AT2 discussion and questions

Features
23

Just like many people are currently cracking the code on interoperability between different blockchains, ultimately I think the same would be true for SAFE. Where there is a will there is a way, and I think different third parties will present distinct methods to solve for this.

2 Likes
24

Aren’t nearly all blockchains superceded by use of a single sequenced immutable datatype on SAFE? Same functionality at least… append entries to a public ledger ad infinitum…

5 Likes
25

Of course not. Who has permission to write the next entry into the data? How do you make sure it is filled with valid entries?

4 Likes
26

Maybe a blockchain record could have an underlying record holding all the signatures from the miners showing the block record as valid. And the software changed to only recognise blocks with enough positive signatures and a way to ignore bad blocks.

3 Likes
27

I keep banging on about ownership rules…

A way to append only if isValid(hash(block) < miningDifficult) could be handy.

2 Likes
28

It’s just a public ledger. The entries could be signed transactions from the originator with a proof that the sender has enough to cover, or a set of signed IOUs just like back in Sumeria. Access rules and concensus are handled by the datatype handling magic that is SAFE plus plus at2.
The simple logic gets wrapped up in an app… i.e. the entire blockchain experience becomes just one of many apps in the SAFE ecosystem.

1 Like
29

Hi,

I found that AT2 could/was patented already CORDIS | European Commission

Quote:

“We devised a generic asynchronous protocol to solve the secure causal broadcast problem, which we called AT2, Asynchronous Trustworthy Transactions, which we patented”

Does it have bad impact on the Safenetwork implementation?

Thanks

7 Likes
30

I tried to search for patents, couldn’t find any, but I only scratched the surface… Anyway, in some places this thing is referred as:

Asynchronous Trustworthy Transfers (not transactions) in case someone else wants to search.

1 Like
31

It is here, the last paragraph: https://cordis.europa.eu/project/id/862082/en?format=pdf

Or do you mean that you are looking for the patent itself?

2 Likes
32

Yeah, the patent itself.

EDIT: They say they patented it, but does it mean the patent was granted, or was it maybe just applied?

EDIT 2: I tried Google search and also European patent database, and couldn’t find a patent with the evident search terms. Also tried with a couple possible applicant names. But then it started to seem that these patent databases are a bit too difficult to understand for the uninitiated and I don’t know if there is a search that would cover them all, so I gave up.

2 Likes
33

For future ref this one is pretty good WIPO - Search International and National Patent Collections
I searched but no luck (just quick search)

4 Likes
34

So if an idea is used prior to a patent application…
Would that be an “all clear”?

35

Depends, in the USA they have a first use type thing, in other countries if any of that data was already public then the patent is bust. Patents are a nightmare, I dislike what they have become. We have pretty much let all our patents go dead. It was costing us almost $100K per year to keep them.

In this case here I doubt it is patented and not sure it could be, at least the ask for auth → get back enough and you are good. We have already got a lot of this kind of pattern in our code already.

13 Likes
37

This was a good read relating to patents and intellectual property in general: Against Intellectual Monopoly

38

You can’t find it because it hasn’t been published yet. First a provisional is filed, then the provisional is converted to a utility after a year. After a bit more than another year post utility filing, the document will be published. So assuming they converted their potential provisional filing to a utility, the timing suggests that it won’t be published for a while (and won’t be allowed for a while longer still).

With that said, I really wouldn’t worry about a potential IP here.

  1. Software patents are notoriously difficult to get right to satisfy (recent and changing) patentability requirements in US and EU alike

  2. Software patents are difficult to enforce (even when granted)

  3. Academic institutions typically only enforce IP against corporate/commercial operations

  4. As David says above, I’m not sure this is patentable given that many of the operations of AT2 are well known and many previously employed on Safe (there comes the patentability issue again)

  5. etc.

12 Likes
39

also it is the specific implementation what can be patented, so if you are using the ideas but having your own way of making that idea reality, then there is no patent infringement.
That’s how the PC clones were even possible.

3 Likes