--- Log opened Tue Jul 22 00:00:48 2025
00:07 -!- Gooberpatrol66 [~Gooberpat@user/gooberpatrol66] has quit [Quit: Konversation terminated!]
02:02 -!- alethkit [23bd17ddc6@sourcehut/user/alethkit] has quit [Remote host closed the connection]
02:15 -!- darsie [~darsie@84-113-82-174.cable.dynamic.surfer.at] has quit [Quit: Avoid fossil fuels and animal products. Have no/fewer children. Protest, elect sane politicians. Invest ecologically.]
02:15 -!- darsie [~darsie@84-113-82-174.cable.dynamic.surfer.at] has joined #hplusroadmap
02:33 -!- etc-vi [~etc-vi@user/meow/girlchunks] has quit [Ping timeout: 252 seconds]
02:46 -!- etc-vi [~etc-vi@user/meow/girlchunks] has joined #hplusroadmap
03:10 -!- alethkit [23bd17ddc6@sourcehut/user/alethkit] has joined #hplusroadmap
03:22 -!- darsie [~darsie@84-113-82-174.cable.dynamic.surfer.at] has quit [Quit: Avoid fossil fuels and animal products. Have no/fewer children. Protest, elect sane politicians. Invest ecologically.]
03:23 -!- darsie [~darsie@84-113-82-174.cable.dynamic.surfer.at] has joined #hplusroadmap
03:58 -!- delthas [16abab341f@2a01:4f9:c010:cf0b::1] has quit [Remote host closed the connection]
03:58 -!- delthas_ [16abab341f@2a01:4f9:c010:cf0b::1] has joined #hplusroadmap
03:58 -!- delthas_ is now known as delthas
04:39 < L29Ah> but did they coerce The Gatalog into stopping publishing them?
05:29 < hprmbridge> kanzure> yudkowsky redemption arc  https://x.com/Rule3O3/status/1947531637211771082
07:01 -!- Bolichon [~Bolichon@187.252.206.18] has joined #hplusroadmap
07:01 -!- Bolichon [~Bolichon@187.252.206.18] has quit [Remote host closed the connection]
07:35 -!- Bolichon [~Bolichon@187.252.206.18] has joined #hplusroadmap
08:34 -!- Bolichon [~Bolichon@187.252.206.18] has quit [Read error: Connection reset by peer]
08:35 -!- Bolichon [~Bolichon@187.252.206.18] has joined #hplusroadmap
09:09 -!- Gooberpatrol66 [~Gooberpat@user/gooberpatrol66] has joined #hplusroadmap
09:46 -!- Bolichon [~Bolichon@187.252.206.18] has quit [Ping timeout: 260 seconds]
10:20 -!- Bolichon [~Bolichon@187.252.206.18] has joined #hplusroadmap
10:24 -!- Bolichon [~Bolichon@187.252.206.18] has quit [Client Quit]
12:21 < kanzure> mainline DHT spec https://www.bittorrent.org/beps/bep_0005.html
12:22 < kanzure> https://www.libtorrent.org/dht_sec.html
12:23 < RangerMauve> Mainline is a GOAT. I feel like hyperswarms improvements on it make it technically superior but it's hard to compete with the stability and spread of mainline
12:26 < kanzure> i have some gaps in my understanding
12:26 < RangerMauve> Feel free to ask Q's. DHTs are an area of expertise for me
12:26 < kanzure> what specifically is the incentive for running the bootstrap nodes or DNS seed nodes? i guess it's cheap to run? but what about legal defense costs? bittorrent.com was running one, or might still be.
12:27 < kanzure> for any other company maybe there's a strong argument that connecting peers together is mostly legally harmless, but for bittorrent.com specifically i'm not sure that argument would hold up to legal attack?
12:27 < kanzure> maybe it would. hrm.
12:28 < kanzure> also i have some confusion around resource exhaustion attacks or cost of running DHT nodes, or otherwise polluting the DHT. what prevents that? or what constraints are imposed? is it just local resource constraints (local policy)? how do you not end up with the public DHT stuffed exclusively with garbage?
