OrbitDB: Peer-to-peer databases for the decentralized web
github.com
github.com
If you try to go with just OrbitDb you will have to handle manage the ipfs nodes and pin data so it stays persistent and loads faster but also handle the permissions too.
But with 3Box makes it easier to do all of the permissions and loading all data faster + always persistent and/or public or private data, and you know what's amazing? the way you can store data? for example if you want to build a chat app you can use Messaging (Threads) to make quickly a chat app! https://docs.3box.io/api/messaging
Enjoy!
CRDTs usually work as last-write-wins, meaning that if you have a key-value store, the last update to update a key 'wins' the value via the way oplog reduction works.
If you reverse that to a FIRST-write-wins log, you can grant permissions and ownership on a first-come, first-serve basis. Revocation, then, becomes the issue. What do you do with the records they already have? Questions like that are plentiful.
The approach most people take is to find workarounds or "good enough" solutions here, either by using encryption and allowing the encrypted data to be public, or by using some sort of other OrbitDB store as their ACL and management, and only giving select keys access to write to said ACL store in the first place.
Adding encryption into the mix though, particularly multi-writer, becomes exponentially harder.
Um.
What I meant, since LWW is nomenclature for an alternative to CRDTs, is that the last writer by _logical clock_ in a CRDT, not by _wall clock_ time, will "win" the key.
I think you're still stepping on rakes...
Can't reply to the comment below, so replying here.
I believe what markhenderson was trying to say is that in OrbitDB, the default merge strategy for concurrent operations is LWW.
The comment above is conflating a lot of things here. 1) determinism is exactly the guarantee one needs for CRDTs, and I'd argue generally is a good thing in distributed system but 2) adding vector clocks (OrbitDB uses Lamport clocks, or Merkle Clocks [1], by default), nor wall clocks, have nothing to do with determinism and in fact there's a good reason to not use vector clocks by default: they grow unbounded in a system where users (=IDs) are not known. In my experience, LWW is a good baseline merge strategy.
I don't think it's at all correct to say that "the winner is essentially arbitrary" because it's not. The "last" in LWW can be determined based on any number of facts. For example "in case of concurrent operations, always take the one that is written by the ID of the user's mobile device", or "in case of concurrent operations, always take the one that <your preferred time/ordering service> says should come first". It'd be more correct say "the winner is based on the logical time ordering function, which may not be chronological, real world time order".
As for the last comment, I'm pretty sure it's a CRDT system :) Want to elaborate your reasoning why you think it's not a CRDT?
[1] "Merkle-CRDTs: Merkle-DAGs meet CRDTs" - https://arxiv.org/abs/2004.00107
As I understand it, the Merkle-CRDT represents a Merkle tree as a grow-only set of 3-tuples. When you add a new event to the thing (as a tuple) you have to reference all of the current concurrent root nodes of the data structure, in effect becoming the new single root node; and your event data, which must be a CRDT, gets merged with the CRDTs of those root nodes. Do I have it right so far?
Assuming yes, let's say you have causality chain like so:
1 --> 2 --> 3 --> 4
`--> 5 --> 6
> you have to reference all of the current concurrent root nodes of the data structure, in effect becoming the new single root node
correct, and more precisely the union of heads is the current "single root node". in practise, and this is where the merge strategy comes in, the "latest value" is the value of the event that is "last" (as per LWW sorting).
> and your event data, which must be a CRDT, gets merged with the CRDTs of those root nodes.
the event data itself doesn't have to be a CRDT, can be any data structure. the "root nodes" (meaning the heads of the log) don't get merged with the "event data" (assuming you mean the database/model layer on top of the log), the merge strategy of the log picks the "last/latest event data" to be the latest value of your data structure.
> It's time to write a new value, so I create a new tuple with references to 4 and 6, and merge their CRDT values.
when a new value is written, correct that the references to 4 and 6 are stored, but the new value doesn't merge the values of the previous events and rather, it's a new value of its own. it may replace the value from one or both of the previous events, but that depends on the data model (layer up from the log).
