IRC channel logs
2025-11-03.log
back to list of logs
<Psudeou>Fun coincidence, I've recently thought about the internet <Psudeou>1) Physical infrastructure of the internet. The physical infrastructure of the internet <Psudeou>Who owns the internet cables? Who maintains them? Who eavesdrops on them? The answers are: corporations and governments. By this time with the age verification ID in swing it should be obvious that there really is no "free internet", all the internet is owned by corporations and governments, they have simply needed time to catch up and expand their jurisdiction to it. We can expect the future to <Psudeou>develop further in this direction. What use will TOR or encryption have when you need an ID to log into the internet? Defending internet freedom is a political battle, and in the long run, states will win because frankly the People don't have the ability to pay attention to everything at once. When they're distracted, states will swoop in and force "age verification" on them, or any euphemism <Psudeou>2) Centralization of command. As we continue on the threat of state control, I need to remind you of the often-forgotten basics of what "state" or "control" means. We live in a world of hierarchically structured organizations. In such an organization, you have one person above you giving you commands. Sometimes, you have a person below you whom you give orders to. Government institutions are <Psudeou>frequently seen as "neutral". This is after all what our schools and parents teach us ad nauseum. The police is there to protect you from criminals. Firefighters fight fires. The laws are there to protect you. But additionally, these organizations serve as the -personal property- of whomever is on the top of the organization. Corporations are beholden to shareholders, government institutions to <Psudeou>other government instiutions. And of course the spidery webs of corruption which infect these positions of power. Positions of power are sought after. Where there is centralization, there is by necessity corruption and abuse of power. <Psudeou>Control means simply force. Laws are en-forced by violence. At the end, if you refuse to play the game, it's a struggle of violence. Individuals get thrown in jail. Protestors get riot police'd. If neither side 'loses' or 'backs down', civil war breaks out. The riots in hong kong are worthy of reference <Psudeou>3) Technological control. As technology advances, power tips in favour of the 'centralized institutions', because they have ultimately control over the flow of goods, i.e. technology. Right now everybody can still buy parts from alibaba or go into an electronic shop to order things, but in 2-3 centuries it's possible that all these 'blindspots' are accounted for, and that the system of 'capitalism <Psudeou>doesn't exist anymore, rather replaced by a new form of feudalism. The centralization of money and corporations witnessed in USA will likely happen everywhere else, eventually fusing state and corporation power seamlessly. Ideas such as "entrepeneurship" will be replaced by "dynasties" to justfiy the new society. A truly globalized world, united under one country may be the worst case scenario, as <Psudeou>dissenters will have nowhere to flee. Remember that Snowden has asylum in Russia, of all possible countries. Competition between states is one reason why people should have good living standards, from the perspective of rulers. <Psudeou>4) Centralized Internet. If the internet is centralized, then very few people can decide over very many people: what they see, thus what they feel, thus what they think. Google can censor your site, if they want to. Facebook can shadowban your account, if they want to. Your state can tell facebook to hide certain groups from you. Etc. Hide embarrasment and corruption, pfizergate. Documents gone. <Psudeou>Even if these possibilities are not grasped now, they can be grasped to their full implication at a later date. It's a question of how far "People Power" is degraded and "State Power" exalted. <Psudeou>Conspiracy theorists are often mocked for paranoid assertions. However, how many of these assertions could be true? UFOs no, but any type of corruption yes. Pizzagate was not real, but the distrust behind it was true, see Epstein. <Psudeou>5) Centralized Communication. Centralized internet means centralized communication. Chat Control is then just the first step towards controlling what people may talk about, what ideas they may share and with whom they may talk. Groups or special interests may be forbidden. China has already reached this point. <Psudeou>6) Psychology of Surveillance. The mere knowledge that they are watched makes people less courageous and more apathetic. They will be afraid to join political groups or expose themselves to new ideas, fearing it may have later repercussions on their career or on their hidden 'citizen score'. The idea that someone is constantly watching what you do is possibly The most powerful weapon of the <Psudeou>central. Escaping from this paralysis will be the first step future protestors will have to overcome. Activists must learn how to get their fellow people to overcome this new affliction. <Psudeou>7) Psychological profiling. Everything you do leaves a profile. Ditto. But can a psychological profile actually be useful? Some people and companies believe that you can already tell 10 years before someone is a criminal. To be honest I can't say how useful or dangerous this information