Conversation
Notices
-
仮面の告白 (lain@testing.pleroma.lol)'s status on Sunday, 04-Mar-2018 21:55:28 UTC 仮面の告白 @augustus hail eris. - Hallå Kitteh repeated this.
-
worst girl 🏴☠️ (xj9@sunshinegardens.org)'s status on Sunday, 04-Mar-2018 23:03:33 UTC worst girl 🏴☠️ honestly, the most interesting #p2p system right now is secure scuttlebutt. the only authority is *you*.
you decide who to replicate.
you are your own CA.
you pick what network(s) you want to using public or private root keys that anyone can generate.Hallå Kitteh repeated this. -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Monday, 05-Mar-2018 04:23:25 UTC Hallå Kitteh @delores @xj9 @benis The actual question: Has anybody adapted IPFS to run on I2P? -
devores (delores@shitposter.club)'s status on Monday, 05-Mar-2018 04:38:23 UTC devores @benis @xj9 gnunet does offer data redundancy, in the sense that you can mirror someone elses files willingly. thats where freenet fails with its data redundancy (forcing users to store shit they dont want). gnunet, you just store what you mirror. thats the ideal.
i have never used gnutella, but i doubt its as active or useful as bittorrent (which im not promoting). if i cant get hundreds of thousands of perfect log/cue flacs, you've just made a shitty soulseekHallå Kitteh repeated this. -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Monday, 05-Mar-2018 05:12:25 UTC Hallå Kitteh @benis @delores I haven't used gnutella since last I used BearShare 15 years ago. How is it objectively superior?
Isn't bittorrent an improvement on gnutella and IPFS an improvement on bittorrent? I agree with @xj9 that trying to keep bittorrent compatibility in IPFS would have been better, but here we are.
Unless gnutella has had serious protocol upgrades since 2000, GNUnet is also an improvement on it. -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Monday, 05-Mar-2018 05:15:52 UTC Hallå Kitteh @ivesen @xj9 ssb is zeronet minus the namecoin plus a bunch of interesting people who want to do, and are doing, cool things with it.
Nobody is doing the #gitssb equivalent on zeronet AFAIK. The split between messages and blobs makes it possible on SSB. -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Monday, 05-Mar-2018 05:20:15 UTC Hallå Kitteh @ivesen @xj9 Ok, that's cool. So they're basically putting up a web-hosted git repo but on zeronet? That gets you part of the way, but pull requests and issues integrated with the network is really sweet. -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Monday, 05-Mar-2018 05:29:40 UTC Hallå Kitteh @benis @delores @xj9 IPFS was designed to have an upgradeable algorithm. I'm assuming bittorrent would be trickier. -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Monday, 05-Mar-2018 05:32:35 UTC Hallå Kitteh @ivesen @xj9 Yes, git is inherently decentralized and you can do pull requests over federated e-mail. Only Linus does that though, everyone else is on some hub, bucket or teacup. I would like the hubs, buckets and teacups to be distributed too, and git-ssb already is. Now it just needs more polish, userops (although that's already pretty decent) and users.
Discovery matters. Thresholds matter. -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Monday, 05-Mar-2018 05:33:54 UTC Hallå Kitteh @delores @xj9 @benis As for the rest of the post: Too incoherent, only read once. -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Monday, 05-Mar-2018 05:49:15 UTC Hallå Kitteh @benis @delores @xj9 The hash includes bits that identify which hash algorithm is used. If you want to start using SHA-512, just implement it and start the hash with 0x1340 instead of 0x1220, and then hope others will implement it or use your implementation. The existing base will not be confused, they just won't be able to parse the hash until they've been upgraded. -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Monday, 05-Mar-2018 05:51:31 UTC Hallå Kitteh @benis @delores @xj9 I don't see much difference between bittorrent and ipfs in terms of cathedraliness or bazaariness, it's just that bittorrent is older and more mature, while ipfs is still in bootstrap. -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Monday, 05-Mar-2018 06:05:18 UTC Hallå Kitteh @benis That's just the same thing with more bytes. -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Monday, 05-Mar-2018 06:09:26 UTC Hallå Kitteh @benis @delores @xj9 So to you bazaar means correct and comprehensive specification and robustness? No wonder I misunderstood you. I thought it was about project governance. -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Monday, 05-Mar-2018 06:22:30 UTC Hallå Kitteh @benis I don't see how it's any more easy or difficult to teach your software 0x14 than to teach it "urn:sha512". The urn namespace is greater, that's the difference.
Human-readable is nice. But I don't believe there will be a generic client smart enough to understand how to use SHA-512 on Gnutella. You need a Gnutella client and it needs to already understand SHA-512.
