NXT :: descendant of Bitcoin - Updated Information

[jl777]

Does anybody know how long the recent fork on the testnet was?

If it was less than 10 blocks, then using 10 blocks for confirming transactions if fine, but if it was more, then 10 blocks is not enough...

This conclusion is not correct.

One does not draw such conclusion from 1 (in words ONE) single example.

[...]
Does anybody see the flaw in the following algo:

A) find the latest version of NRS (probably safer to hardcode this to avoid Evil Bob's tampering)
B) determine what fork the latest versions are on using method TBD
C) warning nodes that are not on the expected majority fork that they are not on the majority fork
D) Using above to generate overall NXT network fragmentation and adjust the number of confirmations needed based on this

Yes.
1) People should be encouraged to write their own node code. NRS is just one of them.
2) A) does not work out. Tampering is still possible.

Thanks for your incisive negative feedback! So it is your conclusion that since it is impossible to make sure it is perfectly accurate that there is absolutely no value in detecting potentially active forks?

What if the fork is caused by upgrading? What if Evil Bob isnt around? What if we could detect things are a bit unstable? There is no value in this?

James

P.S. I am sure everyone is rushing out to write their own NRS as we speak!!

[paradigmflux]

Can anyone in this thread who is mining in my pool please comment on https://bitcointalk.org/index.php?topic=522872.0 ?
Particularily if you're one of the people who have already recieved one of the payments.

[ChuckOne]

Thanks for your incisive negative feedback!

I had to look up incisive.

Feedback is feedback. Do with it what want AND do not evaluate it (first rule of feedback).

So it is your conclusion that since it is impossible to make sure it is perfectly accurate that there is absolutely no value in detecting potentially active forks?

No. There might be value in this. But your workflow is quite error-prone. I would like a simpler solution.

What if the fork is caused by upgrading? What if Evil Bob isnt around? What if we could detect things are a bit unstable? There is no value in this?

Quite some time ago, we used a migration strategy that activates feature F at block B. But distribution of F what at block B - X, where X >= say 20,000.

This way, people have two weeks to upgrade. We need to increase X in the future as the number of NXT nodes increases. More nodes will slow down upgrading.

[mczarnek]

P.S. For nodecoin haters, do you still think nodecoin is bad for NXT?

Is nodecoin good for NXT?

If it helps detect when there are significant forks in progress and especially if your node is on a minority fork, do you think that is good or bad?

Of course, more competition is bad but I still think Nxt has them beat.  Nxt is so much more than just a coin.

[jl777]

Thanks for your incisive negative feedback!

I had to look up incisive.

Feedback is feedback. Do with it what want AND do not evaluate it (first rule of feedback).

So it is your conclusion that since it is impossible to make sure it is perfectly accurate that there is absolutely no value in detecting potentially active forks?

No. There might be value in this. But your workflow is quite error-prone. I would like a simpler solution.

What if the fork is caused by upgrading? What if Evil Bob isnt around? What if we could detect things are a bit unstable? There is no value in this?

Quite some time ago, we used a migration strategy that activates feature F at block B. But distribution of F what at block B - X, where X >= say 20,000.

This way, people have two weeks to upgrade. We need to increase X in the future as the number of NXT nodes increases. More nodes will slow down upgrading.

I look forward to suggestions

[okaynow]

Hey guys, just a heads up, to any of you trading at Cryptostocks:

AlcurEX, a listing for an exchange traded there, has posted a list of coins to be voted by the shareholders and NXT is on the list.
If any of you are holding any AlcurEX shares, please vote for NXT.

Thank you
-Fo-

I did, although I don´t think that they will be able to add it in the beginning.

Thanks, i did so too.
Unfortunately, judging from the discussion in their thread today, my view is the same as yours.
i asked them to check the wiki for info, drop by from here, or pm the devs.
Hope they do, because on that list there are some serious jokes (pun intended)

[Jerical13]

I am not sure I understand what you mean by tagging. To me It would seem be more practical to implement if there were a set of "check the one that applies" categories that would actively limit the types of listings that could be used. Flexibility could be achieved by having a comprehensive set of categories to choose from, that could be added to if needed; and by allowing more detailed descriptions of the asset in the description field during the assets registration.

Tagging means to apply n tags to an object (this case: an asset).