12:29 < L29Ah> it doesn't do anything illegal
12:31 < kanzure> tell that to coinjoiners
12:31 < L29Ah> besides, i host an intermediate tor node for over 10y under my name and nobody cares
12:31 < kanzure> relay or exit?
12:31 < L29Ah> i don't recall any coinjoiner getting persecuted
12:31 < kanzure> isn't there one literally in court right now being prosecuted?
12:31 < kanzure> it's literally day 3 of court
12:31 < L29Ah> relay; i hosted an exit for a while but stopped after Bogatov got jailed
12:32 < L29Ah> link me one then
12:32 < kanzure> https://x.com/L0laL33tz/status/1944882856422973504
12:33 < L29Ah> and finance is other thing, you have lots of AML legislation but not anti communications laundering
12:33 < L29Ah> moreover free speech is a big issue in US so it is unlikely to ban encrypted communications any time soon
12:34 < L29Ah> no mentions of coinjoin there
12:34  * L29Ah has no clue how tornado cash works
12:34 < kanzure> regardless of whether something is actually illegal or even explicitly legal, legal defense is still a thing that has to occur in some cases
12:35 < kanzure> anyway, what about the resource exhaustion attack stuff
12:37  * L29Ah has no clue how bittorrent dht protects itself from ddos spam, mayhaps it doesn't
12:37 < L29Ah> also mayhaps it is too expensive to disrupt for good
12:38 < L29Ah> or nobody cares enough to hate torrenting anymore
12:39 < L29Ah> rutracker is still well reachable despite being the number one target for DDoS for anti-piracy coalitions
12:39 < kanzure> or is this one of those situations where the adversaries haven't actually looked into the details and nobody really cares.
12:46 -!- gl00ten [~gl00ten@2001:8a0:7ee5:7800:46d9:f5c:17a2:432] has quit [Ping timeout: 248 seconds]
12:53 < kanzure> why doesn't kademlia DHT actively rebalance- eg if a node goes offline, shouldn't peer lists or infohash mappings be sent by remaining nearby nodes to a new neighbor node that might be slightly farther XOR distance? i guess that won't work because that neighbor node wouldn't be able to distinguish legitimate behavior from resource exhaustion attack hmm.
12:54 < kanzure> kademlia redundancy at n=8 might be enough especially if doing something like node_id = hash(IP address), and then neigboring nodes in XOR infohash-node-id-distance-space are very likely to be geographically disjoint/decoupled. and maybe 8 is enough to preserve redundancy, although i don't see how without active healing.
13:04 < kanzure> how were torrent client DHT parameters chosen? why 300 kilobytes or 2.5 megabytes and not larger values?
13:05 < kanzure> seems like these values should be selected based off of long-term DHT node participation data, eg the current estimates of tens of millions of DHT nodes
13:05 < kanzure> and/or DHT lookup failure rates across an exhaustive list of infohashes.
13:07 < kanzure> IPFS seems to suffer from similar sybil attacks and content enclipse attacks as mainline DHT. possibly they are using kademlia too.
13:07 < RangerMauve> The incentive for running nodes is to keep the network healthy and to get your tool used so you can get more consulting gigs and b2b enterprise level stuff on top. :P
13:08 < kanzure> is there a particularly large b2b enterprise kademlia market?
13:08 < RangerMauve> Re polluting the DHT, mainline is large enough that it's not worth the cost. Taking down bootstrap nodes permenantly could be helpful but then trackers and PEX can pick up the slack
13:09 < kanzure> AFAIK facebook was using torrents for internal software distribution to their machine cathedrals but in my estimation i would think they did this internally instead of hiring b2b vendors to implement that
13:09 < RangerMauve> Generally the DHT is full of garbage, but you need to constantly republish the garbage every 30-60 mins or so so it's hard to keep it there
13:10 < kanzure> yeah mainline DHT looks much more byzantine and ephemeral than i would have expected. it makes sense tho given the behavior of most torrent clients or users.
