>>3239 You might pick up some ideas here, OP. >>>/robowaifu/237

>>3240 uh...ok thats lot. thx but in looking for digital ones not robotics

>>3233 Update: I removed inc from inc.download_file(f.path) on live 307 in lynxchan.py, this seemed to allow for files to be properly downloaded now, and one whole thread downloaded before getting this error again: "AttributeError: 'NoneType' object has no attribute 'cssselect'" Or check the screenshot

>>3233 Hopefully Robi will post here (or somewhere else) and explain exactly what his code is doing. I don't into Python, but I wrote a C++-based scraper that has been working for about a year and a half now. If I knew what his stuff was doing exactly, then I could probably relate that to my work and explain exactly how I'm doing it. Maybe it would help out some.

>>3235 Post your scraper somewhere?

>>3236 Sure, OK. (It's been readily available all along, actually). I just recently made a version that supports operating on a Raspberry Pi too. BUMP currently supports about 6 different types of imageboard software, and also kind of unifies the archives into a generalized, filesystem-based structure after download. So, it's kind of both a multi-scraper, as well as a generalization of the data storage that should enable easy script-based exporting to any type of other system -- no intermediate database required. >>>/robowaifu/8772

What is this 2009

>>3202 Just host your website as a Tor hidden service. It's free to do so and you don't need to open your ports. I'm not sure if you can avoid DDoS attacks, but it may be the case anyways.

>>3203 This. Somehow in 2020 there's still this mentality of "Tor is only for cp and glowies", like this shit doesn't happen in clearnet. Today's clearnet is pozzed to all hell in terms of censorship and the effort required to host anything safely is stupid. Honestly anyone who gives somewhat of a fuck should host/provide a mirror over Tor.

>>3201 Neocities only requires captcha to sign up, but I agree it's frustrating. I do agree people need their own sites, but there is no way in hell I'm signing up on every site just so I can talk to someone. I'm not bored of anonymous imageboards, I actually prefer them over all other types of social media. I just think we need some kind of federated imageboards system. Then anybody can self-host an imageboard with ease, everybody can setup an imageboard alongside their website. I think it would be very nice. It would have the benefit of federation too. This would also make it easier to keep track of posts, bookmarks, watched threads, etc. without needing to compromise on the anonymity aspect (because of a shared protocol). >>3202 I agree with this, buying a domain requiring personal information is cucked, as is renting a vps. >>3203 Suggests a Tor hidden service, I haven't looked into this, but would this not mean your website is inaccessible to clearnet users? Which is not ideal imo. If I want to host a website to give to normies, I'd host it on clearnet and the tor frontend would come second. For non irl stuff, that's fine to only have the Tor one. I think Tor is a good concept, I haven't used much though. I think the speed is an issue (although I understand why).

>>3215 >I just think we need some kind of federated imageboards system. I agree. Robi had the idea of The Final Solution which seemed vaguely to expand the basic idea of the webring further. My idea is basically to use the Kademlia (DHT) system common in BitTorrent to route users to each other anonymously, each hosting their own .onion service, and sharing a common imageboard directory arrangement across the entire network with each. They could each pull down 'live' board's content, manage it as they like (ie, moderate it themselves), and then publish their own edited version of an amalgam of all /tech/-related boards say, and anyone who cared to could subscribe and pull that user's moderated version of a "mega-/tech/" down onto their own machine as well. Not only would this be a federated system, but also very close to being an automatic, persistent, and durable archiving system as well into the bargain. If Robi ever gets his ass back here, I'd be happy to work with him to get The Final Solution going, and also to offer my own idea's implementation as well.

