News for hackers. Stuff that matters.
Re:RFC 6920: Naming Things with Hashes by btrask on Saturday February 20, 2016 @08:26PM (#11142853)
Lots of comments are asking what problem this solves.

First, content addressing is the best general solution to exactly-once message delivery in a distributed system. Examples of distributed message protocols include (ahem) email, HTTP and RSS. The alternative, UUIDs, is strictly worse in almost every way (which is to say, if you're currently using UUIDs for anything, consider switching to hashes of the content).

Second, content addressing guarantees the integrity of messages, meaning as long as you have the hash, you can get the data from anywhere. This is highly useful for mirroring and failover. Basically, we might be able to solve linkrot to a large degree, and make easier to mirror sites (especially locally, which is useful for high latency and offline).

I wrote up a short comparison of this protocol, along with a few alternatives a while back.[1] (Note that the IPFS comparison is incorrect, it should have an /ipfs/ path prefix.)

[1] https://bentrask.com/?q=hash://sha256/1e9a6b770ef9f1ca894af4...

Re:RFC 6920: Naming Things with Hashes by DyslexicAtheist on Saturday February 20, 2016 @09:18PM (#11142981)
"Oh, you wanted to increment a counter. I'm sorry it's not that simple Dave" -- distributed computing literature
Re:RFC 6920: Naming Things with Hashes by anon4 on Sunday February 21, 2016 @06:13AM (#11144077)
You just contact the central authorisation server and get a universal authorisation request authorisation.

Then, you take the universal authorisation request authorisation token and hand it off to the counter incrementing authorisation server which gives you a counter incrementing authorisation token.

Then, you take the counter incrementing authorisation token to your local counter incrementing server and ask it to increment the counter by 1.

In turn, your local counter incrementing server checks that your counter incrementing authorisation token was indeed signed by the central authorisation server and issues a counter incrementing operation to one of the 8 central counter incrementing servers.

Of course, your local counter incrementing server is itself authorised by the central authorisation server and the counter incrementing authorisation server to issue authoritative counter increment operations; so if its own local counter incrementing authorisation token has expired (5 minutes), it also has to do steps 1-3 to renew its authorisation to perform authoritative counter incrementing operations. This doesn't concern you, as all those requests are buffered so this section is purely informative.

Your own authorisation tokens also have expirations (5 minutes for the unviersal authorisation request authorisation token and ~10 seconds for the specific authorisation tokens) so you best issue your requests promptly and make sure to renew the tokens. Should your token expire before you have completed your counter increment request to your local counter increment server, perform steps 2-4 (or 1-4 if more than 5 minutes have elapsed since you acquired your universal authorisation request authorisation token).

Expect to receive a counter increment confirmation message within the next 2 full moons by carrier pigeon.

Super simple stuff.

Re:RFC 6920: Naming Things with Hashes by fennecfoxen on Sunday February 21, 2016 @10:38AM (#11144878)
> You just contact the central authorisation server

No! No!! No!!!

First, we're doing distributed computing. There is no central authorization server. There are N authorization servers, and by the time you know for 100% sure whether the majority agree on whether you're authorized, someone has revoked the authorization. So take your authorization token, we'll write it down when you use it, and if you're de-authorized we'll adjust the counter after the fact to compensate, during our periodic audit.

But more importantly, the real problem is that your counter increment message was sent to a machine that has been taken offline by an errant backhoe and we don't know whether it took. Fortunately you sent the counter-increment message to three different servers, so once we have time to go through all the counter increment messages to de-duplicate we'll have your count, and the problem today is making sure that we have a full three copies of the message stored instead of just two.

Re:RFC 6920: Naming Things with Hashes by fla on Sunday February 21, 2016 @08:33AM (#11144417)
Don't forget to instantiate a CounterIncrementingAuthorizationTokenFactoryFactory in your local cluster ;)
Re:RFC 6920: Naming Things with Hashes by DyslexicAtheist on Sunday February 21, 2016 @06:49PM (#11147315)
new counter increment pigeon. who dis?
Re:RFC 6920: Naming Things with Hashes by haberman on Sunday February 21, 2016 @03:38AM (#11143810)
If you use a message hash as your unique identifier for exactly-once delivery, your messaging system won't let a user send the same message twice, even if they mean to.

You could wrap the user's message in some metadata including a uuid to ensure the overall content is distinct from anytime the user previously posted that message. But now you're back at uuids, which you say you don't like.

