partition the list into two parts, rather than, say, pick 4 pivots to partition the list into 4 parts in a single pass. (that would not be very quicksort-ish)
16:30:59 <int-e> Or perhaps one could dynamically change the pivot while partitioning (keeping track of the minimum of the larger part and the maximum of the lower part along with the currently selected pivot)
16:37:55 <b_jonas> wow. Conspiracy: Take the Crown is not even in http://magic.wizards.com/en/game-info/products/card-set-archive . They usually put sets there much before they're released.
16:38:09 <izabera> int-e: how would that book keeping help?
16:38:34 <b_jonas> I was wrong. zzo38: release notes (setfaq) for Conspiracy 2 is out: http://magic.wizards.com/en/articles/archive/feature/conspiracy-take-crown-release-notes-2016-08-22
16:38:51 -!- AnotherTest has quit (Ping timeout: 264 seconds).
16:39:44 -!- oerjan has quit (Quit: Later).
16:40:23 <int-e> izabera: it would help defeat awsort because it assumes that there are long runs of comparisons involving the same key
16:52:09 -!- carado has quit (Read error: Connection reset by peer).
16:52:30 -!- carado has joined.
16:59:17 <zzo38> Retro chess problem, to figure out the last few moves: 2QRNbqr/pk1pBppp/RppKp3/P2N4/8/7P/2P1P1P1/5B2
17:00:48 -!- Phantom_Hoover has quit (Ping timeout: 276 seconds).
17:03:36 -!- Jafet has joined.
17:06:41 <Lymia> <int-e> (from an economic and sociological perspective; technically I think it's comparatively boring, though apparently since the two chains are quite similar there are now replay attacks using transactions intended on one chain on the other)
17:06:45 <Lymia> If you were intentionally forking
17:07:01 <Lymia> Couldn't you make the fork have a different hash algorithm or something?
17:07:16 <Lymia> (Or otherwise make it different enough)
17:09:54 <Lymia> I wonder if we could have a different "style" of BFJoust hills?
17:10:27 <Lymia> Monthly tournaments where competitors may be secretly submitted, and are revealed only at the end when the whole tournament is crunched.
17:10:54 <Lymia> This defuses Kiseki and similar nonsense.
17:11:38 <int-e> Lymia: you could and they didn't
17:12:16 <int-e> ref: http://www.coindesk.com/rise-replay-attacks-ethereum-divide/
17:12:48 <quintopia> meh kiseki is not actually a real problem. the problem is when you run your program enough times to push everything else off the hill, which is a jerk move
17:14:02 <Lymia> One instance of the proof of concept, then normal programs?
17:14:13 <Lymia> The scoring algorithm does converge onto 100-0 for everything else if something maintains a 100% win rate.
17:14:15 <Lymia> So... there's that.
17:14:32 <quintopia> but i do think it would be neat to have a micro hill. one that you tweet at, and the entire program is contained in the tweet
17:15:31 <quintopia> Lymia: one of the scoring algorithms does. its an issue i brought up when proposing it.
17:15:47 <quintopia> but there are others selectable from the scores page
17:18:42 -!- carado has quit (Remote host closed the connection).
17:19:11 -!- carado has joined.
17:22:32 -!- AnotherTest has joined.
17:22:45 -!- Lord_of_Life has quit (Ping timeout: 250 seconds).
17:25:38 -!- Lord_of_Life has joined.
17:31:47 -!- Caesura has joined.
17:36:44 -!- ais523 has joined.
17:43:58 <b_jonas> ais523, zzo38: for the M:tG set "Conspiracy: Take the Crown", comprehensive rules update and release notes (setfaq) are out now, see http://magic.wizards.com/en/articles/archive/feature/conspiracy-take-crown-release-notes-2016-08-22 and http://magic.wizards.com/en/gameinfo/gameplay/formats/comprehensiverules
17:44:50 <b_jonas> ais523: the strangest part to me is the artifact card Spy Kit, which is a less general version of an un-card.
17:45:30 <ais523> b_jonas: not really, _____ affected itself, spy kit affects other things
17:45:34 <quintopia> did i ever ask you how you got credit for an integral in the brain age tas?
17:46:02 <b_jonas> ais523: spy kit is weaker because it only affects permanents, whereas the unnamed card works in any zone
17:46:12 -!- Phantom_Hoover has joined.
17:46:26 <ais523> quintopia: I don't think so, but it's fairly simple
17:46:42 <ais523> they wanted a "really complex equation" to work as the answer to one of the questions
17:46:45 <ais523> and knew I was good at maths
17:47:00 <ais523> I thought it'd make more sense to do something that looked hard than actually was hard
17:47:11 <ais523> and for extra fun, I made it a joke equation
17:47:55 <quintopia> it went by too fast to get the joke. how did "they" know you?