13:10 < RangerMauve> The illegal content also isn't stored in the DHT. It's just random hashes mapped to IP addresses
13:11 < kanzure> yes i understand, the legal attacks would either (1) not actually care about that fact, or (2) focus on coordination for copyright infringement activities not actual direct copyright infringement (eg like google search results which have been fought in court with regards to DMCA takedown etc)
13:11 < RangerMauve> Regarding rebalancing, since you need to republish periodically any nodes that go down will be purged from your routing table eventually anyway
13:11 < RangerMauve> I think the packet sizes for UDP datagrams were where they came up with size limits
13:11 < kanzure> oh okay. hm.
13:12 < RangerMauve> IPFS's DHT is way smaller and way less efficient so it's much easier to overwhelm a given node
13:12 < RangerMauve> Game companies and companies interested in measuring internet traffic globally are potential business users. It's hard to break into the market
13:13 < kanzure> what happens if there is no node in XOR space that matches a certain infohash within the required ID/infohash distance metric? 
13:14 < RangerMauve> IPFS has had some effort on the "illegal hashes" use case where they curate lists of "bad hashes" they refuse to serve. Mostly for gateway operators though
13:14 < RangerMauve> YOu just get the closest node if there's no "match"
13:14 < RangerMauve> The closest node you can route to, or more likely "n" closest nodes
13:14 < kanzure> but the closest node won't necessarily have that infohash/relevant peer list?
13:14 < RangerMauve> Yeah
13:15 < RangerMauve> That's why you ask several closest nodes
13:15 < kanzure> why would any of them have it? those closest nodes are also outside the "required distance metric" too
13:15 < RangerMauve> Could you elaborate?
13:16 < kanzure> the scenario is that there is no node in XOR space that matches a certain infohash within the required ID/infohash distance metric
13:16 < RangerMauve> Generally you will rout to the same nodes that your peers would if you're on the same network
13:17 < kanzure> okay but if you announce your torrent to that node (outside the distance metric cutoff, but it's the closest one anyway), that node will just discard your announcement because it sees that it's too far away
13:18 < RangerMauve> I'm honestly not sure. You might be SOL at that point
13:19 < kanzure> weird. it might be possible the actual parameters for "distance" are set such that this is exceedingly uncommon for the whole distribution of infohashes. like maybe even it's based on the bytes prefix, first 8 bytes of the infohash or something, you could imagine that as an alternative to a distance metric that would cause the situation i asked about to not occur.
13:19 < RangerMauve> I've never encountered a situation where the closest nodes all reject hashes
13:20 < RangerMauve> TBH the required distance metric is new to me 😸
13:21 < kanzure> OK so in general the parameters of the system were (somehow) selected such that you can always find close node IDs to almost any arbitrary infohash you might have, so you end up talking to the right peers [even if nobody is currently announcing that infohash as available].
13:22 < RangerMauve> Is that supposed to be in the kademlia spec? https://www.bittorrent.org/beps/bep_0005.html Seems like announce_peer just takes whetever you give it
13:23 < kanzure> chatgpt says that the spec says "Responsibility for an infohash is implicit: a node stores values only if it is among the k closest nodes **that it knows of** to the key." so if it doesn't know of a responsible node, then a DHT node (behaving honestly or responsibly) should then accept responsibility even if it's outside the distance cutoff. cool.
13:24 < RangerMauve> That sounds like bullshit it made up
13:24 < RangerMauve> Err the other thing
13:24 < kanzure> well holding off from whether it is bullshit, is it a good idea
13:26 < RangerMauve> 🤷 It is a neat idea but I haven't heard it used in any contexts
13:26 < kanzure> the spec says the following: "After the search is exhausted, the client then inserts the peer contact information for itself onto the responding nodes with IDs closest to the infohash of the torrent."
13:28 < RangerMauve> The fun DHT stuff is using it for hole punching >:)
13:28 < kanzure> uh i might be wrong about the distance metric thing. weird.
13:29 < RangerMauve> I wouldn't trust a chatbot to give me information :P
13:30 < kanzure> the spec says: " "distance metric" is used to compare two node IDs or a node ID and an infohash for "closeness.""
