How does Maidsafe compare to Freenet, I2P, and IPFS? From what I can gather, Freenet and I2P don’t have the distributed filesystem that Maidsafe has. I think IPFS does have something similar but it doesn’t have the Safecoin incentive system. I could be way off though. Am I close?
Are there other comparable networks?
Freenet uses a cache system as far as I know. So when you upload something, nodes pick it up in cache. If it’s not requested for some days or weeks your data might me lost.
I2P is a bit like TOR. You connect to a node, that node connects you to another node etc. You have 2 proxies before someone can reach you, and the other nodes have the same. so that’s always 4 proxies between 2 nodes connecting. It’s a server-system. So you run a server behind your 2 proxies, when you take it offline it’s gone.
Freenet is a distributed Filesystem in an abstract way. Filesystem maybe for you doesnt mean that you see it on some os as a higlevel path or drive letter on some windows system. Actually there is even a mountable fuse layer for Freenet. Freenet is the very same idea of named data network as the fundamental Maidsafe idea. You upload data into the network (be it Freenet or Maidsafe) and it gets a permanent address and the file gets chunked and split into parts and will get distributed to participating nodes. Freenet is probably the oldest implementation and actually working one that is out there.
I2P is for anonymous torrent filesharing system. Like @anon40790172 said, it can also run servers within I2P. Upload has it’s own tunnel. Download has it’s own tunnel. So it’s technically 4 tunnels per client. Speeds increase as more users uses the system.
Tor is not build for torrent filesharing system. The only purpose is communications.
What are the most significant ways in which Freenet and Maidsafe differ?
Freenet does not guarantee the existence of data. You don’t know if your data will exist or not so the possibilities of Freenet are very limited. In the end is more a exchange and communication system instead that a data management system as Maidsafe.
Got it. Does anyone know more about IPFS and how it compares to Maidsafe?
For one, IPFS has no anonymity guarantees which is a big deal IMO
IPFS is similar to MaidSAFE in that it uses a Peer to Peer system to deliver files efficiently,
IPFS lacks most of the features of SAFE… All of these are possible over IPFS - but they would be add on features - not built in features.
By default IPFS is just a scheme to name and deliver your files. The network does not store your files unless somebody intentionally chooses to “pin” it. I believe they intend to do a “pay for pin” kind of scheme, but that is an add on feature not a built in feature.
IPFS also doesn’t by default self-encrypt your files. I believe it does have point to point encryption.
I believe that IPFS and SAFE share a lot of the same goals. IPFS is developing things one piece at a time in a modular function, while SAFE is trying to build the entire system all at once with all of the defaults set for “SAFE”.
IPFS is attempting to live within existing regulatory and legal frameworks (by supporting features blacklists for example) while SAFE is kinda anarchistic in circumventing authority entirely with an enormous black box…
Personally I think IPFS is going to be more useful for most everyday things. While SAFE will be most useful where extreme security is needed. Right now, I can type ‘IPFS add foldername’ and my stuff is shared with whomever has the hash… That is useful, and for most things self-encryption, anonymity, SAFEcoin, etc would be overkill…
You can download and use IPFS today, while SAFE is still kinda sorta…
Can you create a website in IPFS and making it accessible to everyone who knows the name/domain? Is there an API similar to the one Maidsafe is planning for apps?
Yes. That is really simple. If you have a website in a folder and you type ‘ipfs add’ and the folder path, it hashes the folder, all the files in it and that hash becomes it’s address…
After that anyone running IPFS can look for that hash and the nearest copy will be delivered.
If the folder has an index.html or the like it will open like a webpage.
IPFS has public gateways, so if you go to https://ipfs.io/ipfs/QmSXpNKcmAdUQtEr5CPxXwdAtuy5CnDpZaWNjhJX42qz4m/ for example, even if you don’t have IPFS, it will still find the files and deliver them…
The big caveate is that one of your machines, or somebody who has run ‘ipfs pin add QmSXpNKcmAdUQtEr5CPxXwdAtuy5CnDpZaWNjhJX42qz4m’ needs to be online in order for the file to be findable… There is no automatic network storage. It is all voluntary or for pay – so that nobody accidentally stores something illegal or such…
One of the BIG competitive advantages that IPFS has is that it is very simular on the interface level to the linux filesystem or the traditional HTML system etc… It really is just a drop in replacement, and everything works like it has always worked, just from a P2P provider rather than a Server provider…
OK so IPFS does not have a distributed filesystem like Maidsafe or Freenet. It still uses servers. Does it have an API for building something more sophisticated than static content?