17:48:20 <ais523> dwangoAC (the person who was presenting/commentating) and I have been working together on a TAS for years now
17:48:30 <b_jonas> ais523: that the unnamed card affects only itself barely matters if you're interested only in rules complications. you can use cytoshape to quicksliver elemental to add the unnamed card's ability to almost any permanent, except possibly an aura.
17:48:35 <ais523> (also I was expecting some people to screenshot/rewind to look at it)
17:48:38 <int-e> something like the degree of x in the polynomial (a-x)(b-x)...(z-x)?
17:49:02 <FireFly> yeah, you definitely have to pause to read it
17:49:04 <ais523> b_jonas: well there are things like placing spy kit on a "note this information during the draft" card
17:49:16 <ais523> that doesn't do anything but only because a rule was specifically added to prevent it doing anything
17:49:31 <b_jonas> ais523: I remmeber that one. was that one of the definite integrals that reduce to an integral of an odd function on an integral symmetric around zero?
17:49:52 <ais523> b_jonas: actually the function in question's a constant
17:49:55 <b_jonas> someone mentioned another of those definite integrals recently
17:50:35 <b_jonas> ais523: yes, that's why spy kit is *weaker*. it only changes names of permanents.
17:50:41 <ais523> but a) it's disguised behind a couple of trig identities, b) one half is easier to solve the definite integral of than another so most people who've tried to solve it haven't realised it's a constant function
17:51:08 <ais523> b_jonas: but spy kit is an easy way to get an unusual name combined with an unusual ability
17:51:13 <ais523> you can put it on experiment kraj, for example
17:51:21 <ais523> _____ only has abilities if you cytoshape it or the like
17:51:48 <b_jonas> ais523: or if you steal the ability from the unnamed card with a quicksilver elemental (or experiment kraj)
17:52:02 <b_jonas> ais523: spy kit might be easier to use in practice
17:52:20 <ais523> b_jonas: if you allow un-cards, there's also the spy kit + wordmail combo
17:52:27 <b_jonas> I'm just saying the unnamed card lets you do more rules breakage if you are willing to consider scenareos unrealistic in a game because they require ten card combos
17:52:32 <ais523> I don't think anyone's worked out quite how much of a stat boost that gives yet, but it's a lot
17:53:07 <ais523> also spy kit + magnetic theft + eradicate is a stupid combo that's got a lot of love on Reddit
17:53:14 <ais523> you can't do that with _____
17:53:53 <b_jonas> in http://www.komal.hu/forum/forum.cgi?a=to&tid=7&tc=4021 post [4006] is the definite integral that reduces to an integral of an odd function around a central interval
17:54:35 <b_jonas> ais523: ah yes, magnetic theft. that would be crazy
17:55:04 <ais523> it took people a while to figure out the magnetic theft trick, they were trying to do something with harmless offering before
17:56:18 <zzo38> Rule 612.6 says it count as a text-changing effect.
17:57:19 <b_jonas> ais523: Crumble to Dust could also work, but hard to pull off, you have to animate a land an opponent controls first, such as with Wind Zendikon.
17:57:24 <quintopia> ais523: is the joke just that you used trig functions to represent x^2/4
17:57:42 <ais523> quintopia: the trig functions entirely cancel each other out
17:57:49 <ais523> I used a few of the more obscure trig identities
17:57:56 <b_jonas> it can, but it also needs Magnetic Theft or some such thing
17:58:01 <ais523> (you won't find them directly in most trig identity lists, you have to rearrange the equation somewhat first)
18:00:02 <b_jonas> Dichotomancy should work against a spy kit
18:00:37 -!- carado has quit (Remote host closed the connection).
18:00:51 <b_jonas> oh! Exclusion Ritual works with the Spy Kit on your own bear
18:01:13 <quintopia> ais523: i solved the integral by hand and notice they cancel but i dont understand why
18:01:15 -!- carado has joined.
18:01:15 <b_jonas> just ignore me, I'm stupid
18:01:40 <ais523> quintopia: that's the effect I was going for
18:03:49 <quintopia> ais523: nvm i get it. i was ebtering it into desmos wrong
18:04:12 <b_jonas> Can you have more than four of the unnamed card in a deck if you change its name to different things before the match starts? In fact, can you have the unnamed card in your deck for a Standard constructed tournament if you change its name to Forest before the match starts?
18:04:32 <ais523> b_jonas: I was wondering about that
18:04:45 <ais523> I believe the answer would be yes if you could somehow generate {2} during deck construction
18:04:57 <b_jonas> Wait, I should look that up in maro's old un-faq
18:05:45 <b_jonas> http://magic.wizards.com/en/articles/archive/ask-wizards-june-2007-2007-06-01
18:09:50 <quintopia> ais523: it would have taken less space to do int_blah^(blah+8) 2cos^2(blah*x/2)-cos(blah*x). is the factor of 1/2 just to make it uglier?
