even though I don't think you guys will get too far, I haven't forgotten about the Jow Forumsternet wiki and I have finally had a free day to work on getting the clearnet access back up. (the onion was always accessible but apparently nobody wanted to use it.)
imo you guys should focus on something wildly different than current internet and meshnets, such as using public key hashes as "IP" addresses and replacing layer 3 with something end-to-end encrypted, removing the need for layer 6 shit like SSL/TLS. yes, this would break a lot of current infrastructure and software but it's ultimately the best approach and will future-proof your network, rather than just making another hobbyist meshnet. hint: every college LUG and hackerspace has made a meshnet at one point but they know it isn't sustainable at large scale in its current state.
it's clear most people on Jow Forums don't have the faintest idea of how to begin, so bring in friends who may be interested; people majoring in CS/networking, people already in the workforce, et cetera - so they can give you real-world ideas and expectations. don't keep circlejerking on irc and Jow Forums if you want something done.
and for completeness i'll give out the other related Jow Forumsternet links so people don't forget (i don't run these): git.gternet.me/ ircs://irc.jollo.org:9999/ #gternet
also, if the gternet.me domain owner is lurking, feel free to CNAME wiki.gternet.me to mesh.gentoo.today so the naming is more consistent
Grayson Baker
bump
Christian Williams
Have a pity bump. The effort is good and there has been some great discussion in the past, but you're literally just providing a BATMAN setup script, which makes the mission statement pretty shaky.
Also conversation belongs in threads on this board, not in irc. Keep it where it belongs instead of doing the same thing every community is doing and killing itself with exclusive clubs and 0 visibility due to never being on the host board.
Contacting 3rd parties outside simply asking questions/guidance and posting the result in the corresponding venues is not needed.
OP is right that you need to do more than just a meshnet, the whole web is dogshit and its architecture needs to be changed. Sites should not tell clients what logic to run and logic should not be needed to display shit. A client should automatically generate navigation and formatting based on the site's info and metainfo, such that even with simple wget it should be reasonable to read and navigate a site with no further processing. Not to mention as OP said, inbuilt encryption, and also decentralized hosting and perhaps server-like logic.
Isaac Wilson
>The effort is good and there has been some great discussion in the past, but you're literally just providing a BATMAN setup script, which makes the mission statement pretty shaky. agreed 100%, i kind of have this sentiment in the OP >Also conversation belongs in threads on this board, not in irc. communication is going to naturally be partitioned especially if people bring outsiders who don't use Jow Forums but still know their stuff. the real issue is the amount of shitposting and metaconversation (such as this) going on in these threads that prevent real discussion. >Contacting 3rd parties outside simply asking questions/guidance and posting the result in the corresponding venues is not needed. but you would still suggest we have third-party involvement, surely? it would be a great opportunity to hear all possible suggestions rather than just a subset from a bunch of anons.
as for your other ideas, this is why i stress putting ideas on the wiki, since we can then have ways to move forward. i don't really expect us to make a meshnet right away, and in fact, i believe we should take our time and brainstorm a bunch before jumping to conclusions. so far we don't even have that, we have people who want to see a final product, and that's what's hurting us
Julian Robinson
>we have people who want to see a final product, and that's what's hurting us Which is why we should probably start with a decentralized non-mesh network with symmetric mesh nodes at first. As the project gains in popularity, it will be possible to make more and more use of the mesh nodes. In any case, symmetry is needed to reach critical density, and a working web on a decentralized, internet-based platform is a lot easier to work out before the mesh stuff.
Third-parties can be involved the same way as any other contributor, but you shouldn't just go independent of Jow Forums entirely just so you can work with randos like the tox people basically did. It's not needed anyway as there are many high-quality devs on Jow Forums, believe it or not (in particular there are many confirmed real microsoft devs, as well as at least 1 google dev).
Brody Nguyen
>start with a decentralized non-mesh network cjdns maybe? internet is being used to bridge long-distance cjdns nodes. i know i've suggested VPNs for internet-connected nodes before, but it probably isn't the best solution since IP addresses still have to be assigned.
i have i2p and tor set up on my boxes, and i could easily add something else as long as it's stable, written in a non-meme language (c, erlang, something well-designed for servers), and people are going to contribute to and use it.
i'm already looking at non-infra alternatives to software and services, such as cgit+gitolite to replace github (it's actually really comfy and i want to write a script so other anons can easily put it on their vpses), maybe filesharing/bittorrent-style shit, IM/E-mail, et cetera.
it's all really exciting and it would be more exciting to see others contribute instead of jerking off
Nathan Hill
>cjdns maybe? As part of the stack, definitely for the dns. The data layer can just be ipfs or something similar in a comparable capacity (of course ipfs has its own problems, but it's a start. We don't need a perfect release-quality product day 1, just something good enough to toy around with and actually build stuff off of). We will need a way to get nodes to host arbitrary encrypted content which they can only decrypt if they try to access the file proper (freenet-style) to ensure data good data locality in the network and allow people to contribute without having to setup anything specific like manually uploading their own files or whatever (in particular, without having to know what's available on the network/to explicitly pin stuff/etc.). Maybe even clone and enable automatically serving arbitrary services altogether (with proper automatic local quota setups to prevent DoS), like if someone tries to access a gitolite node, a random node in the network which knows the service serves and processes relevant requests up to quota, and relevant content is synced with the data layer, allowing other nodes which know the service to process further requests even from where the user left off with the previous serving node).
