NXT :: descendant of Bitcoin - Updated Information

[TwinWinNerD]

Hey guys i need some help!

The reddit-tip-bot went opensource now and i need someone to check the code before i release the Bounty to the dev. Can someone please do this?

Please please

The source code resides in https://github.com/nxtip/nxtip and is GPLv2 licensed!

The bot is based on ALTCoinTip, with some extra files that emulate the bitcoind daemon behavior as far as ALTCoinTip is concerned, plus handle the deposits. See the src/ctb/nxtip* files.

Contact me, there might be a bounty for the code checking!


PUSH

[jkoil]

Code:

		for (w = 0, i = 0; i < 32; i++)
			w |= v[i] = tmp1[i];

Is my pascal version correct?

Code:

w:=0;
for i:=0 to 31 do
begin
  v[i]:= tmp1[i];
  w:=w OR v[i];
end;

What exactly is the "if (i==0) { i=0;}" supposed to do???

looks fine, but are you sure your Pascal compiler supports 0 as an array index?
(Can't remember if the starting index could be set ...)

What exactly is the "if (i==0) { i=0;}" supposed to do???

that looks like a place for a "conditional" breakpoint for debugging ...  unless someone had wanted to adjust some timing in the loop

Thanks for the review!

Yes, pascal arrays are 0-based. :-)

ok   apparently you have compiled it already.
I was just remembering that by default Pascal arrays are [1..n] but they can be set   [ 0 .. N ];
Once ported Pascal/C ... maybe it was Pascal to C and the arrays were used as 1..N. So then there was two different ways to be chosen in C.


About OR :
I guess the Pascal compilers are still case-insensitive. And they haven't invented different operators to logical/bitwise OR  and to Boolean OR
 

[rickyjames]

cross posted

Unclaimed coins fund forming commitees.

candidates for the commitees, please apply!

https://docs.google.com/spreadsheet/ccc?key=0AsrKGhOkVcFpdFRFeHNUdWx3Qy1nTjBPLUFNWUJKQkE&usp=sharing

thx

what are the planned scopes and responsibilities for the 3 groups?  seems to me that tech/infrastructure will overlap quite a bit

I think you're right.  Hopefully the techdev fund will support programming and infrastructure fund will support deploying permanent hardware nodes like the Odroid U3  - which I have been converted to from the Raspberry Pi as the low-cost solution for NXT.

http://hardkernel.com/main/products/prdt_info.php

[marcus03]

ok   apparently you have compiled it already.
I was just remembering that by deafult Pascal arrays are [1..n] but they can be set   [ 0 .. N ];
Once ported Pascal/C ... maybe it was Pascal to C and the arrays were used as 1..N. So then there was two different ways to be chosen in C.

In Pascal, characters in strings are 1 based, dynamic arrays are zero-bases and static arrays are based on whatever you defined them.

[Jean-Luc]

OK so as far as web browser javascript transaction generation and signing goes, does it matter where the end client sends their signed transaction to?  do they have to send it to a NRS process that has active forgers, or can they just send it to any NRS system with a synced blockchain?  Im asking because Im considering opening up API for all the 24 VPSs I manage on nxtcrypto.org and am also considering adding SSL certs for that domain into the Jetty for each server as well.

The node doesn't need to be forging. If you send a transaction to any node, it will broadcast it to others, and eventually a forging node will include it in a block.

[pinarello]

cross posted

Unclaimed coins fund forming commitees.

candidates for the commitees, please apply!

https://docs.google.com/spreadsheet/ccc?key=0AsrKGhOkVcFpdFRFeHNUdWx3Qy1nTjBPLUFNWUJKQkE&usp=sharing

thx

what are the planned scopes and responsibilities for the 3 groups?  seems to me that tech/infrastructure will overlap quite a bit

good question, TBD I think.

but you are a tech guy for sure  

edit: maybe one commitee would be a better solution?

[Come-from-Beyond]

According to Wiki TF decides the next forger dependent on the previous block. So, if there is a partition of the network N1 and N2, then the nodes in these two different partitions have different TF outputs.