1 --> 2 --> 3 --> 4
`--> 5 --> 6
no writes are lost. the result in your example depends what 4 and 6 refer to. in a log database, the ordered log would be eg. 1<-2<-3<-5<-4<-6, so all values are preserved. in the case of a key-value store, it could be that 4 is a set operation to key a and 6 is a set operation to key b, thus the writes don't effect each other. if 4 and 6 are both a set operation on key a, it would mean that key a would have the value from 6 and the next write to key a would overwrite the value in a. makes sense?
How do you know that? It's not inferrable from the DAG. Is sequencing also provided "a layer up"?
> if 4 and 6 are both a set operation on key a, it would mean that key a would have the value from 6 and the next write to key a would overwrite the value in a.
Yes, I mean for all of my events to be reads and writes of the same key. And you've proven my point, I think: if the resolution of this causal tree is Last-Writer-Wins, 6-domiantes-4, and "key a [gets] the value from 6", then whichever poor user was operating on the 3-4 causal branch has lost their writes.
This is a problem! If you claim to be a CRDT and offline-first or whatever, then as a user, I expect that the operations I make while I'm disconnected aren't just going to be destroyed when I reconnect, because someone else happened to be using a computer with a lexicographically superior hostname (or however you derive your vector clocks).
And if you want to say something like, well, when the unlucky user reconnects and sees that their work has been overwritten, they can just look in the causal history, extract their edits, and re-apply them to the new root -- there's no reason for any of this complex machinery! You don't need CRDTs to just replicate a log of all operations. Of course, you also can't do any meaningful work with such a data structure as a foundation, because it immediately becomes unusably large.
If it's possible to model operations in a commutative way (eg. instead of assigning values to keys one just stores differences), then the conflict resolution is mathematically proven, just apply all operations in whatever order, they're commutative, great. Of course it doesn't help with "real world data", but that's where mathematically we can use and oracle (the user, or whatever linearizer service we choose).
I jumped the gun there and made an assumption that the value of a node is the LWW ordering :) Ok, so without that assumption, the DAG
1 --> 2 --> 3 --> 4
`--> 5 --> 6
((0, x), a, 1) --> ((1, x), a, 2) --> ((2, x), a, 3) --> ((3, x), a, 4)
`--> ((2, y), a, 5) --> ((3, y), a, 6)
> This is a problem! If you claim to be a CRDT and offline-first or whatever, then as a user, I expect that the operations I make while I'm disconnected aren't just going to be destroyed when I reconnect, because someone else happened to be using a computer with a lexicographically superior hostname (or however you derive your vector clocks).
You're conflating the data(base) model with the log and we can't generalize that all cases of data models or merge conflict are cases of "I expect all my operations to be the latest and visible to me". They are semantically different. If the writes are on the same key, one of them has to come first if the notion of "latest single value" is required. If the writes are not on the same key, or not key-based, multiple values appear where they need to. What we can generalize is that by giving a deterministic sorting function, the "latest value" is the same for all participants (readers) in the system. From data structure perspective this is correct: given same set of operations, you always get the same result. For many use cases, LWW works perfectly fine, and if your data model requires a "different interpretation" of the latest values, you can pass in your custom merge logic (=sorting function) in orbitdb. The cool thing is, that by giving a deterministic sorting function for a log, you can turn almost any data structure to a CRDT. How they translate to end-user data model will depend (eg. I wouldn't model, say, "comments on a blog post" as a key-value store).
If you're curious to understand more, I think the model is best described in the paper "OpSets: Sequential Specifications for Replicated Datatypes" [1]. Another two papers, from the same author, that may also help are "Online Event Processing" [2] and "Moving Elements in List CRDTs" [3] which show how by breaking down the data model to be more granular than "all or nothing", composing different CRDTs give arise to new CRDTS, which I find beautiful. Anything, really, that M. Kleppmann has written about the topic is worth a read :)
[1] https://arxiv.org/pdf/1805.04263.pdf [2] https://martin.kleppmann.com/papers/olep-cacm.pdf [3] https://martin.kleppmann.com/papers/list-move-papoc20.pdf
1. Modeling an append-only log as a CRDT is trivial
2. Building a database on top of a "CRDT" append-only log doesn't make the database a CRDT
3Box also has support for members only OrbitDB threads which can restrict posting to members, and encrypted OrbitDB threads to make posts private to the group. To Mark's point above.