Eitherway, starting with robust, fast, decentralized file sharing is priority number one, followed by static web (new or old format), then dynamic web (i.e. a distributed chan with limited/no js).
Dylan Sullivan
>decentralized file sharing agreed definitely. ipfs, bittorrent, git can all be used for this to back up code, images, media, and whatever else >static then dynamic web static web should be limited to content and the ability to request that content, while dynamic applications would probably be better off with different protocols and programs to handle each (like we have email and xmpp now, we can have an imageboard/news protocol, stuff like that)
Julian Harris
>different protocols and programs to handle each I generally agree with this also. It bypasses the complexity issues in current browsers and allow more efficient, much lighter operations with a standardized, task-specific interface. No more executing arbitrary code on user computers. Also people are used to download apps for everything now, so it remains possible to obtain the critical mass needed to make the whole network work.
Eli Stewart
i see why people want to develop webapps from a portability standpoint: it runs under any operating system with a compliant browser. but the issue is that javascript+HTML+CSS is not meant for this. we would need to recondition developers into developing for an alternate platform. maybe even allow developers to write in their own language and compile down to VM bytecode that can be released to all platforms.
Lincoln Morris
"webapp" development should be different from web service development (the former would refer to the apps on each platform which run a specific paradigm, there can be arbitrary quantity of clients for them that all talk the same protocol, irc or email style, while the other would refer to writing the necessary part to make a service available through these webapps by talking the right protocol). As such, the question of portability becomes disassociated: it's up to the webapp dev to deal with this, either by manually targetting each platform, or using UI frameworks that already enable multi-target development. And yes, surely this should always be as native as possible. For this purpose, native VM + bytecode is perfectly fine (since it's not high-performance), but fucking web tech is retarded and not adapted for sure.
Eli Nelson
>VM bytecode i found LLVM can do this: releases.llvm.org/1.6/docs/CommandGuide/html/lli.html LLVM is already in active development and this would save from reinventing too many wheels. that was kind of wordy and i'm not sure whether i'm understanding you right. but are you saying that we'll need to standardise APIs and protcols across platforms in order for everything to work as expected? if so, yeah, we need to be careful about how we develop this so that it's extensible and we can fallback for older devices that don't support everything.
David Gomez
Hello guys, totally newbie here. I always wanted to go balls deep on networks and seems like this is the project I was wanting for. Problem? I'm completely new to advanced internet stuff, so I would like to start researching the basic knowledge I need to lend a hand. Would you mind to point me out in the right direction?
PS: Being Peeps and Creems discusing here makes me feel bad being Chocolate and asking basic stuff. God damn.
Caleb Evans
I got you a present. Its an easy manual for going over the basics, all the way to some more intermediate stuff. layer01.club/HMXs.pdf
Colton Richardson
Holy shit user I have an overall look and seems really nice. Thank you so much, I'm going to start devouring it tomorrow and will try to learn every page.
Dominic Clark
>I always wanted to go balls deep on networks i'll try to upload some shit maybe, but mesh.gentoo.today/wiki/Resources#Link_dump has a bunch of links. i'll need to find the cisco resources i had saved; there's a lot of useful videos you can probably search on youtube though, and of course a bunch of books
Josiah Davis
No problem user. I don't want to overwhelm you at all, but here is a resource about security as well, which is a good thing to know. layer01.club/W3RC.pdf
Henry Collins
I will be reading in public transport, which i have 2 hours in space of 1 + 1 so I don't think I will get overwhelmed. I just need some powerwill to not drop. Thank you again.
Will be checking this lil by lil, user. Thank you too.
Holy fucking shit. I will need some time time to get and check, but I'm saving all this knowledge forever.
William Diaz
what do you guys think about RISC architectures and open hardware in general? i'd really like to build my own 100% open desktop with a reduced form factor and something like ARM/SPARC, just to use as a development machine probably.
it's clear intel/AMD CISC is beyond repair, unsuitable for auditable hardware, and not many people pay much attention to the hardware they're using. but it's really important to set open hardware precedence at the same time as promoting FOSS.
Leo Cox
I think RISC is a game changer. Hack the planet!
Isaiah Miller
would you choose ARM, SPARC, or something else? i hear SPARC is possibly the most open of all of them, and it's pretty well supported by the UNIX family of systems. i haven't done much of my own research on CPU architectures though so i'm kind of in the dark
Liam Torres
>had a free day to work on getting the clearnet access back About damn fucking time.
John Jenkins
tor access was always there you ungrateful fuck >install tor >add proxy setting >connect to wiki wow hard