18:11:08 <ais523> quintopia: pretty much
18:11:14 <int-e> http://www.mtgsalvation.com/cards/conspiracy-take-the-crown/27511-spy-kit ... mm, the oracle text has a typo, but is it in any way official?
18:20:32 <b_jonas> int-e: there's no oracle text released yet
18:20:49 <b_jonas> int-e: that's not an official oracle text, it's just a text typed from some preview photos for that site
18:21:31 <b_jonas> int-e: you'll have oracle text when it appears in Gatherer, although it will probably be the same text as appears in the release notes http://magic.wizards.com/en/articles/archive/feature/conspiracy-take-crown-release-notes-2016-08-22
18:56:05 -!- hppavilion[1] has joined.
19:00:48 -!- hppavilion[1] has quit (Ping timeout: 244 seconds).
19:11:37 -!- carado has quit (Remote host closed the connection).
19:32:39 -!- hppavilion[1] has joined.
19:48:36 -!- hppavilion[1] has quit (Ping timeout: 250 seconds).
19:50:25 -!- hppavilion[1] has joined.
19:58:50 -!- carado has joined.
20:13:18 -!- sebbu has quit (Ping timeout: 250 seconds).
20:17:45 -!- ais523 has quit.
20:21:56 -!- sebbu has joined.
20:30:27 -!- augur has joined.
20:36:08 -!- MoALTz has quit (Quit: Leaving).
20:37:18 -!- ocharles has quit (Ping timeout: 250 seconds).
20:40:06 -!- ocharles has joined.
20:53:52 -!- hppavilion[1] has quit (Ping timeout: 265 seconds).
20:58:12 -!- carado has quit (Remote host closed the connection).
21:00:38 -!- carado has joined.
21:05:57 -!- carado has quit (Remote host closed the connection).
21:06:30 -!- carado has joined.
21:17:40 -!- hppavilion[1] has joined.
21:28:51 -!- Zarutian has joined.
21:29:46 -!- Zarutian has quit (Read error: Connection reset by peer).
21:29:53 -!- Zarutian has joined.
21:48:48 -!- hppavilion[1] has quit (Ping timeout: 250 seconds).
21:50:22 -!- Reece` has quit (Quit: Alsithyafturttararfunar).
21:53:50 -!- AnotherTest has quit (Quit: ZNC - http://znc.in).
21:53:58 -!- hppavilion[1] has joined.
21:54:40 -!- ais523 has joined.
21:57:41 -!- carado has quit (Remote host closed the connection).
21:58:33 -!- hppavilion[1] has quit (Ping timeout: 240 seconds).
22:19:21 -!- Jafet has quit (Ping timeout: 244 seconds).
22:22:31 -!- `^_^v has quit (Quit: This computer has gone to sleep).
22:44:08 -!- moonythedwarf_ has joined.
22:44:47 -!- jaboja has joined.
23:12:44 -!- Sgeo_ has joined.
23:12:59 -!- boily has joined.
23:20:20 -!- oerjan has joined.
23:23:44 <boily> Phantom_Helloover. I do, somewhat?
23:24:04 <boily> hellørjan. do you crypto, bro?
23:24:43 <Phantom_Hoover> can you get a sense of what https://github.com/PhantomHoover/trepan/blob/master/machine.js#L2044 is trying to do?
23:25:14 <HackEgo> [wiki] [[Special:Log/block]] block * Ais523 * blocked [[User:175.44.4.163]] with an expiry time of 1 year (anonymous users only, account creation disabled): persistent spammer
23:27:20 <boily> Phantom_Hoover: looks like it extracts a 128 bit fingerprint or IV from a 32 bit number.
23:28:39 <Phantom_Hoover> v0 through v3 are 32-bit words in a 128-bit input value
23:31:32 <Phantom_Hoover> see my instinct was just to treat it as a black box and find the solution to the puzzle elsewhere
23:31:52 -!- Sgeo_ has quit (Read error: Connection reset by peer).
23:31:59 <oerjan> helloily. only through putty and my browser hth
23:32:39 <Phantom_Hoover> scroll to address 6666 and overwrite the ! with a space
23:34:21 <boily> what the fungot...
23:34:23 <fungot> boily: mr president-in-office of the council' s conclusions because we do indeed focus upon money rather than making those workers who already have difficulty making sense of all the others, i am delighted to acknowledge the fact that the council must now be content with mere rhetoric. it must stop serving mammon ciba-geigy, that is the origin of the various research agencies which are concerned with land-use planning, to avoid
23:34:48 <Phantom_Hoover> best i can tell it was posted on some of the chans in autumn last year
23:35:28 <Phantom_Hoover> i mention this because then and now i could find no evidence that anyone got any further than me
23:36:01 <Phantom_Hoover> and back then i reverse-engineered and disassembled the source and i couldn't find any way forward
23:36:25 -!- Sgeo has joined.