13:30 < RangerMauve> But yeah the main attacks on DHTs are sybil attacks where you make a bunch of malicious nodes
13:31 < kanzure> obviously that should be solved by requiring PoW(node ID) instead of node ID because maybe IP address allocation is cheap
13:31 < RangerMauve> OR black hole attacks where you make a bunch of nodes with ids close to a target hash or a target node id. Then you have those nodes pretend like there's no peers but each other and no announces for a given hash
13:31 < kanzure> to conduct that black hole attack, you need to actually obtain relevant IP addresses right?
13:32 < RangerMauve> A guy did some calculations for the PoW approach and found that the amount you'd need to make it truly expensive to circumvent would make the UX on end users too awful
13:32 < RangerMauve> Right now mainline node IDs are random
13:32 < kanzure> node ID should be hash(IP address) at minimum, at least one round of hashing, because otherwise you can-- oh they are presently random? huh.
13:32 < RangerMauve> But yeah using your public IP to derive your node id is a fix to make sybil harder
13:33 < RangerMauve> This is why newer DHTs like the one in hyperswarm are appealing to me. Mainline was "good enough" (and still is) but there's been a lot of improvements since
13:33 < kanzure> explain hyperswamp
13:34 < RangerMauve> Another is to "un-announce" yourself from the DHT to keep the nodes clean and less noisy
13:34 < kanzure> most nodes are going to be very byzantine and not have graceful shutdown. why would they be expected to un-announce?
13:34 < RangerMauve> IT's like mainline but newer with fancier features like hole punching
13:34 < RangerMauve> In your app code you can add the unannouce step for graceful shutdowns and it helps 🤷
13:35 < RangerMauve> IT's not a requirement of the network but every bit you add to your application makes it faster. E.g. next time you try to find peers you'll have fewer false positives
13:37 < RangerMauve> False positives due to churn are a major source of performance issues on mainline
13:38 < RangerMauve> https://docs.pears.com/building-blocks/hyperdht
13:38 < kanzure> churn is believable, but how much of a performance impact could that be, simply do multi-threaded DHT crawling to get your info. how hard is it to ping a few thousand peers.
13:38 < RangerMauve> They have _some docs_ but honestly you need to read the code / be in the know to get a lot of this info
13:38 < RangerMauve> Not hard per se, just more latency
13:38 < RangerMauve> The bottleneck is less CPU bound and more network IO bound
13:39 < kanzure> yeah ok it's roundtrips
13:39 < kanzure> so the latency scales, at minimum, with the "DHT depth" or how many hops you need to iteratively query
13:39 < RangerMauve> Yeah
13:40 < RangerMauve> This is also why peer exchange and local peer discovery are handy supplements to the DHT
13:40 < kanzure> how bad is it in practice? in terms of seconds
13:41 < RangerMauve> Unsure tbh, in my not super scientific tests I got like 3-15 ish for querying the dHT. But I was using a super unoptimized impl
13:41 < kanzure> 15 seconds total, or per roundtrip?
13:41 < RangerMauve> Total
13:41 < kanzure> that's nothing
13:41 < RangerMauve> It's a lot compared to website loading 😅
13:42 < RangerMauve> But yeah I agree. It's pretty speedy
13:42 < kanzure> ok maybe for IPFS content retrieval
13:42 < kanzure> most people aren't storing their short-form web content in a way that requires a 15 second DHT lookup
13:42 < RangerMauve> Since you're doing UDP and don't need to do a handshake you can send a shitload of packets and process them as they come
13:43 < kanzure> what is the performance of lookups and retrievals in IPFS land?