This also works for encryption in members threads. To remove a member, you can create a new thread without the member and link the original thread. This gives forward secrecy since new encryption keys are generated for the new thread.
We're working on improving this system over time, but it requires some more advanced cryptography such as proxy re-encryption (like nucypher).
Examples: content about Tiananmen Square 1989.
More advanced moderation tools can be built on top, too.
Some basic questions, as I am still struggling to see the use case of this project.
* How did this project get started? What problem is it trying to solve?
* Are there any real world examples of where this project is used?
OrbitDB got started because we wanted to build serverless applications, especially for the web (ie. applications that run in the browser.) Serverless meaning "no server" and no central authority, ie. something that can't be shut down.
OrbitDB gives tools to build systems and applications where the user owns their data, that is, the data that is not controlled by a service. As a way of simple example, imagine Twitter that doesn't have one massive database for all tweets, but rather you'd have one database for each user.
I wasn't there at the beginning but I believe the project came out of trying to achieve said mutable state within IPFS (which, for other readers is content-addressed and therefore append-only)
http://orbitdb.org lists all of our current users, the biggest one is Metamask by way of https://3box.io. https://tallylab.com is building with it for remote encrypted backup and shared tallies. https://github.com/dappkit/aviondb is a MongoDB-like interface for it.
What I'd like to know is do they support authorization?
Identity providers work by cross-signing an external keypair with the generated OrbitDB key-pairs, and access controllers generally work by exposing a `canAppend` function that facilitates any kind of auth you want to perform. There's support for Metamask for example. OrbitDB is also used _in_ Metamask under the hood by our good friends at 3box.[2]
TallyLab, my application project, uses these in its IAM system and you can see the repos here.[3]
1. https://github.com/orbitdb/orbit-db-access-controllers 2. https://3box.io/ 3. https://github.com/tallylab
So, basically you can have any sort of Web 2.0 or Web 3,0 based authentication/authorization methods. Oauth, JWT, DIDs...anything you can imagine.
Link to Access Controller: https://github.com/orbitdb/orbit-db-access-controllers#creat...
Security is easy in some cases. Sign your log appends. Security around transactions without a central authority is pretty complex. Whats the solution there? Block-chain type things? It would have to be application specific I suppose.
I never gave FirebaseDB a shot because it felt like too much vendor lock in but this might be interesting to try one day.
HOWEVER! Our latest release, 0.23.0, mitigated this by using a power-of-2 skiplist to load things in parallel, which gave us a nice 4-5x boost there.
(Is there a design doc for OrbitDB anywhere?)
You'd be surprised how well JS does on both fronts, in addition to being able to run across platforms :)
Define robust:
1. Strong and healthy; vigorous
- (of a process, system, organization, etc.) able to withstand or overcome adverse conditions.
JS is the most robust language we have according to the main definition, however you measure it. What other language is as alive as JS right now? What other language has as much programmer attention?
JS is also easily robust according to the noted sub-definition. What other language would survive the web? The web, an environment that could easily be described as "very adverse"... What other language is ready to run in a browser where it will be dynamically and continuously mixed with modules from many different sources, without breaking? JS was built for this.
What other popular language is so easily runnable cross platform? What other popular language lets you program with both OOP and functional paradigms while also being as popular as JS? What other popular language lets you monkey patch things to fit a piece of code into any situtation?
Robust means all of these things to me. What does robust mean to you and what is your example of the most robust programming language? You didn't say...
None of these fit the bill: Python, Ruby, C, C++, Golang, Rust, C#.
Any language that has a dynamic & weak type system, and where monkey patching is not only allowed but used widely, cannot make a claim to leading to robust software.
I would argue that languages focusing on correctness are the ones that would have a good claim at producing more robust software. As you've asked, I would say that Rust, Ada, Haskell fit the bill... but even languages that focus on keeping simplicity and boring uniformity at scale, like Go and Java, could still make such claim (and lots of software written in them can be described as robust) a lot more than JS, which offers basically 0 features focused on correctness beyond the bare minimum.
