|Paul (15 Dec 2004 16:16): Should I delete the below messages? They're sort of amusing, though, maybe...|
|Paul (16 Dec 2004 18:48): then again, maybe not...|
|m (16 Dec 2004 20:41): I'm getting spam at my gmail account even though I've never posted the email address anywhere. I wonder if spammers send to all [dictionary]@[popular_email_host].com
addresses and then ignore the bounces.
|Paul (17 Dec 2004 18:11): yes. solutions:
|Paul (17 Dec 2004 18:14): By the way, if any of you are trying to email me, my address is pgd@post.[my college's domain]|
|m (18 Dec 2004 2:01): sender authentication without having personally ok'ed the sender is tricky.|
|m (18 Dec 2004 2:03): Actually, I think collin jackson did his senior project on something related.|
|Paul (19 Dec 2004 12:30): to tag email as from spam-IPs, good-IPs or unknowns...all one needs is sender authentication and a few competing IP-list sites, ne?|
|m (21 Dec 2004 14:31): maybe, but email headers are easy enough to forge. I was thinking that a spammer could just prepend some fake initial header data to the email and pretend to be forwarding on a valid email, but you could filter based on intermediate-node IP. it still seems like it would be straightforward to forge TCP header information and make the intermediate email routing step seem to come from a legit host.|
|m (21 Feb 2005 21:04): how did spammers ever find mancala? Have any of you ever linked mancala from somewhere? Maybe an evil router watches http post traffic and grabs urls with post data associated with them to run through a message-board-bot?|
|Paul (22 Feb 2005 18:35): Well, I certainly haven't tried to hide mancalablog...on the contrary, I originally wanted the archive to be listed in Google, in case some poor searcher wants to know about equilateral triangles on lattice points or Kagawa Compulsary Education..... And spam has hardly been significant...3 posts in 8 months, I think.|
|m (22 Feb 2005 21:24): still, google bots can't find you unless someone already in their (vast) database links you, right?|
|Paul (24 Feb 2005 16:31): Well, you should be able to get such a list by Google-ing for "link:manc.a.la", but it comes up empty. I think Google may now index pages it finds, regardless of whether they're linked from anywhere... But watch out: similar queries often come up empty even when Google has a cached copy of the page with the link!|
|Grant (25 Feb 2005 1:53): Too creepy!|
|zong (25 Feb 2005 10:05): Google's motto, according to Larry Page, is "Don't be evil." Whether it is that or just "Don't appear to be evil," one will never know.
The technical problem with Google is that it is supposed to just measure the properties of the network. But because it does a good job in measuring the network, it modifies the network via a feedback process through it.
|Grant (27 Feb 2005 18:30): Wait, I thought the problem was concerns of privacy and intellectual property?|
|zong (28 Feb 2005 12:16): That comes from Google being (market-wise) powerful, I guess; but that's the same thing. Its power is in modifying the network (here used in a sense that includes traffic patterns), which is a screwy thing to do for a search engine but inevitable for a successful search engine.|
|Grant (28 Feb 2005 19:30): I'm not understanding how you put '(semi-)covertly gathering all sorts of personal information for dubious purposes' under the heading of 'modifying the network.'
I'm not really all that concerned about how Google's presence modifies the internet - stuff like Google bombs and web pages that talk about Google don't bother me at all. I'm more concerned about how Google's presence have effects outside of the web - possibly identifying or implicating users via their searches, or giving government (or even commercial) agents the means by which to do so.
That, and knowing that by sending email to gmail, I could essentially be handing it over to Google for analysis, indexing, and who knows what else.
|m (1 Mar 2005 22:20): seems reasonable|
|Grant (2 Mar 2005 3:53): It is, probably, in some ways just fine. And actually, some of the stuff on google-watch does seem pretty paranoid. One might even say 'crazy.' The points that I thought might be valid follow.
The provisions for disclosing personal information are pretty vague, and suspect to change at any time. Furthermore, 'necessary by law' - who's law, sucker? My law, or The Man's law? No, seriously, though - somebody brought up an interesting point about requests from other countries with somewhat dubious governments.