I don't see any forward compatibility in BEP 0003:
> pieces maps to a string whose length is a multiple of 20. It is to be subdivided into strings of length 20, each of which is the SHA1 hash of the piece at the corresponding index.
And I don't see any other mention of "hash" on http://www.bittorrent.org/beps/bep_0000.html , except mentions of infohash that don't seem to affect the core protocol. -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Monday, 05-Mar-2018 06:31:14 UTC Hallå Kitteh @benis @xj9 esr's text is about the governance of a project and its source code. It's about whether you develop in private and just publish tarballs or whether you publish your source control repo and liberally accept patches.
In esr's terms, basically every free software project on github is a bazaar project, unless it's for show only and they don't accept pull requests. The bazaar won before the last century was over.
If you invent your own meanings for the terms, I cannot take responsibility. -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Monday, 05-Mar-2018 06:32:46 UTC Hallå Kitteh @benis Here is the timestamp for the death of the Cathedral:
> In April of 1999, the Free Software Foundation (the official sponsors of GCC) dissolved the original GCC development group and officially handed control of the project to the the EGCS steering team.
https://www.safaribooksonline.com/library/view/the-cathedral/0596001088/apcs02.html -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Monday, 05-Mar-2018 07:06:26 UTC Hallå Kitteh @benis ASCII is not more semantic than binary. I agree that a larger namespace provides more extensibility than a smaller namespace, but I don't think multihash will run out of that byte any time soon, and I don't think the management of the semantics of that byte will have a huge ideological clash. If it does, fine, replace it with something better.
Maybe do like microsoft did and replaced their fourCC with a GUID. By the time we need 512-bit hashes I guess a 128-bit identifier isn't quite as insane, but even at that point it still seems quite overkill. And "urn:sha512" is already 80 bits.
I'm a big fan of the http://wiki.c2.com/?PowerOfPlainText , but it only makes systems more understandable for humans, not machines. And while I do think some people go on a djb-like crusade against plaintext in protocol, I don't think it's entirely unwarranted when you're planning to shove terabytes around. And I'm also a big fan of http://wiki.c2.com/?YouAintGonnaNeedIt . -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Monday, 05-Mar-2018 07:15:55 UTC Hallå Kitteh @benis
My, those certainly are a lot of BEPs. And they all hardcode sha1 and 20-byte infohash except the one that doesn't mention hashes at all.
I get it, you can make propose to enhance bittorrent by making bittorrent enhancement proposals.
You can improve IPFS by discussing within the IPFS community too.
I was specifically replying to this statement from you:
> Like wasn't SHAttered already? Why isn't Bittorent or IPFS ftm using another algorithm!?!? Gnutella can use any agnostically. It's not rigid like the formers.
I have shown you how IPFS isn't rigid about its algorithm and you have shown me that bittorrent is. When I was "assuming bittorrent would be trickier", apparently I was right.
I don't think this discussion is going anywhere productive. Thanks for triggering me to read up on esr and bittorrent. -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Monday, 05-Mar-2018 07:16:46 UTC Hallå Kitteh If anyone except me was wondering where specifically esr (via Gerald M. Weinberg's "The Psychology of Computer Programming") got his Kropotkin quote, it's in "Memoirs of a Revolutionist", chapter 3, section VII, paragraph 6:
https://theanarchistlibrary.org/library/petr-kropotkin-memoirs-of-a-revolutionist#toc31
> Having been brought up in a serf-owner’s family, I entered active life, like all young men of my time, with a great deal of confidence in the necessity of commanding, ordering, scolding, punishing, and the like. But when, at an early stage, I had to manage serious enterprises and to deal with men, and when each mistake would lead at once to heavy consequences, I began to appreciate the difference between acting on the principle of command and discipline and acting on the principle of common understanding. The former works admirably in a military parade, but it is worth nothing where real life is concerned, and the aim can be achieved only through the severe effort of many converging wills. Although I did not then formulate my observations in terms borrowed from party struggles, I may say now that I lost in Siberia whatever faith in state discipline I had cherished before. I was prepared to become an anarchist. -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Monday, 05-Mar-2018 07:40:29 UTC Hallå Kitteh @benis I was hoping to get your technical insights on why Gnutella is superior to everything, but all I got was word salad and "it's semantic".
But it's good to know that Gnutella is available for i2p. I wasn't previously aware, so thanks for that. I will keep that in mind next time I'm looking for something.
IPFS over i2p shouldn't be very different from IPFS over TCP/IP, it's just that nobody has yet bothered. And I guess it would be pretty harshly segregated from the clearnet IPFS, but then so is the case with bittorrent.
