Radicle: A peer-to-peer alternative to GitHub • Hacker News Board

Re: Radicle: A peer-to-peer alternative to GitHub

#161

Post by simias » Sat, Dec 05, 2020, 4:35 PM UTC

htrap wrote: post #134

simias wrote: post #128

Earlier quoted context omitted.

I'm not sure I see the point of building a complex platform with governance and whatnot. What I want is an easy way to publish my projects and for other people to contribute, I need something more like email (independent, self-hosted servers with a well defined protocol to communicate between them) than Facebook. I don't care about the governance of my git remote, or getting money out of it. I care about it being rel…

I feel that there is a real need for permanent Storage with respect to Open Source. Code breaks when old packages are unpublished or repositories deleted. Push once and fetch forever solves this. Also Centralized solutions are providing open source collaboration tools for free, storage for free, because of their revenue from enterprise customers. What happens when they decide to shut down? or change their policies? o…

With git it's rare for a project that's actually in use to go completely memory-holed, every contributor effectively having a local copy of the resource.

Using git (generally github) repositories for dependency management is, IMO, a hack and so it's not surprising that it often breaks. I like the way buildroot handles it (I'm sure they're not the only ones, but that's the one project I'm most familiar with):

- The buildroot buildbot fetches third party packages dependencies and archive them.

- When you build your buildroot image locally, it attempts to fetch from the third party directly. If the file doesn't exist anymore, it falls back onto the buildroot cache instead.

You could also easily add your own caching layer in there if you wanted too. I think that's distributed computing at its best: simple and robust, with a clear and easily understandable architecture. No blockchain-based proof-of-stake distributed storage, just a series of wget. And of course since everything is authenticated with a strong hash it's always perfectly safe.

Re: Radicle: A peer-to-peer alternative to GitHub

#162

Post by vicek22 » Sat, Dec 05, 2020, 4:42 PM UTC

simias wrote: post #128

htrap wrote: post #106

We are building https://gitopia.org - Permanent Data Storage provided by Arweave - Works from within git with the help of git-remote-helper `npm install -g git-remote-gitopia` so no need to learn new tooling - Built-in incentivization to token holders who also take part in the governance of Gitopia - Token holders share revenue made by the platform - You can mirror your GitHub repositories now using the Github Mirror…

I'm not sure I see the point of building a complex platform with governance and whatnot. What I want is an easy way to publish my projects and for other people to contribute, I need something more like email (independent, self-hosted servers with a well defined protocol to communicate between them) than Facebook. I don't care about the governance of my git remote, or getting money out of it. I care about it being rel…

You might be interested in https://github.com/MichaelMure/git-bug. It stores issues directly in the git repository so the whole issue tracker is as distributed as the rest of your code. I can't speak to the UX though.

Re: Radicle: A peer-to-peer alternative to GitHub

#163

Post by adkadskhj » Sat, Dec 05, 2020, 4:48 PM UTC

StavrosK wrote: post #159

simias wrote: post #128

Earlier quoted context omitted.

I'm not sure I see the point of building a complex platform with governance and whatnot. What I want is an easy way to publish my projects and for other people to contribute, I need something more like email (independent, self-hosted servers with a well defined protocol to communicate between them) than Facebook. I don't care about the governance of my git remote, or getting money out of it. I care about it being rel…

How about hosting your own Gitea instance? It works quite well.

I think that works - but what GP mentioned is where it fails:

> Frankly I would be perfectly fine with the current situation where you have a bunch of effectively centralized code hosting solutions (github, gitlab, bitbucket etc...) if you could trivially move your project from one to an other. For the code it's easy, git is built that way. For issues, PRs and the like it's trickier.

Maybe there's a nice way to Gitea distribute this information among Gitea instances, though it doesn't seem advertised as such - as this would basically make Gitea some neat Federated tool i think.

Ultimately i'd like Git[ea|whatever] to be just a frontend for a database that behaves like Git. In the way same way that Github's Source Viewing is just a frontend for Git. You don't worry about moving your Source data between Github and Gitlab, so why are we worrying about universal data like Bug tickets, Feature tracking, etc.

I'm a massive fan of systems that behave like Git and Scuttlebutt. Which is to say, they're dumb - simple. Git can be pushed and pulled from basically everything. There's no complex suite of nodes around the world that are expected or assumed to operate for any Git functionality. In the same way Scuttlebutt - which offers P2P layers, is similarly dumb (though less so, unfortunately) when compared to more complex P2P offerings like IPFS.

In my ideal world we'd have a database to pair with Git that would have some very basic schemas to complement Git. Possibly even baked into Git. Such that when you move from Github to Gitlab to Gitea - everything truly essential comes with. Some things might change, like your CI if it was bound to Github - but still. Losing some things vs losing everything.

I personally am less concerned about using Github/etc. Ie i'm not dying for someone to give me a new Github. I'm seeking a way to reduce lockin.

NOTE: I'm working on a distributed database that fits my needs on my above design goals. Really it's just Git + some additional data structures which allow for more data types being stored, such as binary and structured data to build foundations for SQL layers and etc. It's not intended for general use, but i'd love to see someone pick up the idea and run with it. "Git for Data" has been done a couple times, NomsDB and DoltDB namely, but they still felt like they weren't Git-like in that they wanted to centralize - probably for SaaS reasons.

Re: Radicle: A peer-to-peer alternative to GitHub

#164

Post by djsumdog » Sat, Dec 05, 2020, 4:49 PM UTC

xipho wrote: post #151

simias wrote: post #128

Earlier quoted context omitted.