> (does not fail or lose data, or get into a corrupt state easily)
The data is stored on IPFS primarily (Pick a number of different ipfs varieties, rust, golang, js). Oplog specific data is stored in a datastore-level instance, which utilizes c++ leveldown bindings. Code that is well tested and has never been proved to lose data due to being written in JS alone. Aside from that the programming language does not determine whether a developer creates crappy code. The developer does, horrible, inefficient, bug filled code can exist in any number of languages. Its really a matter of preference.
Side note: I used to absolutely hate JS until a began programming in it for many months. Originally programming Java
Nope. I actually just disagreed with you. I think JS is a robust language and I said why very clearly and politely.
You, however - well I stopped reading after your first sentence.
Good luck to you!
You didn't say why JS is robust, you twisted the word's definition to fit your point. Don't do that.
If your way of debating is to leave the debate when someone says something you don't like, then good luck working in IT.
It simply leads to more robust software because compiler rejects large set of incorrect programs that would otherwise end up in production. Your test coverage has to be much larger just to verify that your program is sane on the most basic level.
Rusts compiler is even stricter than most.
Not to mention resilience to change, JS is simply a blunder
How does OrbitDB compare to GUN?
I went to a meetup where the creator of GUN gave a presentation. To this day, it was the single strangest presentation I've ever experienced.
The presentation had numerous obvious errors. The audience kept pointing out major misunderstandings and errors about distributed systems. I remember one awkward moment where someone pointed out a glaring error in the author's model of conflict resolution that undermined one of his key points, which the GUN creator tried to gloss over as quickly as possible. To top it off, he ran out of slides before he finished his talk because he hadn't bothered to complete his whole slide deck before presenting. Instead, he just tried to improvise the last 1/3 of his talk about GUN.
It was so bad that we all walked out of there wondering if we had just been trolled. To this day, I'm still surprised when I read about GUN on Hacker News.
You don't need to take my word for it. I encourage anyone considering GUN to open up the source code and look through some random files: https://github.com/amark/gun/tree/master/src
I'm not even a Gun fanboy. I've just used it successfully and it has done for me what it claims to do.
Could it be that Hacker News has a vested interest in seeing competitors to IPFS fail? I'm not sure, but either way, your comment grosses me out.
But I don't see why you couldn't try GUN if you wanted to. It's disturbing how HN seems to demand a kind of "brand loyalty" regarding technology. Just use what works.
Key-Value databases, feeds, and other data model types that OrbitDB supports by default, are all built on that log. You can also create your custom database types, ie. custom data models.
I once tried to fix a bug that I found in GUN but I gave up after just trying to figure out what the code is supposed to do.
old school javascript like this have an advantage: you can ship as is, without the need to obfuscate under the excuse of performance.
if you load two hundred build dependencies and then write twenty word long variables, it might look nice for you at development, but a user trying to debug or validate code on the fly would be in a very bad position as you would have to 'minify' (i.e. obfuscate) to not have a 70mb production file.
Code like this is very readable if you know some conventions, things like `cb` being a callback pointer. Its not even close to the cryptic knowledge needed to read assembly on weird platforms, for example. And a far cry from obfuscated.
> load two hundred build dependencies and then write twenty word long variables [...] a very bad position as you would have to 'minify' (i.e. obfuscate) to not have a 70mb production file
You defeated your own argument here. You're saying it's nice that the code is obfuscated already (short variables) so you can get better performance, it's really the same thing - except doing it by hand has a lot of downsides.
Minifiers and source maps have been around for a long time to get you the best of both worlds, understandable code in development, minified code in production (even though gzip alone gives you 80-90% of the gains). There is absolutely no reason to write obfucasted code like this [1] where you need to guess the meaning of thirty different one-letter variables. Grep'ing the code becomes impossible. This has nothing to do with 'old school' JS, but universal code standards.
( Not sure "obfuscated" is the word I'd use, though. )
(which also seems to be putting https://ipfs.io under strain. Sorry, ipfs.io!
And if you want to operate over a distributed network, which means you WILL have network partitions, then you are subject to CAP and will need eventual consistency mechanisms.
This is a false. None of the things you listed preclude consistency.
> And if you want to operate over a distributed network, which means you WILL have network partitions, then you are subject to CAP and will need eventual consistency mechanisms.
That's not how CAP works. Plenty of distributed CP databases exist.