13:43 < RangerMauve> There's ways to speed it up too. If you make your bucket sizes larger and prepopulate your table you can get sub 1s latency
13:43 < RangerMauve> Total shit
13:43 < kanzure> cool, cool
13:43 < RangerMauve> They use TCP connections and like three layers of abstraction on the wire
13:44 < RangerMauve> But you can also get it to sub 20s if you optimize a bit
13:44 < RangerMauve> My web browser is pretty speedy with public content after I tuned it a bunch
13:44 < kanzure> juan benet always seemed like the kinda guy that discovered adderall only yesterday and didn't figure out how to use it right
13:44 < kanzure> i can't keep saying that because it's been almost 10 years now
13:45 < RangerMauve> ipns://agregore.mauve.moe/ took me like 15 seconds to load actually
13:45 < RangerMauve> But I also have a seedbox for it and optimized it. 🤷 It's harder to get home-to-home data transmitted
13:46 < kanzure> have you seen PKARR or PKDNS
13:46 < RangerMauve> FWIW the IPFS folks are still doing improvements. The QUIC based overhaul should help
13:46 < RangerMauve> I know the guy that made PKARR and he talked about it a while ago but I haven't kept up to date
13:48 < kanzure> i have been looking at mutual credit protocols (or lack thereof) which is why i was looking at the DHT spec.
13:48 < RangerMauve> What's your use case?
13:49 < RangerMauve> And how do DHTs relate?
13:49 < kanzure> use case is decentralized p2p credit
13:50 < RangerMauve> Oh! Also the veilid DHT is pretty cool. https://veilid.gitlab.io/developer-book/concepts/dht.html Unlike the others it has IP privacy
13:50 < RangerMauve> It's still kinda slow but at least the slowness comes with privacy
13:51 < kanzure> my intuition would prefer something like DHT for routing instead of lightning-style/tor-style public directory
13:52 < RangerMauve> Like for connecting folks or for storing attestation that the transaction happened?
13:52 < kanzure> even ant pheromone or floodfill routing might be better than central directories haha
13:53 < kanzure> nah just reviewing p2p things really.
13:54 < RangerMauve> Yeah I kinda want to experiment with p2p in the nostr ecosystem as a bridge to decentralized currencies
13:54 < kanzure> i have a few ideas for solving this problem https://fiatjaf.com/3cb7c325.html he published a "solution" here https://github.com/fiatjaf/trustlines but backtracked here https://github.com/freedomlayer/offset/issues/196#issuecomment-490911189
13:57 -!- TMM [hp@amanda.tmm.cx] has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
13:57 -!- TMM [hp@amanda.tmm.cx] has joined #hplusroadmap
14:01 < kanzure> there has been an astonishing lack of progress in mutual credit protocols.
14:01 < kanzure> partly it seems because many of the advocates of federated bookkeeping seem to actually hate money lol
14:02 < RangerMauve> It sounds hard so I'm not surprised. Trustlines is interesting
14:02 < RangerMauve> At the conf I went to recently I learned about how bitcoin bonds work and that's been eye opening
14:03 < RangerMauve> The idea that you could stake some amount of money and say "if I'm lying submit proof to the smart contract and you'll get this money" and use that as an alternative to "trust" is interesting
14:03 < L29Ah> bookkeeping sunk in state regulations ages ago, in my experience
14:03 < kanzure> yeah i am having trouble reconciling whether fiatjaf believes trustlines works or if i should believe his comment in 325.html that solving timeouts is insufficient
14:04 < kanzure> RangerMauve: ah, you mean fidelity bonds
14:05 < kanzure> "The big issue is that we don't have an independent judge to assert, for example, that D has indeed "revealed" p to C in time. C must acknowledge that voluntarily. C could do it using messages over the internet, but these messages are not reliable. C is not reliable. Clocks are not synchronized." okay well even with the bitcoin blockchain, you're not going to get the blocks all at the same ...
14:05 < kanzure> ...time, there is still clock drift.
14:05 < kanzure> block relay is actually pretty fast on the bitcoin network in general, but it's not instantaneous
14:05 < L29Ah> most people i know who do it only do it because it is required by law and strive to delegate it as much as they can; federated bookkeeping is something from the parallel world where tax authorities don't dominate the corporate landscape
14:06 < kanzure> L29Ah: the whole field of accountancy looks like a joke anyway (dollars are not constant value and they use dollars for some reason)
14:07 < L29Ah> ledger is multicurrency and currency-agnostic, one defines them ad hoc when sees fit
14:09 < kanzure> RangerMauve: for the problem of decentralized commit one possible solution might be FROST/MuSig2 style threshold multi-party multi-signatures. so the commit happens only if everyone collaboratively constructs the signature all together. it can stall and fail, but not in a way where one party is on the hook and the other isn't.