I also heard that personal communications ceased to be protected by (our) federal law after 90 days or something? Which means that the US government could legally peruse email via a subpoena rather than a search warrant? I end all my sentences in question marks? Like this?
I can't say what protections would be given for foreign (non) citizens - presumably, less.
As far as the actual search engine goes, I am personally of the opinion that that sort of thing should fall under fair use. But I don't kid myself for one second - Google is copying other people's works (commercial and otherwise) without any sort of permission, and it is making a profit out of it.
But, like, I still use it...
|m (2 Mar 2005 10:47): yes. presumably less.|
|m (3 Mar 2005 18:10): whoah! google maps is straight sexy.|
| (3 Mar 2005 23:04): Yeah. I just wish it had maps of Japan. That'd be awesome.
I have a hard time finding things, sometimes. It's only mitigated somewhat by the fact that I don't have a car.
|m (7 May 2006 0:29): god damnit|
|Paul (11 May 2006 9:22): What the hell? It's not like this site has any pagerank or readership. Any suggestions?|
|Paul (11 May 2006 12:10): okay, adding a simple spamfilter, sigh|
|Paul (12 May 2006 5:54): tweaking the spamfilter again :(|
|Paul (28 May 2006 18:53): I am disappointed to report that we've gotten 58 spam posts since I tweaked the spamfilter. The domains they link to are registered to a guy in Minsk, Belarus. How many person-hours and social interactions is this one guy probably wasting/ruining? Like, just protecting mancala from him took me almost an hour iirc. I guess spam isn't as bad of a utilitarian evil as war etc., but it ranks right up there.|
|Grant (29 May 2006 17:35): I'm pretty sure they're both bad, but that's about as far as I'd be willing to take that comparison.|
|m (10 Feb 2007 21:47): subject text: Is he Pythagorean|
|R (17 Jun 2007 20:38): Subject: I guess some people expect their software to automatically do whatever they think it should do, and be perfect and contain no bugs.|
Fucking awesome. Ripped from thomasvs' Advogato page. Home of other cute rants like:
Never ever ever install stuff from source to /usr unless you are completely sure that this is what you want and the only way to work around something. If you don't know if it is, the answer is it isn't.
|m (28 Jun 2007 9:25): today's spam to make it past the filter is|
(a) upsetting: "Subject: Who has the bigger pen1s.? g"
Damnit. I always suspected.
(b) absurd: invitation.pdf
It's an incoherent kidnapper's ransom-note style pdf telling me to buy some sort of stock. I don't know. (don't worry -- there's nothing embedded in the pdf besides a little image data)
|m (26 Jul 2007 23:14): Shit. R, your dancing post from a couple weeks ago tripped p's spam filter. I don't know on what impulse, but I was browsing his spam.dat file that logs everything here flagged as spam. So, uh, maybe I can just cat it onto the end of blog.dat? I'm a little afraid of breaking mancala, though.|
|m (26 Jul 2007 23:18): not quite.|
P: chmod blog.dat g+w
|R (7 Dec 2007 18:59): Subject: "Men think with a penis - the bigger penis, the smarter man."|
Subject: "Big penis is great - it's an axiom."
|G (8 Dec 2007 19:37): Those are awesome. Maybe I should start reading my emails, too! Maybe I should start digging around in this pile of shit, looking for gold!|
|G (9 Dec 2007 1:32): I, uh, didn't mean that to say that those emails you guys sent me aren't awesome. They're great. Really.|
|m (9 Dec 2007 14:27): G's .procmailrc:|
|m (9 Dec 2007 14:34): To actually match the character class and to avoid vacuous matching, that sed line should really instead read|
| sed -re "s/[[:alpha:]]+/shit/g"
|G (18 Dec 2007 20:10): Hey, I got one!|
"You Dont please with your male aggregate size."