It really depends on your use case. If all you want is to host static websites with variable availability, IPFS is fine.
However, safe net potentially offers much more. Of course, this may be overkill for the above use case for some, but it is essentially for other use cases.
Hosting blockchain like databases, data with one owner, personal data, high availability, no censorship, fast transactions, economically sustainable (safecoin), network layer authentication, etc. Powerful, powerful features there.
To add, I think a big use case for safe net will actually be largely unnoticed. I suspect it will become a back-end store for many apps which want a simple way to load/save app data with security and high availability.
In many respects, other cloud data services can already do this, but they are centralised and corporate. Having something which is distributed, open source and easy to register for etc are strong drivers.
So, the headline apps may be pure safe net technology, but there will be many apps which interact in a smaller, but still essential way.
I agree. I think it will be the go-to back end for apps, watch your back AWS…lol!
It really isn’t “Variable availability” If you have an ISP account you type “pin” on the command line and it is available. I think there will be lots of other solutions for permanent storage.
It is no more constrained to static sites than SAFE is. It has a DNS system of it’s own, and it it is at it’s core built on the same technology of GIT… Anything that you can do with JavaScript you can do on IPFS, and you can do most anything with Javascript.
I do agree that SAFE will be better for high security stuff. You can just put it on SAFE and forget it…
I see SAFE as a big unknown, mostly because it success is dependent on it’s economic model, and many folks want to monkey with that. We have no idea what it will cost to store stuff on SAFE…
Having a “whole elephant at once” solution is somewhat risky in that if one leg is crippled, the whole elephant is crippled… IPFS is modular, and you can drop in different encryption, different economic models, different redundancies, different structured data models, on a user by user, case by case basis based on what makes sense in the situation vs what makes sense for the most paranoid on the network.
The biggest advantage is that you don’t need an additional API… It all can work in a manner quite compatible with existing web technologies. I think if SAFE is going to compete with IPFS on day to day stuff it will need to become nearly as easy to use, from both the front end and the back end… IPFS already has a pretty big head start and is being implemented in all kinds of Ethereum projects as well as being included in products like FreeNAS etc…
I think this is an unfair criticism of safe net. Certain things really need to be at the core. I suspect IPFS will struggle to just bolt on some of those features. Safecoin, for example is key to having a market for storage and had to be part of the core. Likewise, the security model is integral to the core.
We need extendable platforms, bit we also need firm foundations. Yes, it takes time to build those, but often it is worth it - it isn’t easy to create truly modular and extendable applications either. There will be many modules and applications on top too, if maidsafe continue to deliver.
Time will tell.
So long as it works, it is affordable, and it makes it to market in time, SAFE should be the better solution for a lot of things.
Although the 'seamless integration" isn’t likely to compete…
Another factor that gives IPFS an advantage is that the user does maintain a lot more control – If they want 10 copies in 10 different geographically distributed data centers, they can make that happen. With SAFE they must trust the black box that is SAFE to make all those decisions for them, and if somehow SAFE fails, they are SOL…
I would agree the IPFS isn’t there yet. They have the torso of the elephant, but are missing most of the legs in any tangible fashion… The torso is mighty useful though as it sits…
I see now that Maidsafe and IPFS are pitted against each other from an architectural standpoint. One is building by way of a more modern method to plug into existing technology as much as possible while the other has started over from the ground up. I certainly subscribe to the former method typically but what can be accomplished by starting over that can’t be accomplished by “modularizing from here”? Sometimes a lot. This will be fascinating to watch.
I do think IPFS’s modern approach will give it an advantage in people’s minds from a philosophical standpoint. SAFE will need to win them over by being awesome.
Do you have some risk with Oracle Lawsuits on Javascript as with Java?