14:09 < kanzure> i have not analyzed this thoroughly at the moment so nobody should hold me to that being a good/actual solution.
14:10 < RangerMauve> Yeah I don't think I know the problem space enough. I've had to do threshold multi party agreements on causality though!
14:11 < RangerMauve> https://consento.org/docs/group/
14:11 < RangerMauve> Didn't end up getting deployed anywhere
14:11 < kanzure> i have another possible solution that avoids decentralized commit (not ready to share details); the MuSig2 style approach could be interesting although complex.
14:12 < RangerMauve> We used append only logs to keep things simple
14:13 < kanzure> that consento page does not seem to be informed about social/collaborative key recovery schemes
14:13 < RangerMauve> What do you mean?
14:14 < RangerMauve> We knew of shamir secret shariing for example
14:14 < kanzure> the goal is key recovery, right? at minimum there is shamir secret sharing 
14:14 < kanzure> https://developer.blockchaincommons.com/csr/
14:14 < RangerMauve> In this case the goal was consensus on group membership changes over time and on top of that consensus on order of operations
14:15 < kanzure> or if you distrust electrical computers https://secretcodex32.com/
14:16 < kanzure> there are many consensus algorithms available if you mean a BFT or paxos thing
14:17 < RangerMauve> Those are for leader election, no?
14:17 < RangerMauve> This is leaderless
14:17 < RangerMauve> Hmm nvm misremembering
14:18 < RangerMauve> 🤷 Eh, the project ended up falling apart anyway
14:18 < RangerMauve> G2G read through logs and determine some weird discovery issues, TTYL~
14:42 -!- gl00ten [~gl00ten@88.214.180.228] has joined #hplusroadmap
15:51 -!- gl00ten [~gl00ten@88.214.180.228] has quit [Ping timeout: 252 seconds]
16:27 -!- darsie [~darsie@84-113-82-174.cable.dynamic.surfer.at] has quit [Ping timeout: 276 seconds]
17:16 -!- stipa_ [~stipa@user/stipa] has joined #hplusroadmap
17:19 -!- stipa [~stipa@user/stipa] has quit [Ping timeout: 252 seconds]
17:19 -!- stipa_ is now known as stipa
17:27 -!- gl00ten [~gl00ten@2001:8a0:7ee5:7800:46d9:f5c:17a2:432] has joined #hplusroadmap
17:27 -!- gl00ten [~gl00ten@2001:8a0:7ee5:7800:46d9:f5c:17a2:432] has quit [Remote host closed the connection]
17:27 -!- gl00ten [~gl00ten@2001:8a0:7ee5:7800:46d9:f5c:17a2:432] has joined #hplusroadmap
18:23 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has quit [Quit: Gateway shutdown]
18:27 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has joined #hplusroadmap
19:03 < fenn> "By law, all food additives must be listed on the label. In 1940, a Federal ruling prohibited listing the vitamin content of alcoholic beverages on the label as this would imply that drinking alcohol is healthy — an improper inducement; thus, added vitamins cannot be listed on the label. However, this would violate the first law concerning food additives in general : Therefore, vitamins cannot
19:04 < fenn> be added to alcoholic beverages."
