Follow

Here we go, after 1.5 years and a pile of c++ code later I'm pleased to announced my basilisk story is FINISHED!

"Basilisk collection - From Wikipedia, the free encyclopedia"
suricrasia.online/unfiction/ba

if you're wondering what c++ code I'm referring to, I'm talking about this code I wrote in order to help me find good hashes for the story github.com/blackle/Basilisk-Hu

@SuricrasiaOnline adskjsf i was wondering if you actually calculated those hashes, that's wild

@snailerotica it's been running on a single thread on my computer for the last year (because if my computer goes idle it crashes for some reason)

surprisingly the smallest hash I was ever able to produce occurred within days of running it full time

@SuricrasiaOnline this is very good

also this is extremely nit-picky but aren't cryptographic hashes designed to be slow

@SuricrasiaOnline your background section says that an ideal cryptographic hash function is fast

@monorail oh right, that comes almost directly from the beginning of the Cryptographic hash function wiki article en.wikipedia.org/wiki/Cryptogr

@monorail I think it depends on the hash function, some are designed to be fast and others are designed to be very slow

@SuricrasiaOnline @monorail yes

hashes used for storing passwords are generally the only case where you *want* it to be slow

@monorail the fast ones are good for when you want to hash a giant file to check a signature on it, the slow ones are for passwords/proof-of-work systems

@SuricrasiaOnline gotcha :blobthumbsup:

i wouldn't have said anything if it didn't seem like you were going for extremely hard scifi and that caught me off guard, so i'm glad to know it was the me who was in the wrong

@monorail thanks, yeah I'm hoping it doesn't have too many inconsistencies

@SuricrasiaOnline @monorail there are hash algorithms specifically designed to be slow, mostly to make brute forcing them difficult so you can't e.g. crack a password easily. SHA-256 is not one of these and this is one of the complaints about Bitcoin vs other proof of work cryptocurrencies, which led to e.g. Litecoin using scrypt. That said, the whole concept of burning compute power for, uh, making free money?? is an inherently flawed concept regardless of the proof of work system used.

@SuricrasiaOnline @monorail alternative hashes used sometimes require excessive amounts of memory to make parallelizing them difficult regardless of how fast the algorithm itself is. Funny how there's a bunch of research to make proof of work harder to break, which is ironic considering proof of work is only useful for terrible things in the first place.

@monorail no, not all of them. from the cites in the article, they say that sha-256 is quick to verify, but slow to compute a partial hash inversion. see "proof of work".

@monorail cryptographic hashes for storing passwords absolutely should be slow, though. I think the standard for that is b-crypt?

@monorail @SuricrasiaOnline actually they aren't, at least not as long as it does not impact security. Key derivation functions on the other hand, works in the same kind of way, but are deliberately slow to compute (and generally configurable in term of computation time required)

@SuricrasiaOnline holy shit, even after clicking the link knowing this was fiction I *forgot* partway through and went to consult the sources in amazement. Excellent job.

@SuricrasiaOnline I fucking love this. This is amazing. It's written so well, the subject matter is accurate, and I love how it plays with the boundary between the real world and fiction, like how some of the Wikipedia page links are real and fake, and some of the sources are real and fake, and some of the named attacks. I appreciate that the sample of basilisk.txt given in the article is accurate.

@SuricrasiaOnline ngl this reads like a good SCP story, and i mean that in a good way

Good god this is well-done

@SuricrasiaOnline why does the "Wikidata item" link in the sidebar go to the page about Unix? :p

@pcy that was the template I used lol, I guess I missed a spot when I was removing references to it

@SuricrasiaOnline I was so excited for the basilisk trashing the shitcoin, and now I only feel emptiness inside.

@SuricrasiaOnline This is brilliant, and well worth the compute time for the extra verisimilitude.

@SuricrasiaOnline this is awesome! I thought it was real for a bit! I love the writing!

@SuricrasiaOnline this is weirdly unsettling for how dry it is, good job

@leo @SuricrasiaOnline you know seeing other unfiction like this has me thinking about the alameda weehawken burrito tunnel, which has a wildly different tone but is unmistakeably unfiction idlewords.com/2007/04/the_alam

@shoofle @leo "it took six months to persuade suspicious taqueria owners to switch to a salsa with lower magnetic permittivity" oh my GOD

@SuricrasiaOnline @leo RIGHT?! this is one of my favorite pieces of writing

@shoofle @SuricrasiaOnline @leo imagine that I am able to express how fantastic that burrito tunnel article is and I wrote those words here

@shoofle @SuricrasiaOnline this is amazing

We really have a unique situation here - a population of fifteen million people without access to high-quality Mexican food.

@leo @SuricrasiaOnline for me the most horrific part of this story were the two words "red motivation"

just, everything it *implies*

@SuricrasiaOnline @leo like the fact that it's a thing people do *and* that it's euphemistically referred to as 'red motivation' *and* that it's accepted enough that the euphemism passes without comment

@SuricrasiaOnline Great story.

Personal opinion: I think it can be even better with even more technobabbles by mentioning the rumor that its creator has a large quantum computer capable of running the Grover's algorithm, and how cryptographers say it doesn't actually work, first because such a computer doesn't exist, also because it requires XXXX qubits and XXX millions Toffoli gates, also because 256-bit hash still has 128-bit of preimage resistance post-quantum. Nevertheless, you can write that some researchers from PQCrypto 2020 raised the concern that the unknown attack may gain a quadratic speedup on a quantum computer, raising questions to the current belief that doubling the hash size is effective to stop quantum attacks.

You can cite DJB's trolling paper: cr.yp.to/hash/quantumsha3-2010

@SuricrasiaOnline absolutely love the citation on the speculation that the basilisk collection is from satoshi

@SuricrasiaOnline uhhh could I strongly encourage you to NOT rip off wikipedias layout wholesale? a lot of links go back to wikipedia which is very confusing and kinda makes your site seem like a scam (I know it isn't).

if you want it released as a wiki-esque site, consider deploying mediawiki perhaps? Would probably be a lot easier than whatever you're using at the moment too.

Anyway, nice work.

@techpriest yeah, I guess it might be wrong to use the layout but I really want the immersion that it provides

I may add a internet-archive-esque banner at the top that clarifies that it's fictional

@SuricrasiaOnline yeah that would help a lot. it confused me at first to the point where id try looking up the page on wikipedia (caution with clicking random links yknow...) then got really confused when it didn't show up there.😅

a banner indicating it would probably help a bit.

@SuricrasiaOnline this is SO GOOD and I still want to know what a Hand of God attack is

@SuricrasiaOnline this is great, and I was rotfl at the Satoshi bit at the end

@unascribed gonna disappear into the mountains with a stack of textbooks on preimage attacks and I'm not coming back until I build the basilisk myself

Sign in to participate in the conversation
Cybrespace

cybrespace: the social hub of the information superhighway jack in to the mastodon fediverse today and surf the dataflow through our cybrepunk, slightly glitchy web portal support us on patreon or liberapay!