Brave Adds Native IPFS Support

  • 🔧 Actively working on site again.

Never Scored

True & Honest Fan
kiwifarms.net
Joined
Aug 21, 2019
The newest version of Brave has apparently added native InterPlanetary File System support.

I will admit, when I first read the headline I didn't know what in the fuck IPFS was. After doing a little reading it seems it's a peer-to-peer protocol that can be used to distribute static content including webpages. I don't know a lot about the project. The first concern with any kind of peer to peer deal like this is obviously child porn, but they say a node can just share the content it is interested in while everyone share an index of who is sharing what. I know a lot of people here are concerned about censorship on the internet. If other browsers pick up on this is this a potential game changer, or is this just some fucking bullshit?

IPFS Homepage
 
I'm by no means versed on it as I just learned of it today as well. It appears to be like a more seamless version of torrenting. While it doesn't look like it's explicitly an 'anti-censorship' mechanism, it could help smaller persecuted sites serve a larger client base since classic server scaling (like AWS) is denied to them.

Neat I guess, we'll see.
 
The newest version of Brave has apparently added native InterPlanetary File System support.

I will admit, when I first read the headline I didn't know what in the fuck IPFS was. After doing a little reading it seems it's a peer-to-peer protocol that can be used to distribute static content including webpages. I don't know a lot about the project. The first concern with any kind of peer to peer deal like this is obviously child porn, but they say a node can just share the content it is interested in while everyone share an index of who is sharing what. I know a lot of people here are concerned about censorship on the internet. If other browsers pick up on this is this a potential game changer, or is this just some fucking bullshit?

IPFS Homepage
To understand this, it's best to start with a comparison to another 'distributed file store'- FreeNet.

Both act as stores of files, which are eventually located through hashes that match to the file's content, but if you are running a site that will have its content updated, you can have a more permanent address that can be updated to refer to the new hash addresses that describe the new content.

You can also have a domain name that points at a IPFS hash, which is pretty cool. It would be nice if someone would provide a 'dynamic DNS' type service for that.

Differences:
  • IPFS generally involves direct connections between nodes to exchange data (less privacy, more speed).
  • IPFS does not usually store content that you yourself do not request (and even stuff you do request will eventually stop being shared unless you specifically 'pin' it, unless it's accessed often). That's good if you get caught up in some police raid and weren't accessing child porn.. there won't be child porn on your PC as there almost certainly would be if you used FreeNet. It's not good if you're raided for sharing bombmaking plans on IPFS, and you get raided and their presence on your PC proves you shared them.
I've tinkered with IPFS for a bit, it's a little clunky. There's also not much reason to use it, but that's a separate issue.
I somewhat agree, but this starts to address issues that hinder adoption. Now if only someone would write a tutorial that actually explains how fucking IPNS works.
 
I've tinkered with IPFS for a bit, it's a little clunky. There's also not much reason to use it, but that's a separate issue.
What I’m wondering is when P2P standards of this type will be solid and full-featured enough that they can be used to host dynamic content. Web forums and wikis with databases, etc. It seems like they’re mostly geared towards serving static pages.
 
What I’m wondering is when P2P standards of this type will be solid and full-featured enough that they can be used to host dynamic content. Web forums and wikis with databases, etc. It seems like they’re mostly geared towards serving static pages.
Yeah so far the only real-world usage of IPFS has been as file hosting for libgen.is content.
 
  • Informative
Reactions: Drain Todger
What I’m wondering is when P2P standards of this type will be solid and full-featured enough that they can be used to host dynamic content. Web forums and wikis with databases, etc. It seems like they’re mostly geared towards serving static pages.
I don't think anyone's yet come up with a good way of maintaining shared state in a distributed system that's open to the public. You can store your static data in IPFS or wherever, but it seems like the only 3 options for your dynamic "index" of content are:
- Just accept the single point of failure and have a centralized database server somewhere (or multiple copies with some system of replication)
- Set up a distributed application on a blockchain and come up with some way not to be 51% attacked on the chain you use.
- Embrace the anarchy and let everyone be their own curator of the IPFS jungle, similar to Freenet/FMS.
 
https://ipfs.io/ipfs/bafykbzacedxcwpn2khjrnitotxpb5qkz6uwloyoxlzrjmkqz437v6jadcrnuo
How much you want to get these go-tards have never written in real languages?
https://github.com/ipfs/ipfs/blob/master/README.md said:
 
  • Horrifying
Reactions: moocow
What I’m wondering is when P2P standards of this type will be solid and full-featured enough that they can be used to host dynamic content. Web forums and wikis with databases, etc. It seems like they’re mostly geared towards serving static pages.
There's a lot of time complexity involved with rapid updating and deletion. Think about how much computation is required for rapid insertion, search and deletion; it would be extremely costly so cheap dynamic hosting would be a while away I think.
 
Last edited:
  • Agree
Reactions: Drain Todger
I don't think anyone's yet come up with a good way of maintaining shared state in a distributed system that's open to the public. You can store your static data in IPFS or wherever, but it seems like the only 3 options for your dynamic "index" of content are:
- Just accept the single point of failure and have a centralized database server somewhere (or multiple copies with some system of replication)
- Set up a distributed application on a blockchain and come up with some way not to be 51% attacked on the chain you use.
- Embrace the anarchy and let everyone be their own curator of the IPFS jungle, similar to Freenet/FMS.

There's a lot of time complexity involved with rapid updating and deletion. Think about how much computation needs to be done for rapid insertion, search and deletion; it would be extremely costly so cheap dynamic hosting would be a while away I think.
It seems like they're working on decentralized database tech for IPFS and other decentralized web platforms, actually.



None of the stuff seems mature or high-performance enough to use for serious web apps yet, though.
 
It seems like they're working on decentralized database tech for IPFS and other decentralized web platforms, actually.



None of the stuff seems mature or high-performance enough to use for serious web apps yet, though.
This looks really good. I'd still expect things to be quite costly if you plan on doing heavy amounts of read/write operations for the meanwhile. Probably needs to implement some sort of caching system so that frequently used content gets stored on the best performing servers. That's far beyond my current comprehension or any of my load-balancing experience.
 
  • Agree
Reactions: Drain Todger
Back