19:04 < fenn> and so we still have alcoholics developing severe neurological disorders despite it being easily preventably with the addition of thiamine
19:04 < fenn> preventable*
19:52 < fenn> "(dollars are not constant value and they use dollars for some reason)" if you have a better estimate of value i'm all ears
19:54 < fenn> i once got confused why we don't use an estimate of absolute value and after a long time i realized there is no such thing; value is always relative. it's like why there is no absolute time or frame of reference; clocks tick at a different speed depending on your speed and local gravity
19:56 < fenn> your value for something is what you'd be willing to trade for it. it only get complicated because you can trade something you don't really value for something you do value
19:56 < fenn> so then you have to think about how other people value the thing
20:05 < L29Ah> it depends on what you need this estimation for
20:09 < L29Ah> pushing an USD estimation everywhere where a number and a type of asset is enough can be counterproductive
20:14 -!- TMM [hp@amanda.tmm.cx] has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
20:14 -!- TMM [hp@amanda.tmm.cx] has joined #hplusroadmap
20:29 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has quit [Remote host closed the connection]
20:29 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has joined #hplusroadmap
21:40 < hprmbridge> kanzure> scroll to "money is the language of accountancy" https://beyondmoney.net/resource-links/e-c-riegels-valun-mutual-money-plan/
21:41 < hprmbridge> kanzure> "an unstable unit plagues the accountant with a confusion of tongues. This year’s statement is written in a tongue different from last year’s and perhaps even last month’s. Figures are not merely black and red; they are also gray and pink. Taxes are impossible of estimation because when the government runs a deficit there is a hidden tax that manifests itself in inflation. "
21:43 < fenn> better hoard chocolate bars
21:44 < hprmbridge> kanzure> nah someone will simply attack you by hoarding cocoa and distorting your books
21:44 < fenn> they will ban chocolate except for its artistic value
21:46 < hprmbridge> nmz787> IDK what context is but oh no my chocolate!
21:46 < hprmbridge> kanzure> all accountancy and books should be grounded in physics and net energy or material or something
21:46 < hprmbridge> kanzure> track every electron. no stray electrons.
21:47 < hprmbridge> kanzure> conservation of mass seems like a reasonable accounting principle to me
21:50 < fenn> i'm pretty sure this valun concept is retarded but i'm too lazy and apathetic to really read and think about it
21:51 < hprmbridge> kanzure> oh, yeah don't bother
21:52 < hprmbridge> kanzure> underneath there's a group insurance scheme, and the participants have to submit their books as evidence of their productivity or something, and other weird stuff
21:52 < fenn> tracking all depenencies in the economy is not a bad idea, but it is a lot of work currently
21:53 < hprmbridge> kanzure> when I looked, it did not seem to be adversarially sound
21:53 < fenn> the "you can only trade dollars for valuns, not valuns for dollars" was what got me
21:55 < fenn> was all the 2009 open money stuff a response to the recession?
21:56 < fenn> it seems like everything post-bitcoin has a very different quality of thinking, less theoretical and more adversarial
21:56 < hprmbridge> kanzure> not all of it, but yes a lot of attention got dumped in their direction
21:57 < hprmbridge> kanzure> bitcoin from day one outclassed all of it, merely by having working code
21:57 < fenn> proof of concept is a quality all its own
21:57 < hprmbridge> kanzure> and i was sick of all the open money garbage.....
22:05 < hprmbridge> kanzure> what is bram gonna do about chia market action. it looks like a steady decline.
22:05 < hprmbridge> kanzure> he tells me, a few minutes ago, that he has been working on a p2p poker protocol over chia, and wants to switch to atomic swaps next. although I thought chia already had swaps..
22:08 -!- darsie [~darsie@84-113-82-174.cable.dynamic.surfer.at] has joined #hplusroadmap
22:11 < fenn> i'm just glad chia failed and didn't spawn armies of hard drive consuming zombie AIs
22:13 < hprmbridge> kanzure> hey shouldn't hard drive manufacturers adopt his proof of space to advertise their bleeding edge storage capacities? instead of just marketing take our word for it btw you can't buy it yet she goes to another school?
22:14 < hprmbridge> kanzure> oh you can network it across a farm. so no way to tell it's a real single device unless its capacity is orders of magnitude larger than a farm could otherwise offer.
22:14 < fenn> if they sell a device that doesn't exist they can be sued for fraud
22:14 < fenn> otherwise, shame on you for buying into tech hype
22:14 < hprmbridge> kanzure> marketing is not the same as selling
22:15 < hprmbridge> kanzure> okay my brain juice is fried and is not computing right. oh well  farewell.
22:16 < fenn> tangentially related https://www.bahjeez.com/the-great-microsd-card-survey/
22:16 < fenn> sleep tight, don't let the brain bugs bite
--- Log closed Wed Jul 23 00:00:49 2025