Re:RFC 6920: Naming Things with Hashes by btrask on Sunday February 21, 2016 @05:34AM (#11144009)
You need to think about what makes messages distinct. Content addressing will help you as long as you have a rigorous definition of identity, which you might otherwise be doing ad-hoc when you assign UUIDs.

If the same message is meaningful at different points in time, your hashed content should include a timestamp (and of course you need to choose an appropriate level of precision). If your message is meaningful in different contents, it needs to reference that context somehow (ideally by hash).

The payoff for doing this well is 1. an elimination of double-sent emails, double-posts, etc. (when the software might not know if it went through the first time), and 2. the ability to deduplicate across a network partition, for example if a user makes the same change on their computer and on their phone and then syncs them.

Re:RFC 6920: Naming Things with Hashes by zAy0LfpBZLC8mAC on Sunday February 21, 2016 @09:04AM (#11144519)
... except wallclock timestamps for ordering of events in a distributed system is a terrible idea in most cases for the simple reason that (a) it's not a (even close to) guaranteed unique event identifier (the same time happens in many locations), (b) you usually don't have a clock of sufficient precision to even reliably establish a total order of distinct moments and (c) even if you have, that order does not necessarily preserve causality.

Now, UUIDs don't solve (b) or (c) either, but at least they actually give you a reliable identifier for events.

Re:RFC 6920: Naming Things with Hashes by nhaehnle on Sunday February 21, 2016 @09:53AM (#11144702)
The context here is events originated by a single actor, so wallclock is much more reasonable than in the general distributed system case. You might still be worried about the user accessing your system simultaneously from multiple devices, so a randomly generated ID is probably still useful.
Re:RFC 6920: Naming Things with Hashes by zAy0LfpBZLC8mAC on Sunday February 21, 2016 @10:12AM (#11144778)
First, it's usually impossible to tell actors apart, so it's not particularly useful to build an algorithm that depends on doing so.

Second, while the actor might be the same, the time source usually is not, so it still doesn't help you.

Third, if actor and time source are indeed the same, you really are using a counter, the fact that it's correlated with wallclock time is coincidental.

Re:RFC 6920: Naming Things with Hashes by philsnow on Sunday February 21, 2016 @04:07PM (#11146532)
Piling on, even if the actor and time source are the same, you have to be really careful about how your use "wall clock time", because if you're just asking the OS what time it is, a lot of platforms allow the clock to move backwards, not move, etc.
Re:RFC 6920: Naming Things with Hashes by btrask on Sunday February 21, 2016 @09:23AM (#11144569)
No one is talking about ordering events.
Re:RFC 6920: Naming Things with Hashes by zAy0LfpBZLC8mAC on Sunday February 21, 2016 @10:13AM (#11144784)
... in which case my post explains why UUIDs are the better solution.
Re:RFC 6920: Naming Things with Hashes by btrask on Sunday February 21, 2016 @10:23AM (#11144827)
Content hashing doesn't rely on timestamps to act as a unique identifier. It relies on all of the content, which may (or may not) include a timestamp, depending on your application. To the degree that collisions occur, that's a feature, not a bug. If you get undesirable collisions, you aren't hashing the right data.

Edit: to clarify, content hashes are just like UUIDs except you can get useful collisions if you want them.

Re:RFC 6920: Naming Things with Hashes by schrodinger on Sunday February 21, 2016 @09:27AM (#11144590)
Pretty sure messages are ordered events in this scenario, no?
Re:RFC 6920: Naming Things with Hashes by btrask on Sunday February 21, 2016 @09:41AM (#11144652)
They're events, but no ordering is assumed, guaranteed, promised, implied, or enforced. The timestamps, if you choose to use them, are only for determining message identity (equality) and thus can be as accurate or inaccurate as relevant for your application.

You don't have to use timestamps if they don't make sense. For example, what is the "identity" of a message in an instant messaging program? If you just use the message content, then duplicate messages don't work. It would make sense to include the sender and the time written (which is different from the time sent!) as well. You might also include the recipient, depending.

Note that instead of using timestamps, you could use the hash of the message being replied to. However, IM software usually doesn't have a system for real replies, so doing that properly probably isn't worthwhile. (Edit: however, hashes are very useful for systems like email, forums, etc., where reply targets are explicitly defined.)