I'm not sure I see the point of building a complex platform with governance and whatnot. What I want is an easy way to publish my projects and for other people to contribute, I need something more like email (independent, self-hosted servers with a well defined protocol to communicate between them) than Facebook. I don't care about the governance of my git remote, or getting money out of it. I care about it being rel…

> I need something more like email (independent, self-hosted servers with a well defined protocol to communicate between them) The problem with email servers is you need a special type of sys admin to maintain them properly, they are not for the light at heart when used anywhere beyond the most trivial case. Any server that grows to some well-used size has a myriad of problems (getting outright blacklisted, etc).

Well, that's e-mail. Modern federated systems don't have a different server for sending and one for receiving. ActivyPub implementations are usually a single application with a database (like Pleroma) or a single app plus db/reddis/elasticsearch(optional) like Mastodon. Mastodon has official Docker containers, and it's not difficult to build one for Pleroma.

So you can make something "like e-mail" that isn't as bad as SMTP/IMAP/SPF/DKIM/etc... I've been considering hosting my own Gogs or Gitlab or one of the other locally hosted git platforms. I'd like to see something that allows pull/merge requests between them (you'd need some spam prevention of course; maybe require a message and a follow before people are allowed to push an request to your server).

This project ... doesn't seem like it does that at all. It's a desktop application .. with no real web view into your projects. I feel like it's missing a component, a service run in a docker container that you can program with your Device ID and push your public repos to for others to see.

Re: Radicle: A peer-to-peer alternative to GitHub

#165

Post by theferalrobot » Sat, Dec 05, 2020, 4:54 PM UTC

nikitaga wrote: post #38

What a nice funky design. That front page has a soul, unlike almost everything else I see nowadays.

Totally agree, design today takes itself way too seriously to the point that it all looks like it came from the same corporate hell hole. This is a breath of fresh air.

Re: Radicle: A peer-to-peer alternative to GitHub

#166

Post by JJJollyjim » Sat, Dec 05, 2020, 4:55 PM UTC

b_fiive wrote: post #110

Aeolun wrote: post #78

How and why is literally everything peer to peer in the tech community also somehow connected to ethereum?

quick hand raise for the IPFS community, which is yes p2p, no blockchain. I'm working a "git for data" project atop IPFS and aren't hawking a coin. I'll readily admit there are very few yes-p2p and no-blockchain projects out there, but we exist

In fairness, filecoin is linked on the IPFS homepage (though it is less prominent than I remember)

Re: Radicle: A peer-to-peer alternative to GitHub

#167

Post by 2bitencryption » Sat, Dec 05, 2020, 4:57 PM UTC

viraptor wrote: post #39

nikitaga wrote: post #38

What a nice funky design. That front page has a soul, unlike almost everything else I see nowadays.

There's a post about the design here: https://radicle.community/t/the-journey-to-radicles-new-bran...

this is just me, but I'm honestly way more interested in this design post than I am about the product itself.

love the design - it's honestly something I've never seen before (at least, not in the context of software). I love how it rejects the tired old aesthetic of the "software tool front page" and goes in a totally different direction while still getting the point across.

Re: Radicle: A peer-to-peer alternative to GitHub

#168

Post by olah_1 » Sat, Dec 05, 2020, 5:01 PM UTC

jasonpeacock wrote: post #145

olah_1 wrote: post #143

Earlier quoted context omitted.

> - Permanent Data Storage provided by Arweave This kind of thing always makes me sweat. Are deletes at all possible? What if someone accidentally pushes keys to the repo? On regular github, at least you can nuke the whole repo and start over if need be.

Those keys absolutely need to be rotated, regardless of whether you delete the commit or repo after accidentally pushing them. At which point you may as well leave them there...

I agree, but is it possible to delete at all on ARWeave? Let’s say someone accidentally puts their full name and address in a repo.

Re: Radicle: A peer-to-peer alternative to GitHub

#169

Post by labawi » Sat, Dec 05, 2020, 5:02 PM UTC

jasonpeacock wrote: post #145

olah_1 wrote: post #143

Earlier quoted context omitted.

> - Permanent Data Storage provided by Arweave This kind of thing always makes me sweat. Are deletes at all possible? What if someone accidentally pushes keys to the repo? On regular github, at least you can nuke the whole repo and start over if need be.

Those keys absolutely need to be rotated, regardless of whether you delete the commit or repo after accidentally pushing them. At which point you may as well leave them there...

Keys probably aren't the best example. What about PII, accidentally committed "test" data, documents, nudes, whatever..

Re: Radicle: A peer-to-peer alternative to GitHub

#170

Post by whichquestion » Sat, Dec 05, 2020, 5:06 PM UTC

simias wrote: post #128

htrap wrote: post #106

We are building https://gitopia.org - Permanent Data Storage provided by Arweave - Works from within git with the help of git-remote-helper `npm install -g git-remote-gitopia` so no need to learn new tooling - Built-in incentivization to token holders who also take part in the governance of Gitopia - Token holders share revenue made by the platform - You can mirror your GitHub repositories now using the Github Mirror…

I'm not sure I see the point of building a complex platform with governance and whatnot. What I want is an easy way to publish my projects and for other people to contribute, I need something more like email (independent, self-hosted servers with a well defined protocol to communicate between them) than Facebook. I don't care about the governance of my git remote, or getting money out of it. I care about it being rel…

The best solution from my perspective for the community would be to have a standard that handles issues/prs and the like that could be taken from one code hosting solution to another. Just like how you can take your code with ease from one solution to the next by cloning.

It’s the same idea that we already use for our code but applied to all the other bits that are necessary for maintaining projects.

There are some pretty obvious problems with actually implementing this, however. One of which comes down to getting all the existing code hosting solutions to agree on a standard. As they could simply create ancillary standards to differentiate thenselves. Not to mention all the work involved in implementing this when most people are accepting of what we have now (until it bites them somehow).