is, but it's something to keep in mind, namely that the state knows your personality and habits. <Psudeou>8) Pornography. There is no reason to believe that any state will ever cut down on CP. Or furries. Or.. whatever other weird stuff exists. Why? Because you can use it to blackmail people. <Psudeou>If any person becomes 'dissident', there will be enough record to take them down. <Psudeou>9) Police States. None of the aforementioned fears are irrational. There will come a time where some people will literally live under 1984. Arguably, north koreans/chinese already do that. But possibly the worst surveillance state in existance was: the German Democratic Republic. An extremely bloated, extremely corrupt surveillance holds the entire population in hostage with soviet 70s technology. <Psudeou>Anyways, I got off-topic. My point is this: while the internet is extremely important, we need to keep in mind that it merely exists on top of political foundations. <Psudeou>Any activism on the internet itself will be restricted by 'the rules', and if We don't make them, then They will. <Psudeou>OK, everyone's afk. Feel free to copypaste if there was any insight in that. <sshine><awkward silence after wall of text> <dthompson>uh, gonna need a tl;dr on that if there is to be any real discussion <dthompson>I want to keep this channel focused on discussions of spritely technology <kestrelwx>dthompson: Hi! I get field access of a null object on `ffi_handler` when trying to replicate what you had in the gist and the January blogpost. The code to set it seems to be only reachable from `load_extension` and `load_main`. So should I patch reflect in my repo and ship that or maybe move imperative code from top level and then `load_main`, get a reflect object and then load my entry point? <dpk>tl;dr cypherpunk believes gubberment is out to get them (and apparently everyone else) <dthompson>kestrelwx: worth a shot. sounds like the chicken-and-egg problem that I run into sometimes. need to devise a better way around it. <dpk>dthompson: in your blog post you write ‘Goblins currently uses VLists for its functional hash table needs, but HAMTs are better suited for the task.’ – what was specifically wrong with vhashes? <dpk>i’m idly considering a vhash SRFI as a functional *and* insertion-ordered counterpart to SRFI 146 functional but randomly-ordered hashmaps (HAMTs) and SRFI 250 non-functional but insertion-ordered hash tables <dpk>but i have insufficient experience with using them for real so would be interested to hear what you have to say :-) <dthompson>iirc vhashes are like alists in that if you set the value for the same key n times then there are n key/value pairs in the vhash, so there's a memory growth problem <dthompson>if insertion ordering is important then a vhash seems like a good fit <dthompson>I never really need that so HAMTs have been great for me <dpk>yeah, the double-keys problem is an issue. also the fact that performance degrades if you too often cons multiple times onto the same tail <dthompson>kestrelwx: one thing you should try is to take the return value of Scheme.load_main and access it's 'reflector' property <dthompson>say your top-level program returns a procedure: <dthompson>const [proc] = Scheme.load_main("foo.wasm"); <dthompson>you will now have access to a reflection instance via: proc.reflector <kestrelwx>My `bv.obj` is an empty object(`{}`). Wonder how that came to be. <kestrelwx>Oh, maybe it's just opaque. But I get a TypeError in `bytevector_length` . <kestrelwx>dthompson: Couldn't I access the reflector of the Bytevector itself inside the call? <dthompson>yeah iirc the object your import function is a "raw" wasm reference <dthompson>the object *in* your import function, that is <kestrelwx>Yeah, I just had to wrap the reflect into a thunk because it was undefined in the closure. <kestrelwx>So my bv constructed from the raw value was getting `undefined` reflect. <ridley>Can anyone tell me, in goblins, where the code is that determines whether or not to include an answer-pos in op:deliver? Does goblins always enable pipelining on an op:deliver or determine it call by call? I spent some time trying to read the code but feel like I'm missing something. <ridley>I may have found an interesting edge case in the spec that may warrant adjustment but am not quite sure yet. <dthompson>ridley: tsyesika would know best but is out this week so I will take a look and see <ridley>If it's too much to ask I can also wait for tsyesika <ridley>it's not blocking for me this week <dthompson>we have different data types for "regular" messages and messages related to questions <dthompson>for the <questioned> type of messages, we look up the associated answer-pos to send <ridley>oh that's what I was not understanding, thanks! <ridley>my scheme is usually good enough to understand the goblins code with enough effort but there are certainly still gaps in my understanding <ridley>I think the thing that tricked me is that the <questioned> stuff is actually defined in goblins core and I haven't delved too deep in there yet <dthompson>yeah I guess the types are defined there for the sake of defining some polymorphic procedures that extract common information from any kind of message <ridley>anyway, this gives me enough for now, thanks <ridley>Am I correct that <-np is just <- but where np stands for "no pipelining"? <ridley>seems it is "doesn't expect result" <dthompson>you send the message but you're chill about it. it's the "no problem" operator