So, there are two cases: (next blocks/accounts are C, C' (part of N1) and D (part of N2); TF1 and TF2 are the output of TF of N1 and N2 respectively)
1) TF1 says C; TF2 says D => branches will get longer as foreign blocks get rejected
2) TF1 says C; TF2 says C' => C will get accepted; C' will get rejected as N1 expects C => What does N2? I would say: penalize C', ignore C and move on on its branch.

Still don't see how this conflict can be resolved. I need some idea for that.

Nxt uses Bitcoin's consensus rule. How would Bitcoin resolve this conflict?

IIRC, each node analyzes the known graph and assumes the longest chain as the correct one.

The same in Nxt.

[jkoil]

ok   apparently you have compiled it already.
I was just remembering that by deafult Pascal arrays are [1..n] but they can be set   [ 0 .. N ];
Once ported Pascal/C ... maybe it was Pascal to C and the arrays were used as 1..N. So then there was two different ways to be chosen in C.

In Pascal, characters in strings are 1 based, dynamic arrays are zero-bases and static arrays are based on whatever you defined them.

ok, looks like that you have it ongoing already on the good rails

[Come-from-Beyond]

OK so as far as web browser javascript transaction generation and signing goes, does it matter where the end client sends their signed transaction to?  do they have to send it to a NRS process that has active forgers, or can they just send it to any NRS system with a synced blockchain?  Im asking because Im considering opening up API for all the 24 VPSs I manage on nxtcrypto.org and am also considering adding SSL certs for that domain into the Jetty for each server as well.

Any node can be used.

[opticalcarrier]

OK so as far as web browser javascript transaction generation and signing goes, does it matter where the end client sends their signed transaction to?  do they have to send it to a NRS process that has active forgers, or can they just send it to any NRS system with a synced blockchain?  Im asking because Im considering opening up API for all the 24 VPSs I manage on nxtcrypto.org and am also considering adding SSL certs for that domain into the Jetty for each server as well.

The node doesn't need to be forging. If you send a transaction to any node, it will broadcast it to others, and eventually a forging node will include it in a block.

ok, ill pursue getting SSL running on those VPSs, and opening up API for allowedBotHosts to * and leave allowedUserHosts nulled out to prevent remote forging, unless someone can think of a reason to open that up as well

IIRC, each node analyzes the known graph and assumes the longest chain as the correct one.

The same in Nxt.

Anon136's video described it as the chain that had the most stake involved in forging the blocks was the 'correct' one, even though there may be a longer chain.   

[Come-from-Beyond]

Anon136's video described it as the chain that had the most stake involved in forging the blocks was the 'correct' one, even though there may be a longer chain.   

It's slang. "Longer chain" == "chain with the most stake involved".

[rickyjames]

edit: maybe one commitee would be a better solution?

Ummmmmmmmmmmmmmmmmmmmmmmmmm ............ I don't think so.

[fmiboy]

UPDATE Clienxt GUI

version 0.0.6 [codename] Lite

Since last release, I got number of feedbacks, feature requests, bug reports. Thank you everyone for attention and using Clienxt GUI.

changelog:

* TestNet and Local - from this version on, user can change client from testnet to local and vice versa, simply by changing default values (Settings->Default settings) and restart client.
* Auto Update for Nxt - need some testing, With one click, check latest version, if yes, removes local nxt folder from /lib/ folder and downloads latest from download.nxtcrypto.org and unzips nxt folder in /lib/. Before update, please stop the server from Console window if you using Local version. after update is complete simple Restart server from Console window again. (from right click menu) In future, updates will keep database.
* Fixed Mac (OSX) cmd+c, cmd+v, cmd+x commands for all textfields (feature requested by swartzfeger and antanst)
* simple lightweight NXT Tray icon, minimized window will stay in System tray (near clock)! (feature requested by Mario123)
* Multi-thread support for all heavy operations
* Optimized transactions table (1 table has messages/assets/other transactions)
* Optimized assets table
* User can see his Assets in Overview window (account label, asset name, quantity)
* Fixed some Russian language words
* Support for German language (thanks to teletobi 11810049982317883239)
* Support for Portuguese language (thanks to gracias ...)
* Added support for Generate&Decode Tokens (in Unlock window)
* QR address generator (in Unlock window)
* Removed Watch Window and added option in simplified Unlock Window.
** Forging with multiple accounts by Unlocking them... (There is timer bug, yet to be fixed! And after restart need unlocking again working to get around this)
** Voting System (available features get All Polls, create Poll, Cast a Vote) more coming

NOTE: Voting System and Asset Exchange works only on testnet for now.