If you do want ordering, BYO. Content addressing and UUIDs are equivalent in that regard.

Re:RFC 6920: Naming Things with Hashes by zAy0LfpBZLC8mAC on Sunday February 21, 2016 @10:26AM (#11144840)
> The timestamps, if you choose to use them, are only for determining message identity (equality)

Which simply doesn't work because timestamps are not unique to location. Every clock experiences every timestamp, and thus potentially assigns any timestamp to an event, so multiple clocks can end up assigning the same timestamp to distinct events, and thus identifying them.

> Note that instead of using timestamps, you could use the hash of the message being replied to.

If your message isn't unique, the message you are replying to probably isn't either, so it's useless to use it as an identifier.

> If you do want ordering, BYO. Content addressing and UUIDs are equivalent in that regard.

No, they are not, at all, they aren't even the same kind of thing. UUIDs are unique identifiers, you generate them to label things that couldn't be distinguished without them. Content addressing is nothing more than summarizing a given object in a way that (almost certainly) preserves identity. So, one creates distinction, the other maintains identity.

Re:RFC 6920: Naming Things with Hashes by btrask on Sunday February 21, 2016 @10:37AM (#11144872)
> Which simply doesn't work because timestamps are not unique to location.

Yes, I know. This is okay. Whatever time you think you wrote the message at is fine. Even if your clock is off. Even if you lie. That is (part of) the message content (in this case).

The time the message is sent is meaningless and can't be used, I agree. The time the message was written, or created, or "thought to be created" is what matters (or doesn't).

> If your message isn't unique, the message you are replying to probably isn't either, so it's useless to use it as an identifier.

Git does this and works great. If your entire message history (via hash chain) is identical, then you really do have duplicate messages.

> UUIDs are unique identifiers, you generate them to label things that couldn't be distinguished without them. Content addressing is nothing more than summarizing a given object in a way that (almost certainly) preserves identity. So, one creates distinction, the other maintains identity.

I agree, you're right. However, this is a feature.

Re:RFC 6920: Naming Things with Hashes by zAy0LfpBZLC8mAC on Sunday February 21, 2016 @11:22AM (#11145063)
> Yes, I know. This is okay. Whatever time you think you wrote the message at is fine. Even if your clock is off. Even if you lie. That is (part of) the message content (in this case).

Yes, it's part of its content, sure. That doesn't make it an identity of the event that originated it.

> The time the message is sent is meaningless and can't be used, I agree. The time the message was written, or created, or "thought to be created" is what matters (or doesn't).

But still doesn't provide an identity of the event.

> Git does this and works great. If your entire message history (via hash chain) is identical, then you really do have duplicate messages.

But that is simply because you have defined message identity in this context to mean just that, which is fine because of the semantics of the "messages" that git is made to manage. What it does not is provide identity of commit events. You can not, for example, implement a distributed counter in git that works by committing changes to a file that add a line with a number in it. If two git instances commit adding the same number based on the same parents at the same time with the same authorship information [...], the merge result might not reflect all counter increment operations that happened.

> I agree, you're right. However, this is a feature.

Well, it just is what it is, and each has its uses, but they are mostly orthogonal. It's just that when a UUID is part of an identity, summarizing that identity is easier done by dropping everything but the UUID instead of hashing everything.

Re:RFC 6920: Naming Things with Hashes by btrask on Sunday February 21, 2016 @12:00PM (#11145246)
> You can not, for example, implement a distributed counter in git that works by committing changes to a file that add a line with a number in it. If two git instances commit adding the same number based on the same parents at the same time with the same authorship information [...], the merge result might not reflect all counter increment operations that happened.

Yes. That's why content has to be defined at the application level, based on what the application is trying to accomplish. If absolutely every message is logically unique, then under a content addressing system each message needs to include some random data, which will result in random hashes equivalent to UUIDs. That is the worst case scenario for content addressing.

Even so, it's still "as good or better" (aside from performance concerns, which I don't mean to dismiss).

