chanc3r
|
|
January 21, 2014, 05:08:34 PM |
|
I don't think there is an api call to filter AM's where recipient = X and sender = Y, right? (this should be added)..
Just call getAccountTransactionIDs for the account then loop and find the first one where recipient == avatar_account. Erm what's the processing time for every thousand transactions... sounds a bit like avatar mining
|
|
|
|
Come-from-Beyond
Legendary
Offline
Activity: 2142
Merit: 1010
Newbie
|
|
January 21, 2014, 05:08:41 PM |
|
Get rid of referencedTransaction.
|
|
|
|
Come-from-Beyond
Legendary
Offline
Activity: 2142
Merit: 1010
Newbie
|
|
January 21, 2014, 05:09:23 PM |
|
Guys, u were talking about splitting the bounty to 4 parts. Did u come to a decision? If u r agree to split the bounty then I won't spend time on speedtests. Well I wouldn't mind. We all spend a lot of time coding. I guess hoax and gimre don't mind too, but I haven't heard Jaguar's opinion yet. Ok, let's wait for Jaguar.
|
|
|
|
jl777
Legendary
Offline
Activity: 1176
Merit: 1134
|
|
January 21, 2014, 05:10:29 PM |
|
Does anybody know what time the strange fractional orders on dgex got cleared last night? For a few hours bter was trading 50% above dgex. That is clearly unsustainable and my guess is that the bter market got arbitraged down to the dgex price. Since the dgex price seemed to be stuck, this means the downward move from .00016 was not due to big seller, but arbitrage, or perceived arbitrage. The NXT/CNY value peaked at .85 and dropped to match the dgex price.
It seems the dgex price is unstuck now and this bounce at .00016 is the second time, regardless of the reason. It is not uncommon for a price to bounce three times before breaking out. I think it is good that it has come back down a bit as the rise was too extreme. Gather some support at .0001 and gradually get back to .00016, bounce back a bit and then finally breakthrough. That would be a sustainable pattern that would then turn .00016 from a price ceiling (resistance) to price floor (support).
Of course, this is crypto, so anything can happen, but I think there is a significant amount of fast money parked in NXT right now, so it could follow more standard technical patterns
James
|
|
|
|
ferment
Full Member
Offline
Activity: 168
Merit: 100
IDEX - LIVE Real-time DEX
|
|
January 21, 2014, 05:11:22 PM |
|
I don't think there is an api call to filter AM's where recipient = X and sender = Y, right? (this should be added)..
Just call getAccountTransactionIDs for the account then loop and find the first one where recipient == avatar_account. Erm what's the processing time for every thousand transactions... sounds a bit like avatar mining It's fast particularly if you cache and use timestamp. Or, a service provider could certainly maintain an http accessible cache. If you're not into the decentralized thang.
|
|
|
|
msin
Legendary
Offline
Activity: 1470
Merit: 1004
|
|
January 21, 2014, 05:11:37 PM |
|
Guys, u were talking about splitting the bounty to 4 parts. Did u come to a decision? If u r agree to split the bounty then I won't spend time on speedtests. Well I wouldn't mind. We all spend a lot of time coding. I guess hoax and gimre don't mind too, but I haven't heard Jaguar's opinion yet. Ok, let's wait for Jaguar. He's busy with NEM. Question for CFB; Is a Reputation System being developed by JLC or You? (or plans to develop) I was going to offer a bounty for a working version, implemented using public AM.
|
|
|
|
jl777
Legendary
Offline
Activity: 1176
Merit: 1134
|
|
January 21, 2014, 05:13:07 PM |
|
Guys, u were talking about splitting the bounty to 4 parts. Did u come to a decision? If u r agree to split the bounty then I won't spend time on speedtests. Well I wouldn't mind. We all spend a lot of time coding. I guess hoax and gimre don't mind too, but I haven't heard Jaguar's opinion yet. Ok, let's wait for Jaguar. How much work is it to benchmark all 4 on the same platform? I would like to know the TPS and I think it makes sense for there to be published performance and also test results. Since it is JS, doesn't that mean it is browser specific? Do all four run on all popular browsers? James
|
|
|
|
BitcoinForumator
Legendary
Offline
Activity: 1120
Merit: 1000
|
|
January 21, 2014, 05:14:50 PM |
|
Does anybody know what time the strange fractional orders on dgex got cleared last night? For a few hours bter was trading 50% above dgex. That is clearly unsustainable and my guess is that the bter market got arbitraged down to the dgex price. Since the dgex price seemed to be stuck, this means the downward move from .00016 was not due to big seller, but arbitrage, or perceived arbitrage. The NXT/CNY value peaked at .85 and dropped to match the dgex price.
It seems the dgex price is unstuck now and this bounce at .00016 is the second time, regardless of the reason. It is not uncommon for a price to bounce three times before breaking out. I think it is good that it has come back down a bit as the rise was too extreme. Gather some support at .0001 and gradually get back to .00016, bounce back a bit and then finally breakthrough. That would be a sustainable pattern that would then turn .00016 from a price ceiling (resistance) to price floor (support).
Of course, this is crypto, so anything can happen, but I think there is a significant amount of fast money parked in NXT right now, so it could follow more standard technical patterns
James
I enjoy reading your analysis. Oh and btw...Frictionless has another project going on: http://cryptosource.org/ixcoin-a-very-old-bitcoin-clone-ready-to-skyrocket/He's trying to dethrone Bitcoin, Mastercoin and now NXT. Will he succeed? Stay tuned...
|
|
|
|
Come-from-Beyond
Legendary
Offline
Activity: 2142
Merit: 1010
Newbie
|
|
January 21, 2014, 05:19:58 PM |
|
He's busy with NEM. Question for CFB; Is a Reputation System being developed by JLC or You? (or plans to develop) I was going to offer a bounty for a working version, implemented using public AM.