For instance:

asset1: furniture, chair
asset2: table, furniture
asset3: furniture


People can easily filter for furniture, chairs or tables without loosing flexibility.

Many studies already showed that facet search (aka filtering) is the future and many webshops like ebay and amazon apply this way.

Hierarchical searches are not practical anymore. Best example: You want to buy wine for your best friend. You know he loves wine from Italy.

There is a shop offering wine. It applies a hierarchical search:

red
 - sweet
   - Italy
   - Portugal
 - semi-sweet
   - Portugal
 - dry
   - Portugal

white
 - sweet
   - Italy
   - Greece
   - Portugal
 - dry
   - Greece
   - Portugal

rosé
 - sweet
   - Greece
 - dry
   - Italy
   - Greece

How do you find what you need? You have no chance but to look it all through.

When using a tagging system the flexibility is built in. No categorization is better or more top than another.

Just my 2 cents.

I see what you mean. Like a text search where a user could enter a keyword or phrase. I think that is a good idea. Actually both  category indexing and your idea of tagging/ text searching would compliment each other. Categories would be better for broader searches and be more complicated for specific usage, and vice versa. I wonder if it would be possible to do your tag search by searching for key phrases or terms in the asset description, instead of creating a tagging application. Users could list there identifying terms in their description and a search application could identify these terms and index corresponding listings....possibly.

[L5Society]

(in words ONE)

This was unnecessary and dickish.

Don't be a dick (second rule of feedback)  

[Sebastien256]

@weshley
here are some more asset categories:
Various language of the asset description: e.g. english, french, spanish, etc....

That will come eventually.

[mczarnek]

Could someone here settle something for the paper I'm writing for the infrastructure committee?  https://bitbucket.org/nxtinfrastructure/committee/issue/19/nxt-energy-efficiency-paper-secondleo

Is forging easily split across parallel processors?

Seems to me like what the forging is doing is simply verifying individual transactions using Curve2259, therefore each transaction should be able to be easily handled in parallel.  Right? or am I missing something?

[EvilDave]

Since many here are sick of my posts, I've posted an "article" on http://reddit.com/r/nxt that addresses some of my concers about the Nxt ecosystem, mainly regarding fees, entitled "Nxt is the Bitcoin of the Future: Both in Good and Bad Ways".

http://www.reddit.com/r/NXT/comments/20u7ts/nxt_is_bitcoin_of_the_future_in_both_good_and_bad/

Just letting you folks know in case anyone wants to exercise their fee speech or present a rebuttal.

Hi Brian. Thanks for posting your concerns in detail. I found it easier to digest than trying to follow the back-and-forth nature of discussion here.

I agree that the fee thing needs to be sorted out with an eye towards future price appreciation. Unfortunately, I don't think there is a simple solution. Charging fees based on bandwidth usage or size of transaction seemed like an elegant solution when I first heard about it. However, as nxt intends to compete with a variety of financial systems, each with complicated fee structures of their own, I'm concerned that some great nxt services will be "priced out" of the market, while other services will be unnecessarily cheap.

When thinking of nxt as bitcoin in the future, I think the analogy breaks down when talking about fees. Specifically, the future of bitcoin assumes much lower price volatility. Therefore, denominating fees in bitcoin instead of fiat isn't a big deal. For us, we expect high volatility for the near term. Denominating fees in terms of fiat (or gold or any other asset with lower volatility) might make sense in the near term.

With regards to charging fees for placing an order in the asset exchange, I think that there is room for the nxt AE to improve upon the way markets currently work. If you follow nanexllc on twitter (his work has been featured in the Wall Street Journal, amongst other publications), you will learn that high frequency trading firms have been abusing the stock markets in the United States. They take profit, and justify this profit by claiming to increase liquidity and lower the bid-ask spread. However, in reality, the displayed liquidity (order book) can rarely be accessed by real trades, and orders often disappear before a counterparty can fill them. I know that setting walls is a strategy for day trading. However, for a wall to be effective, it has to be large. This implies that the fee would be tiny in comparison to the size of the wall.

Additionally, Nanex's recent studies show that the bid-ask spread has not tightened over the life of the current regulation (Reg NMS). Based on these issues in current stock markets, I am suspicious of HFT terrorizing our infant network, and I think that charging fees for placing an order prevents both spam and these predatory trading practices. We aren't designing the AE for the profit of algorithmic day traders.

