It’s definitely a good read. I like the approach, from peercoin’s point of view it’s well isolated, no requirement on the core system. I am going to think about it a bit more to see if I could produce some more abstraction.
In the section regarding double spending, you mentioned transaction timestamp. The first come first serve rule best uses the serialization order of transactions, instead of transaction timestamp. The serialization order is (block height, transaction sequence number in block, transaction output sequence number in transaction).
[quote=“Sunny King, post:21, topic:3821”]It’s definitely a good read. I like the approach, from peercoin’s point of view it’s well isolated, no requirement on the core system. I am going to think about it a bit more to see if I could produce some more abstraction.
In the section regarding double spending, you mentioned transaction timestamp. The first come first serve rule best uses the serialization order of transactions, instead of transaction timestamp. The serialization order is (block height, transaction sequence number in block, transaction output sequence number in transaction).[/quote]
Logic behind preventing double spend is simple, client needs to deduct which messages are legit and which try to fool him.
This is done in following logic:
- check which PSA message is older (this can naturally be done by checking block height first) - older message is correct
- if there is more than one relative PSA message in the same block, use timestamp to see which one is the oldest (first one written by minter is the oldest)
What would be the benefit of using “serialization order of transactions” over comparing timestamps?
Is there something I am missing here?
[quote=“peerchemist, post:22, topic:3821”]What would be the benefit of using “serialization order of transactions” over comparing timestamps?
Is there something I am missing here?[/quote]
All transactions in the same block have the same timestamp if I’m not mistaken.
The serialization order sorts transactions in the same block, while timestamps can’t.
Yes, you are both right.
I remember specifically bisecting one block and comparing tx timestamps (albeit only two of them) to confirm my theory on how to verify tx’s in the same block.
Now, I’ve checked again. It seems I have failed at recognizing the same numbers the first time. :-/
Hi Peerchemist,
Have you heard of “Open Asset Protocol” https://github.com/OpenAssets/open-assets-protocol/blob/master/specification.mediawiki?
It seems very similar to PSA (haven’t studied it in depth).
A reference implementation on the Bitcoin blockchain exists in python https://github.com/OpenAssets/openassets.
It seems like you could start from that implementation and port it to Peercoin.
Or is there a substantial difference?
[quote=“hrobeers, post:25, topic:3821”]Hi Peerchemist,
Have you heard of “Open Asset Protocol” https://github.com/OpenAssets/open-assets-protocol/blob/master/specification.mediawiki?
It seems very similar to PSA (haven’t studied it in depth).
A reference implementation on the Bitcoin blockchain exists in python https://github.com/OpenAssets/openassets.
It seems like you could start from that implementation and port it to Peercoin.
Or is there a substantial difference?[/quote]
I know about it, I did some reading on it after I had idea to create PSA to compare the idea. Open assets is basically a play on “colored” coins idea.
So it is quite different and I am convinced that PSA will work better as it is even simpler and more robust.
Too bad they have already taken the name “open assets” though, as that sounds like something that has been agreed upon to become the standard.
Maybe PSA should become “Peerassets” as in “peer2peer assets” which also seems like something fit for standard.
There is not much PSA implementation can use from that library though, due to differences in protocol, fee dynamics of underlying blockchain and all PSA will have to start from scratch.
This one seems to be a simple asset implementation too: http://www.omnilayer.org/
I trust that you know the differences, I wasn’t sure if you heard about it. For now, I don’t see the differences myself, but will probably change when I study them further.
BTW: I found these simple asset implementations by reading http://coinsecrets.org/, it shows the latest OP_RETURN transactions on the Bitcoin blockchain.
[quote=“hrobeers, post:27, topic:3821”]This one seems to be a simple asset implementation too: http://www.omnilayer.org/
I trust that you know the differences, I wasn’t sure if you heard about it. For now, I don’t see the differences myself, but will probably change when I study them further.
BTW: I found these simple asset implementations by reading http://coinsecrets.org/, it shows the latest OP_RETURN transactions on the Bitcoin blockchain.[/quote]
Omni, former MasterCoin uses “auxiliary” currency in form of Omni tokens which are required for something. So it might be simple but it is needlessly complicated and introduces another currency.
A paper describing some ingenious performance optimizations for PSA will be released sometime in next 48h. I will leave the honor of presenting it to the author, so I will not say much for now. 
After that paper is released I will start wrapping up this draft into final paper with proposed v1.0 of protocol and I will propose development roadmap. I hope final paper and protocol will be defined in next 10-14 days.
One more thing; PSA will be re-branded as PeerAssets with subtitle “peer-2-peer Assets” before final release which should make it more obvious that the protocol can work on any blockchain based on original Bitcoin technology, Bitcoin included, while retaining association with Peercoin (peer-2-peer Coin).
It is important to avoid excluding this protocol from other blockchains and communities as we all strive to the same goal, while choosing different road.
PeerAssets (peer-2-peer Asset) should be movable between blockchains as their users/issuers please. Assets must be transferable between their blockchain hosts to enable thriving and resilient economy. A protocol scheme which will enable moving the PeerAssets based Asset to different blockchain will be described in future releases.
This statement will hopefully encourage adoption of the protocol for those who doubt in Peercoin and it’s blockchain as it will enable them to simply move on and continue their asset elsewhere if they feel so without loosing the asset value and surrounding community. It will naturally enable movement from other blockchains to Peercoin as well.
Here is it: Cryptoblog - notícias sobre bitcoin e criptomoedas!
Not sure if they are that ingenious though, but I learned a lot by writing it.
This idea was eventually evolved in PeerAssets.