Reputation System is not going to be implemented in foreseeable future.
|
|
|
|
chanc3r
|
|
January 21, 2014, 05:20:24 PM |
|
Yes, this would be the solution - if that alias of account number was locked and only the account itself can own it, then we can do this. If it is squattable, then it isn't possible to do.
CFB, what are your thoughts?
Squatting is allowed. Let's find another way. Of course it is - squatting is one of the value's of aliases... So lets not call it an alias (even if that might be its implementation) you have previously proposed locking 10 accounts in a special way for a specific reason.. So why not give all accounts a single new attribute which would support client/app/service development? Alias might be the basis of its implementation but you could call it what you like.
|
|
|
|
msin
Legendary
Offline
Activity: 1470
Merit: 1004
|
|
January 21, 2014, 05:20:42 PM |
|
He's busy with NEM. Question for CFB; Is a Reputation System being developed by JLC or You? (or plans to develop) I was going to offer a bounty for a working version, implemented using public AM.
Reputation System is not going to be implemented in foreseeable future. Thank you.
|
|
|
|
Come-from-Beyond
Legendary
Offline
Activity: 2142
Merit: 1010
Newbie
|
|
January 21, 2014, 05:20:45 PM |
|
How much work is it to benchmark all 4 on the same platform? I would like to know the TPS and I think it makes sense for there to be published performance and also test results. Since it is JS, doesn't that mean it is browser specific? Do all four run on all popular browsers?
James
I tested only signing in Chrome.
|
|
|
|
wesleyh
|
|
January 21, 2014, 05:25:43 PM |
|
Yes, this would be the solution - if that alias of account number was locked and only the account itself can own it, then we can do this. If it is squattable, then it isn't possible to do.
CFB, what are your thoughts?
Squatting is allowed. Let's find another way. Of course it is - squatting is one of the value's of aliases... So lets not call it an alias (even if that might be its implementation) you have previously proposed locking 10 accounts in a special way for a specific reason.. So why not give all accounts a single new attribute which would support client/app/service development? Alias might be the basis of its implementation but you could call it what you like. +1 -- just use a special format of alias, like suggested @ followed by account ID. It is my understanding that @ is not allowed at the moment, so the code could be changed so that only the account itself can create @ACCOUNT_ID.
|
|
|
|
Mistafreeze
|
|
January 21, 2014, 05:26:00 PM |
|
Yes, this would be the solution - if that alias of account number was locked and only the account itself can own it, then we can do this. If it is squattable, then it isn't possible to do.
CFB, what are your thoughts?
Squatting is allowed. Let's find another way. Of course it is - squatting is one of the value's of aliases... So lets not call it an alias (even if that might be its implementation) you have previously proposed locking 10 accounts in a special way for a specific reason.. So why not give all accounts a single new attribute which would support client/app/service development? Alias might be the basis of its implementation but you could call it what you like. +2
|
|
|
|
NxtChg
|
|
January 21, 2014, 05:28:24 PM |
|
Just some stats: here's how my 9000 unique IPs so far are distributed:
|
|
|
|
jl777
Legendary
Offline
Activity: 1176
Merit: 1134
|
|
January 21, 2014, 05:28:32 PM |
|
How much work is it to benchmark all 4 on the same platform? I would like to know the TPS and I think it makes sense for there to be published performance and also test results. Since it is JS, doesn't that mean it is browser specific? Do all four run on all popular browsers?
James
I tested only signing in Chrome. If these four will be reference implementations recommended for other developers to use, I think it would make sense for someone to spend the effort to benchmark and test on various browsers. All four implementations appear to be done in different ways, which means different performance and also potentially different compatibility. Would be bad PR if any of the reference code has significant compatibility issues. Maybe someone can offer a small bounty for this? I am about to go on the road again, so I won't be able to test this, but I am sure there are people eager to earn a bit of NXT. I don't even know how one would test this, so maybe someone that does could post testing instructions? James
|
|
|
|
Come-from-Beyond
Legendary
Offline
Activity: 2142
Merit: 1010
Newbie
|
|
January 21, 2014, 05:30:34 PM |
|
If these four will be reference implementations recommended for other developers to use, I think it would make sense for someone to spend the effort to benchmark and test on various browsers. All four implementations appear to be done in different ways, which means different performance and also potentially different compatibility.
Would be bad PR if any of the reference code has significant compatibility issues. Maybe someone can offer a small bounty for this? I am about to go on the road again, so I won't be able to test this, but I am sure there are people eager to earn a bit of NXT.
I don't even know how one would test this, so maybe someone that does could post testing instructions?
James
The other developers should do this work by themselves.
|
|
|
|
rriky92
|
|
January 21, 2014, 05:31:02 PM |
|
Why Only 1000 bytes for arbitrary message?
|
|
|
|
ferment
Full Member
Offline
Activity: 168
Merit: 100
IDEX - LIVE Real-time DEX
|
|
January 21, 2014, 05:31:38 PM |
|
|
|
|
|
Come-from-Beyond
Legendary
Offline
Activity: 2142
Merit: 1010
Newbie
|
|
January 21, 2014, 05:32:17 PM |
|
Why Only 1000 bytes for arbitrary message?
An arbitrary number was taken.
|
|
|
|
|