My 2 cents on HFT: it's becoming very obvious that trading bots are being used to HFT the fuck out of some exchanges. Cryptsy trading is now becoming almost impossible for manual traders in popular markets. I spent 15 frustrating  minutes trying to place a buy order for LTC->NXT last night, got topped by 50 satoshis on every buy order I placed wihtin less than a second of placing the order. In the end, i just hodled the LTC, which made me cry a little this morning.

I started off not liking the idea of a fee on the AE for just placing an order, but HFT is taking predatory capitalism to new heights (or lows, depending on your point of view) and fees may well exclude these traders from the NXT AE, which is a good thing IMHO

[Mises_77]

Since many here are sick of my posts, I've posted an "article" on http://reddit.com/r/nxt that addresses some of my concers about the Nxt ecosystem, mainly regarding fees, entitled "Nxt is the Bitcoin of the Future: Both in Good and Bad Ways".

http://www.reddit.com/r/NXT/comments/20u7ts/nxt_is_bitcoin_of_the_future_in_both_good_and_bad/

Just letting you folks know in case anyone wants to exercise their fee speech or present a rebuttal.

Hi Brian. Thanks for posting your concerns in detail. I found it easier to digest than trying to follow the back-and-forth nature of discussion here.

I agree that the fee thing needs to be sorted out with an eye towards future price appreciation. Unfortunately, I don't think there is a simple solution. Charging fees based on bandwidth usage or size of transaction seemed like an elegant solution when I first heard about it. However, as nxt intends to compete with a variety of financial systems, each with complicated fee structures of their own, I'm concerned that some great nxt services will be "priced out" of the market, while other services will be unnecessarily cheap.

When thinking of nxt as bitcoin in the future, I think the analogy breaks down when talking about fees. Specifically, the future of bitcoin assumes much lower price volatility. Therefore, denominating fees in bitcoin instead of fiat isn't a big deal. For us, we expect high volatility for the near term. Denominating fees in terms of fiat (or gold or any other asset with lower volatility) might make sense in the near term.

With regards to charging fees for placing an order in the asset exchange, I think that there is room for the nxt AE to improve upon the way markets currently work. If you follow nanexllc on twitter (his work has been featured in the Wall Street Journal, amongst other publications), you will learn that high frequency trading firms have been abusing the stock markets in the United States. They take profit, and justify this profit by claiming to increase liquidity and lower the bid-ask spread. However, in reality, the displayed liquidity (order book) can rarely be accessed by real trades, and orders often disappear before a counterparty can fill them. I know that setting walls is a strategy for day trading. However, for a wall to be effective, it has to be large. This implies that the fee would be tiny in comparison to the size of the wall.

Additionally, Nanex's recent studies show that the bid-ask spread has not tightened over the life of the current regulation (Reg NMS). Based on these issues in current stock markets, I am suspicious of HFT terrorizing our infant network, and I think that charging fees for placing an order prevents both spam and these predatory trading practices. We aren't designing the AE for the profit of algorithmic day traders.

My 2 cents on HFT: it's becoming very obvious that trading bots are being used to HFT the fuck out of some exchanges. Cryptsy trading is now becoming almost impossible for manual traders in popular markets. I spent 15 frustrating  minutes trying to place a buy order for LTC->NXT last night, got topped by 50 satoshis on every buy order I placed wihtin less than a second of placing the order. In the end, i just hodled the LTC, which made me cry a little this morning.

I started off not liking the idea of a fee on the AE for just placing an order, but HFT is taking predatory capitalism to new heights (or lows, depending on your point of view) and fees may well exclude these traders from the NXT AE, which is a good thing IMHO

I have been trying to buy NXT today on Bter... annoying as hell. Still holding my BTC. I feel your pain brother!

[mczarnek]

My 2 cents on HFT: it's becoming very obvious that trading bots are being used to HFT the fuck out of some exchanges. Cryptsy trading is now becoming almost impossible for manual traders in popular markets. I spent 15 frustrating  minutes trying to place a buy order for LTC->NXT last night, got topped by 50 satoshis on every buy order I placed wihtin less than a second of placing the order. In the end, i just hodled the LTC, which made me cry a little this morning.