23:36:54 <Phantom_Hoover> boily, the next steps are, btw: scroll up from the skull and you'll see a thought bubble from it saying '00005fe4'
23:37:46 <boily> it says 00005fe4 indeed. I am mystified.
23:37:47 -!- Sgeo has quit (Remote host closed the connection).
23:37:56 <Phantom_Hoover> write that address (little-endian ofc.) where it says 'code offset' on line 00006640 and overwrite the space
23:38:05 -!- Sgeo has joined.
23:38:56 <Phantom_Hoover> do that and it'll fill out the password box below the skull for you
23:39:32 <Phantom_Hoover> then you do basically the same procedure but with 61b0 rather than 5fe4
23:40:11 <Phantom_Hoover> that turns the screen red and writes 'keep going' above the skull, and that's as far as i've got
23:40:47 <oerjan> <Lymia> The scoring algorithm does converge onto 100-0 for everything else if something maintains a 100% win rate. <-- pretty sure i discussed at some point how to avoid that ruining the hill
23:41:28 <boily> I tried 0x00000000 as an offset. it freezes :D
23:41:57 <oerjan> by dividing into strongly connected components, essentially. although it was probably not implemented.
23:42:48 <oerjan> it wouldn't solve the problem of someone submitting several copies of the same program, anyway.
23:43:33 <Lymia> Per-user program limit?
23:43:38 <Lymia> If someone submits more than, say, 5 programs
23:43:47 <Lymia> Bump the lowest scoring program that they've submitted
23:43:52 <Lymia> Rather than the lowest scoring programp eriod
23:44:04 <oerjan> Lymia: that's not immune to nick changes.
23:44:16 <olsner> Phantom_Hoover: the magic numbers seem to contain digits of pi at least, the rest might be a simplification of a random hash function or something... maybe it's been broken in a way that you can generate a colliding password easily
23:44:29 <Lymia> It also ends the era of ais523
23:45:09 <ais523> right, BF Joust may be broken at this point
23:45:18 <ais523> I think we may need to keep submissions secret, at least in the details
23:45:43 <Phantom_Hoover> olsner, i googled the magic numbers and they're identical to those in ftp://ftp.maths.tcd.ie/src/security/ssh-2.0.13/lib/sshcrypt/dlfix.c
23:47:17 <Lymia> ais523, my proposal is monthly hills.
23:47:32 <Lymia> Through the month, people can submit programs which are kept secret until the evaluation at the end.
23:47:35 <ais523> the original BF Joust used weekly competitions, IIRC
23:47:42 <ais523> so monthly competitions would make sense
23:47:46 <Phantom_Hoover> basically my fear is that, because i can't solve the puzzle by doing stuff in the editor, maybe you have to cryptanalyse it to solve it?
23:47:56 <ais523> we should still probably have an instant-feedback hill for practicing on
23:48:05 <ais523> but possibly with an unofficial rule that you shouldn't metagame on it
23:48:17 <oerjan> Phantom_Hoover> that turns the screen red and writes 'keep going' above the skull, and that's as far as i've got <-- what happens if you try 61b0 + 2*(61b0-5fe4) ?
23:49:01 <Lymia> Phantom_Hoover, is this meant to be a revere engineering challenge?
23:49:57 <Phantom_Hoover> not in any place i could find that was indexed by google
23:50:33 <Lymia> Well, from revere engineering, what's the general nature of the puzzle?
23:51:59 <Phantom_Hoover> oh of the VM? it's a custom RISC-y CPU using memory-mapped IO
23:52:16 <Lymia> Is that what's being displayed on screen?
23:52:27 <Lymia> (So reverse engineering just gets you the VM spec, boring)
23:53:04 <oerjan> <Phantom_Hoover> is that a joke oerjan <-- no, but i hadn't seen your googling comment
23:53:26 -!- hppavilion[1] has joined.
23:54:21 <oerjan> Phantom_Hoover: it's just the simplest interpretation of "keep going" i could think of.
23:54:42 <Phantom_Hoover> the puzzle, as far as i can tell, is to decrypt the block of data from 60f0 to 65bf
23:56:17 * oerjan isn't actually looking at the puzzle, mind you
23:56:24 <Phantom_Hoover> there's a breadcrumb trail of hooks in the hex editor that will decrypt a password box and fill it with a password that will then decrypt down to 61cf, which reveals the hook that turns the screen red
23:56:59 <Phantom_Hoover> the problem is that from raking through the disassembly i'm confident that there are no more hooks in the editor left to activate after that