>>3215 >Suggests a Tor hidden service, I haven't looked into this, but would this not mean your website is inaccessible to clearnet users? I only suggested running an onion site since it is a realistic way for a person with zero resources to run a website. The larger your audience, the more it's going to cost to run a service for them. If you need clearnet people to access an onion site, there are free web proxies that MITM an onion site to allow clearnet users to access it. With a quick search, I found this example: https://onion.pet/ . Onion sites are actually the biggest thing I like about tor. These days everyone uses huge monolithic services, but onion sites allow anyone to host a site from their own bedroom. At least in theory. In practice there are very few onion sites worth visiting. >Secondly I don't entirely understand the anonymity aspect. Tor BROWSER suggests to not modify anything, keep it out of full screen, etc. I understand this, because you don't stand out. But then there's people who wrap their entire systems in the Tor network, or OS's that do it like Tails (I think it was tails? One of the distros does by default). How does this not massively make someone stand out? I actually don't know how the tor client works exactly, but I would hope that every program that uses tor would get separate paths through the tor network. If a same path is used by different programs connecting to the same destination, then indeed that would be bad for anonymity. However, TorBrowser doesn't actually use the system's tor client, but it comes with its own tor client. As a result, this shouldn't be a problem, if we ignore timing attacks. The approach I take to anonymity is that I see it as a gradient rather than as a binary value. Even if I'm not getting good anonymity every time I go online, it's still better than never being anonymous. If your threat model is different, for example, if you're trying to build your own wikileaks, then I wouldn't trust Tor to be anonymous enough for that. The reality is that I'm running a bloated OS written in an unsafe language dating back to 1970s, on hardware that has NSA backdoors, running not so great implementations of bloated protocols and standards. I cannot really expect perfect privacy, security, or anonymity, while still using the internet. If I can just make the enemy's life more difficult, then I have succeeded.

>>1054 You too Anon.

>>3146 If you read the thread, that's the reason why I asked. Some things can be standardized, not all. "General" standards are the worst of all. http://harmful.cat-v.org/standards/ Most websites should be rewritten into real programs.

>>3086 Living example (with the exception being that it only allows phoneposting): https://wilchan.org/b/

>>3198 >cuckflare no thanks you!!

>>3198 >wilchuj it's not only bad, it's also full of dumb polish femboys

>>3205 >full yes >polish yes >femboys yes Where is the bad part?

>>3134 heh, welcome to the internet troonfriend.

>>3134 no, not dead. almost dead. the whole project was starving because of lack of maintainers. before the shutdown pretty much the only person doing anything with the repo was dstftw and he only bothered to show up like once a month to merge a few critical bugfixes. other stuff like fixes to less important extractors, improvements and new extractors were brushed aside and ended up hanging on the pr's list for eternity (iirc there was something like 700-ish pr's in september). i tried adding a new extractor some time ago and even though it got reviewed by someone, it just hanged there until the dlc fork came along and finally merged it. so yeah, not exactly dead, more like a bleeding corpse.

>>3143 derp, i meant >>3140

install python open terminal pip install youtube-dl or pip3 install youtube-dl

it's back https://github.blog/2020-11-16-standing-up-for-developers-youtube-dl-is-back/

>>3151 I did, only 480p .m3u8s keep popping up, which is weird.

when i set my hostname on my linux machine, shouldn't other systems be able to just type my hostname and connect to it? for example if i set my hostname to "box" shouldnt i be able to type on a computer on the same network "ping box" and get responses from "box"? i used to be able to do this with my file server but one day hostnames just stop resolving.

richard stallman uses jitsi im pretty sure

>>3151 Not under the media tag? >>3153 RTFM: hostname, dns, domain name

Are the current Wyzecam V2's able to use cfw?

What are your (creative) solution for dynamic ip?

>>3081 Create an overlay network where the current address is bound to a public key, so that when the physical address changes all it takes is to change the associated value and since the overlay address is a crypto key I can encrypt data for free, too. Not my idea, but it's cool.

>>103 i dont have anything cool to say except devuan, nice also >don't let this board get more dead then it already is than*

>>3082 What is the resolver?

>>3084 As long as some kind of table can be built, resolving the overlay address to get the physical address is easy, just store it as a value in the table with the overlay as a key. It's pretty much what also happens with routers and IP addresses. With an overlay network the already existing protocols can't really be used, but it can be worked around by using a dht or something like that.