Re:RFC 6920: Naming Things with Hashes by redka on Sunday February 21, 2016 @04:09AM (#11143874)
A counter solves this without UUID
Re:RFC 6920: Naming Things with Hashes by ketralnis on Sunday February 21, 2016 @04:21AM (#11143885)
Counters have a host of other problems, like making sure that you never issue the same counter twice in a system without a central authority (or just one that doesn't desire a single point of failure)
Re:RFC 6920: Naming Things with Hashes by ec109685 on Sunday February 21, 2016 @02:41AM (#11143702)
There are applications where you want to be able to update a piece of content of the document without changing how it is addressed.
Re:RFC 6920: Naming Things with Hashes by jjtheblunt on Saturday February 20, 2016 @09:54PM (#11143078)
Content addressing most definitely does NOT guarantee the integrity of a message, because hashes are non isomorphic homomorphisms. Cf. an article on "cryptographic hash"ing, for instance, which are hashes with strong statements about how hard it is to find distinct messages with the same hash, etc. .
Re:RFC 6920: Naming Things with Hashes by btrask on Saturday February 20, 2016 @10:09PM (#11143112)
I'm not sure I understand your criticism, but I should've clarified "content addressing using long, cryptographically secure hashes". Most of the major projects are using SHA-256.
Re:RFC 6920: Naming Things with Hashes by jjtheblunt on Sunday February 21, 2016 @02:50AM (#11143722)
The point is that hash functions of fixed bit length b have 2^b possible values, but are used to "represent" bit sequences of length longer than b, of which there are more than 2^b values. Collisions necessarily occur, but finding colliding values isn't necessarily easy (cf. cryptographic hashes), which means if you want to tamper with a message, which has an associated hash, without being detected, you're constrained by the hash function to tamperings which produce the same hash.
Re:RFC 6920: Naming Things with Hashes by dchest on Sunday February 21, 2016 @04:55AM (#11143948)
Which basically means that it guarantees the integrity of message, where "guarantee" is defined as computational infeasibility to produce different messages with the same hash value. You know this stuff, why are you arguing semantics?
Re:RFC 6920: Naming Things with Hashes by jjtheblunt on Sunday February 21, 2016 @12:19PM (#11145362)
I hear you regarding infeasibility, though right now there's another article about content hash collisions being used to distribute a tampered-with Mint Linux image (MD5 collisions).

It's "easier" to get the computational resources to fabricate a collision there, of course, but who is to say how much compute power state level adversaries bring to bear?

Re:RFC 6920: Naming Things with Hashes by dchest on Sunday February 21, 2016 @04:03PM (#11146512)
To be fair, MD5 was broken in 1996. If your content addressable system uses crypto broken 10 years ago, of course, you can't guarantee integrity :-)

who is to say how much compute power state level adversaries bring to bear

Not enough to break modern hash functions.

Re:RFC 6920: Naming Things with Hashes by drdaeman on Sunday February 21, 2016 @03:59AM (#11143851)
I guess, the most widely used content-addressing system is BitTorrent.

Wonder, if anyone had already did a research how many different torrents with the same `info_hash` one can find out there.

Re:RFC 6920: Naming Things with Hashes by dchest on Sunday February 21, 2016 @04:59AM (#11143955)
None. Accidental collision resistance of SHA-1 is 2^80.

SHA-1 is "cryptographically broken": it's estimated that you can produce collision with about 2^60 operations, however nobody has produced them yet.

Re:RFC 6920: Naming Things with Hashes by versteegen on Sunday February 21, 2016 @07:58AM (#11144317)
The question is about duplicate torrents (ie content addressing), not hash collisions.
Re:RFC 6920: Naming Things with Hashes by versteegen on Sunday February 21, 2016 @04:26AM (#11143895)
I had to look it up, and found info_hash is the SHA-1 of the 'info' section of the metainfo file, which lists the files and the SHA-1 hashes of each of their pieces. So you are asking how many torrents have exactly the same file data but different metadata. So yes, interesting question.

If two torrents differ only in metadata, will any bittorrent clients actually allow the data be shared (without manual intervention like symlinking files)?

Re:RFC 6920: Naming Things with Hashes by akvadrako on Sunday February 21, 2016 @05:11AM (#11143978)
That would be an SHA-1 collision which has never been reported and would break lots of encryption in common use.

You guys are worrying about the wrong thing. Any hash considered secure will have zero collisions, at least in your light cone for trillions of years.

If current hashes are shown to have weaknesses they will be deprecated, probably before even a single collision takes place.

Re:RFC 6920: Naming Things with Hashes by versteegen on Sunday February 21, 2016 @07:55AM (#11144311)
No, read what I wrote again. Same files, different metadata (eg. tracker URL, comments, creation date)