|G (22 Oct 2008 1:17): Also, it looks like your spam filter doesn't catch complete gobbledygook.|
|m (22 Oct 2008 16:20): Find the spam filter on grza and take a look at it -- it's pretty goofy.|
|m (2 Nov 2008 6:25): I was just thinking: bayes filtering can't catch clever messageboard spam, because the spam can simply use the messageboard itself to generate its own bayes-likely message, with a few smarmy urls in it.|
I kind of want to see that: a generic mancala message
|g (2 Nov 2008 13:11): But the spam we're actually getting isn't that - it's just nonsense with a bajillion URLs that don't even have link text.|
Also, generic mancalablog post in 3... 2... 1...
|Paul (3 Nov 2008 15:52): using the phpbb-style urls as a trigger for the spamfilter seemed promising, but Jeffrey's spam doesn't have any of them...|
If you've ever tried dadadodo, you'd find some good random mancala messages. Here's one coming up...
|Paul (3 Nov 2008 16:01): ~$ curl -s mancala.grza.net/?t=test |grep : |cut -f3- -d: |dadadodo - |head -20|
dadadodo: reading stdin...
Datte datte, datte. Apple used to. Paprika is no fun of clicking the
image ssh to just like being have it beyond the second or I still
remember suggest this be able to eternally battle mysterious terrorist
timebots, not hang of power you know why did it would bother spamming a
Yakuza But talk to this be it micycle? Nice to basic syntax. Test
the order cursor at maybe nobody got put their a rimming mazing, is
|Paul (3 Nov 2008 16:03): ~$ curl -s mancala.grza.net/?s=20000 |grep : |cut -f3- -d: |dadadodo - |head -20|
dadadodo: reading stdin...
........... 2309 lines
Work on some sort of keeping it to buy a while, the indirect, so.
Also probably be a Remote ssh y; and the government's purpose is
back in their account on all of the all of economics wifi signal?
Besides unla. Perhaps I'm humor you being dicks? Discuss.
Particularly good npr points in Seattle.
The some soundfonts or you a writ of Content? Minnesota? It
anywhere like leeks. Royal Tenenbaum Rollicking good the short
very old, to offered in line and say it's Thanksgiving here is now
Paul, has is long typing this lady site links I refuse to buy of
your bb I thought that! And yikes. Jesus. Sorry I agree: the
executive branch is it working. You wanted to implement the whole
thing I've ever again I knew that, fine; with regex's?
G's procmailrc: syntax these machines I hear was funny;
though. That many buffers are Obama's, campaign of them is
Stupid, but I saw that it's a standard system
and as a href and seems okay, preview. I
|Paul (3 Nov 2008 16:05): Conclusion: "Why did it would bother spamming a Yakuza?"|
|G (3 Nov 2008 16:35): Holy shit, that spam is awesome.|
|m (11 Nov 2008 20:58): :<|
|G (12 Nov 2008 0:44): |
|Paul (13 Nov 2008 13:52): Manually moved spam into blogspam.dat. Honeypot field pending.|
|Paul (16 Nov 2008 5:54): One honeypot field added. Now, can I still post?|
|Paul (16 Nov 2008 5:55): And...can I still post after previewing? Yes.|
|Paul (16 Nov 2008 5:57): And, can I still post from Elinks? Yes, but it shows the honeypot, so just don't put anything in it.|
|m (16 Nov 2008 7:56): Well done, dex. Stupid spam-flood.|
|G (16 Nov 2008 11:33): Nicely nicely.|
|m (29 Dec 2008 1:09): Subject: Your powerful uprise will surprise women|
[pretty great sight/slant-rhyme of uprise and surprise, even if it's a stretch to translate "uprise" to "boner".]
|m (15 May 2009 13:34): Subject: Hannibal the Great's Favorite sexual Positions and Libido Enhancers From Hitsory|
|R (11 Feb 2010 16:49): "Your shlong can be shlonger"|
"Like a drilling machine in pants"
|G (15 Mar 2011 17:46): I just really like this subject line.|
|Paul (23 Jan 2017 20:38): what happens when you reply to spam email rotfl|