IPFS already offers clear advantages over bittorrent: Thanks to the public gateway, I can submit #hpr episodes over IPFS, but I can't do that over bittorrent. -
本当にプルルート (benis@social.i2p.rocks)'s status on Monday, 05-Mar-2018 07:58:07 UTC 本当にプルルート @clacke
>superior to everything
Your words, not mine. I specified Gnutella was superior, for I2P, in designed, and explained why next posts.
In I2P, you want to have as many things, well your entire suite of applications running on it, since it is an overlay network.
Therefore, if you want an exercise from me (i.e. productive discussion), is to federate on I2P any OStatus|ActivePub server. @jeff already did for pleroma, and it's upstream.
Doing so, will prove to you everything I've said this entire thread about networking and testing. It doesn't have to come from my words, but your experience.
>nobody has yet bothered.
take, a serious guess why, ignoring my background.
>it would be pretty harshly segregated from the clearnet IPFS
oh, are your lightbulbs finally turning on?!? please keep mentally exploring why IPFS fails from ground up, up down, left right and in between in it's idea, and worst, it's implementations.
>IPFS already offers clear advantages over bittorrent
Please, enlighten me
>public gateway, I can submit #hpr episodes over IPFS, but I can't do that over bittorrent.
You either live on a cave, or you need to clean up your bias:
https://webtorrent.io/faq /Who is using WebTorrent today?
lit. @jeff made this in a few weeks: https://gitgud.tv
and peertube blazes IPFS many many many times.Hallå Kitteh likes this.Hallå Kitteh repeated this. -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Monday, 05-Mar-2018 08:08:14 UTC Hallå Kitteh @benis
>> nobody has yet bothered [to port IPFS to i2p]
> take, a serious guess why, ignoring my background.
I don't know your background, but my assumption has been all along that anybody who is interested in anonymous file sharing would simply use freenet instead of IPFS. IPFS optimizes for performance and in doing so ignores anonymity and deniability.
But somewhere along the line, somebody will probably find IPFS over i2p the right balance for their use case, and they will port it just like @jeff did with Pleroma. After all, it was possible with bittorrent and IPFS isn't too different.
BitTorrent not working for HPR uploads simply means the form and the process assumes an http*:// URL. WebTorrent and PeerTube have nothing to do with it. -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Monday, 05-Mar-2018 10:37:51 UTC Hallå Kitteh @benis @jeff
> You specifically ignored BEP 17 & 19
Those are about http seeding. I was talking about seeding over IPFS (or, analogously, bittorrent) and somebody kindly providing a gateway server that would allow an http client to get the file I'm seeding.
I guess I could run my own zeronet gateway, but ipfs.io is just allows me to be lazy.
> why do you expect IPFS will work I2P, when even bittorents have a hard time working, as I specified?
They work well enough for me except when they don't, and then that's just a lack of seeders.
I didn't ignore PeerTube and WebTorrent. I think they're awesome and I want them to succeed, they just don't have anything to do with the random use case I mentioned. -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Tuesday, 06-Mar-2018 03:09:06 UTC Hallå Kitteh @jeff @benis Which level is that? Either way your computer connects to a swarm and downloads a thing.
Besides, js-ipfs is already a thing, and it supports WebRTC. -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Tuesday, 06-Mar-2018 03:16:10 UTC Hallå Kitteh @benis @jeff I'm looking at gitgud.tv and I'm seeing a frontend that talks WebTorrent. It's very impressive with live streaming and stuff, but I don't see download links for dumb clients like curl that only speak http, which was specifically what I was talking about. -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Tuesday, 06-Mar-2018 03:29:19 UTC Hallå Kitteh @benis What are the differences between WaveGuide and PeerTube? Is WaveGuide more focused on live streaming? -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Tuesday, 06-Mar-2018 03:32:22 UTC Hallå Kitteh @benis @jeff ... or is the CODE_OF_CONDUCT.md the most important difference? -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Tuesday, 06-Mar-2018 16:25:27 UTC Hallå Kitteh @jeff Glad to hear a concrete data point!
Did you find it worth looking at where the bottleneck was, or was it just a case of screw this, let's use what works now? -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Tuesday, 06-Mar-2018 16:26:23 UTC Hallå Kitteh @jeff Aha, that's cool. So the site does in fact host all the content? Or are the webseeds elsewhere? -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Tuesday, 06-Mar-2018 16:26:39 UTC Hallå Kitteh @jeff Gotcha. -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Tuesday, 06-Mar-2018 16:32:48 UTC Hallå Kitteh @jeff But if the original uploader stays on the swarm, the host won't have to do much, right? -
Hallå Kitteh (clacke@social.heldscal.la)'s status on Tuesday, 06-Mar-2018 16:34:14 UTC Hallå Kitteh @jeff Right, ok.