I started off not liking the idea of a fee on the AE for just placing an order, but HFT is taking predatory capitalism to new heights (or lows, depending on your point of view) and fees may well exclude these traders from the NXT AE, which is a good thing IMHO

hmm.. maybe we should charge a very small fee for placing an order and see if that prevents that, the normal transaction fee for it going through (maybe minus the initial fee?), as that does sound like an issue.

[Sebastien256]

My 2 cents on HFT: it's becoming very obvious that trading bots are being used to HFT the fuck out of some exchanges. Cryptsy trading is now becoming almost impossible for manual traders in popular markets. I spent 15 frustrating  minutes trying to place a buy order for LTC->NXT last night, got topped by 50 satoshis on every buy order I placed wihtin less than a second of placing the order. In the end, i just hodled the LTC, which made me cry a little this morning.

I started off not liking the idea of a fee on the AE for just placing an order, but HFT is taking predatory capitalism to new heights (or lows, depending on your point of view) and fees may well exclude these traders from the NXT AE, which is a good thing IMHO

hmm.. maybe we should charge a very small fee for placing an order and see if that prevents that, the normal transaction fee for it going through, as that does sound like an issue.

if it cost less to the blockchain to place an order than execute the order. I agree that it should be cheaper, otherwise, i don't agree for the moment.

[opticalcarrier]

Since many here are sick of my posts, I've posted an "article" on http://reddit.com/r/nxt that addresses some of my concers about the Nxt ecosystem, mainly regarding fees, entitled "Nxt is the Bitcoin of the Future: Both in Good and Bad Ways".

http://www.reddit.com/r/NXT/comments/20u7ts/nxt_is_bitcoin_of_the_future_in_both_good_and_bad/

Just letting you folks know in case anyone wants to exercise their fee speech or present a rebuttal.

Ive not noticed anyone say they were sick of your posts.  I read your writing and I 100% agree with it being ridiculous to try to adjust fees in client as NXT fluctuates in its value per fiat, which is why I've been saying we should determine how to untie fees from fiat.  But I must question your understanding of a blockchain though, if you are suggesting that we make it free for people to issue bid/ask orders on the AE.  You must remember that one of the key functions of fees are spam control.  Also, you are not considering the full functionality of AE - you seem to think its only a currency exchange market when really its an 'asset' market. Its not like there will be assets created on any of the centralized exchanges that people can operate on.  And later when/if the AT currency exchange portion comes, fees on it are just going to have to be implemented for orders.

The bottom line is that AE is not built for HFC, and the con to a decentralized exchange is fees per order instead of fees per % of successful trade.  Of course the reverse holds for centralized exchanges.  Different use - different purpose.

I still say that we should implement fees as low as possible to encourage utility.  With high usage that low fees can bring, forgers will make up the difference and it will be win/win for all involved.

[opticalcarrier]

Since many here are sick of my posts, I've posted an "article" on http://reddit.com/r/nxt that addresses some of my concers about the Nxt ecosystem, mainly regarding fees, entitled "Nxt is the Bitcoin of the Future: Both in Good and Bad Ways".

http://www.reddit.com/r/NXT/comments/20u7ts/nxt_is_bitcoin_of_the_future_in_both_good_and_bad/

Just letting you folks know in case anyone wants to exercise their fee speech or present a rebuttal.

make much sense what you say,
very good article.

Do You think about this proposal?

market of byte size on the blockchain Vs Nxt

I don't know if that will fix the issues i mentioned in the article. I'm not even sure what makes one transaction have more bandwidth than another unless you are talking about large amounts of text or whatnot.  I don't even want to get into this too much here because I'll probably be accused of throwing FUD. I'll be happy to field any questions on reddit though.

is someone threatening you in PM or something?  Im just not seeing any conflict in the thread.  When I refer to fee based on transaction size, I refer to size in blockchain, which is exactly linear with network bandwidth.  I have stopped responding to igmacas suggestions as I just cannot make heads nor tails from them; I cannot tell if there is just a translation issue or if he is just throwing out keywords and/or suggestions and hoping we will figure it all out based on his general suggestion, because he gives no method or model to his suggestions.

[bitcoinpaul]

Not sure about the tagging thing, I would just perhaps do some global categories only, like "currency", "commodities", etc... (and not allow the user to choose custom ones).

Hm, how would your client know that "fakecoin" or "fakeNXT" are currencies?

Asset issuers has to say it.

If it's not true, we could use service providers or something.