I know a lot changes in this release, but was going to wait until VS is completely released, currently it has two features create a poll and cast a vote, when more APIs are available it will implemented. Enjoy your testnet and public net use

------
Download
------
https://bitbucket.org/fmiboy/clienxt/downloads/clienxt_006.zip

------
Usage
------
Simply unzip archive and

- Windows users - double click Clienxt.jar and enjoy
- Linux&Mac users - in terminal ./Clienxt.run & and enjoy

------
Screenshots
------






------
Reports
------
report feature requests, bugs https://bitbucket.org/fmiboy/clienxt/issues?status=new&status=open


------
Support
------
Translators are very much welcome to contribute. PM here or fmiboy@mail.ru

Any donation is appreciated...
13792774143018875909

If you like work of Translators donate some
teletobi 11810049982317883239 (German)
garcias .#. (Portuguese)


Thank you for taking time to read this post and for using Clienxt GUI!

[Jean-Luc]

Question about the java API.
Are there any plans to extract the functionality of the handlers in in nxt.http into java interfaces.
My use case is to create a payment transaction but i dont want to go through the HTTP layer. Currently to send money from the java api you would have to duplicate all the steps taken in for instance nxt.http.SendMoney.

The Java API is still a work in progress, in fact I am just in the middle of another refactoring for 0.8.0. I am switching to embedded jetty, so Nxt will no longer be a servlet. I will also improve the API while at it.

For the SendMoney example, what kind of an API would you find useful? Right now the steps you would need to duplicate are not that many, basically you need to do:

Code:

        Transaction transaction = Transaction.newTransaction(Convert.getEpochTime(), deadline, publicKey,
                recipient, amount, fee, referencedTransaction);
        transaction.sign(secretPhrase);

        Blockchain.broadcast(transaction);

The rest is validating the user input and returning error messages, which is application specific. If you submit an invalid transaction, it will not be accepted, and a lot of validation is done immediately when the Transaction object is constructed (such as for non-negative amount and fee, deadline, timestamp). Other validations, such as the account having enough funds are only done when the Transaction is about to be included in a block (because the account balance may have changed in the meantime).

[Damelon]

New Nxt Clients - updated Press release:
https://docs.google.com/file/d/0BwAGADgnQcrtYW9aZDZLQ1R0MVE/edit

So community, what do you think now?

Nice document. You could however add some Offspring screenshots from here https://bitbucket.org/incentivetoken/offspring-public/wiki/Home.

thank you, so the version with Offspring:
https://docs.google.com/file/d/0BwAGADgnQcrtR3ZCdVRkNExZSVE/edit

Does it for me
Doing an article on Offspring today

[Jean-Luc]

Cool to see another web client, however it seems to be using api calls that will be removed soon (getInitialData / processNewData)

I will not remove those if other clients are relying on them. But those are not API calls, those are user interface calls. In 0.8.0 you would be able to selectively enable/disable the http API and the UI, so don't bundle them together. And they will be on different ports.

[Anon136]

NOTE: Voting System ... works only on testnet for now.

whats the voting algo like? 1 nxt 1 vote + tx fee?

[wesleyh]

Cool to see another web client, however it seems to be using api calls that will be removed soon (getInitialData / processNewData)

I will not remove those if other clients are relying on them. But those are not API calls, those are user interface calls. In 0.8.0 you would be able to selectively enable/disable the http API and the UI, so don't bundle them together. And they will be on different ports.

Hmm, I had this in my client first, but cfb told me it would be removed because it takes too much resources. Which is why I switched to the official API.

[CommanderVenus]

Are we stuck on block 69524 or is my server just not working correctly?

[Anon136]

Are we stuck on block 69524 or is my server just not working correctly?

im on 69526