Service provider is 3rd party. Has an api that returns information about the assets, like: "scam", "ok", "not_checked", "incorrect_information", etc.. It's a centralized service, but there could be more than 1 service provider. (If anyone is interested in becoming a service provider that is).

Interesting concept. I think this could be service providers. Not baked into the core of Nxt or only doable by the committees. This screams credit rating agencies and all the implications with it, but I like it. In the client, the user has options to select lists from the different trusted service providers who rates assets.

I see these rating agencies getting very rich Of course, there could be rating service providers who don't accept any paying and are very trusted by the community, and thus frequently used by issuers.

The soft version of this would be the rating through the NXT holders via voting.

[bitcoinpaul]

Since many here are sick of my posts, ...

Mmmmhhh, others have different opinions. But sick of your posts? 

[Damelon]

Since many here are sick of my posts, I've posted an "article" on http://reddit.com/r/nxt that addresses some of my concers about the Nxt ecosystem, mainly regarding fees, entitled "Nxt is the Bitcoin of the Future: Both in Good and Bad Ways".

http://www.reddit.com/r/NXT/comments/20u7ts/nxt_is_bitcoin_of_the_future_in_both_good_and_bad/

Just letting you folks know in case anyone wants to exercise their fee speech or present a rebuttal.

Ive not noticed anyone say they were sick of your posts.  I read your writing and I 100% agree with it being ridiculous to try to adjust fees in client as NXT fluctuates in its value per fiat, which is why I've been saying we should determine how to untie fees from fiat.

I just want to add support for the untying idea, too.
Nxt should make sense in relation to itself. Trying to keep matching it to what happens with it in relation to outside entities will have us debating and jumping to keep up all the time.
If anything, it introduces extra variables that we *don't* control into the mix. Occam's Razor suggests we look to minimise the variables.

[mczarnek]

NxtBook
Actually not true about nxtbook, you will be able to store that much info eventually using Nxt's decentralized storage, basic idea is you pay people for storage space and then you store your data across multiple harddrives, so that if any one of them goes down, others are holding onto it and if someone goes down, you copy it to another machine from one of the ones still there.  So the idea is that you only need to store it on 3 or 4 machines at a time and you'd paid to keep it online.  The more you are willing to pay per byte, the more machines and therefore security we can offer.

Similiar to multi-sigs you wouldn't have to store the entire data on every machine, just enough so that 2 out of 5 machines could reconstruct it or whatever.

If people hook up 2 or 3 TB drives and allow Nxt to use them for the decentralized storage system and get paid for them, they could make money off of the process and we'd end up with a lot of available storage space.

This could be used to host such images and will eventually be implemented.  It would also allow us to provide an email service and store encrypted emails in a distributed manner, so google can't read your emails for example.  I think that decentralized storage should be a big priority once asset and decentralized exchanges are finished.  Especially because once we have decentralized storage working, then we can create the distributed computing option, and Nxt will become a supercomputer running across multiple forging machines.

p2p storage would be interesting. It would nice if something like Symform could be built on top of Nxt. Symform, is already offereing p2p storage:
http://www.symform.com/

Very interesting.. will have to further investigate, I see us running something very similar in the near future, except paying people for their storage space.

Going back to the idea of running a supercomputer on top of Nxt, I had a thought, internet bandwidth is going to continue steadily increase for the next few years to the point where most people should be able to easily run a desktop remotely. ( http://blogs.cisco.com/news/the-2017-internet-a-look-at-the-future-courtesy-of-the-cisco-vni-forecast/ ).  I already do sometimes when working remotely from home, I'll just log into my work computer and work from there.  

What if we tailored this decentralized computing to allowing people to buy essentially terminals that are bare-bones machines, probably $100 each, they plug into the Nxt network and essentially run their computers remotely paying by the hour or maybe more likely by the number of instructions executed?  That way people using more processor intensive processes would still be paying more and counting instruction cycles would probably prevent limit cheating.. idk maybe not.

We could probably begin with Linux and turn it into a decentralized operating system.  Then you shift people around between computers and different cores as needed, when only running a word processor someone could run it on one core of someone's machine and run a different person on a different core.  In fact every thread the program runs could potentially be run on different machines.

Still a baby idea and needs further development and analysis and would require a lot of work and steps along the way and may not be worth it but could be very cool and help Linux better compete with Windows.  I suspect once I finish with my current business idea I'll analyze this one further.