Testing BitShares XT Launch...

[BTSdac]

Does test network still works?

Btw with latest changes i cannot import pts wallet anymore (its encrypted with pass)

New test network with full DPOS + P2P connections will be out next week.

happy  to hear this news  and  hope it is a sucess one

[xeroc]

I am fine with command line

[muse-umum]

New test network with full DPOS + P2P connections will be out next week.

It seems that the web gui is still not ready. How can we vote without gui？Command line？

[alt]

New test network with full DPOS + P2P connections will be out next week.

good news!

[xeroc]

Holy moly ... wohooo

[bytemaster]

Does test network still works?

Btw with latest changes i cannot import pts wallet anymore (its encrypted with pass)

New test network with full DPOS + P2P connections will be out next week.

[alexxy]

Does test network still works?

Btw with latest changes i cannot import pts wallet anymore (its encrypted with pass)

[vikram]

I just added some extra code to save the wallet on Ctrl+C. 

The key generation algorithm is deterministic so you you can always regenerate your change addresses.

can your exception handling cater for the scenarios that the client crashed or computer power off suddenly?

power failure cannot be handled. Only way to is to sync wallet changes immiediently on disk

It should be saving after every major wallet operation...   One thing that I have realized is that we need to switch to using a database for the wallet so that we can do partial saves with proper logging without the performance hit of having to encrypt / save the entire wallet.

Now automatically saves after every command that would require the wallet to be unlocked: https://github.com/BitShares/bitshares_toolkit/commit/1300929e64bed39f24e4f137bc2171dda9af2758

We may want to save on other big operations as well, e.g. rescan.

[alexxy]

where is the sources for new wallet?

[sudo]

 

the bts wallet downloaded from“http://bitshares.org/downloads/” Windows: May 9, 0:21:21 bts_wallet.tar.gz
seems like it's different with what you compiled &tested？
the command  such as getinfo  is not existed

D:\bts>bts.exe data
================================================================
=                                                              =
=             Welcome to BitShares XT                          =
=                                                              =
=  This software is in alpha testing and is not suitable for   =
=  real monetary transactions or trading.  Use at your own     =
=  risk.                                                       =
=                                                              =
=  Type 'help' for usage information.                          =
================================================================
listening for rpc connections on 127.0.0.1:1666
connecting to bitshares network: 107.170.30.182:8765
connected to bitshares network
help
Commands:
 quit
 login  - enter your password to load your wallet file
 unlock - enter your password to unlock your private keys
 lock   - lock your private keys
 importkey PRIV_KEY [label] [rescan]
 importwalletkey PRIV_KEY [label] [rescan]
 import_bitcoin_wallet WALLET_DAT - load a bitcoin-qt or PTS wallet.
 balance         -  print the wallet balances
 newaddr [label] -  print a new wallet address
 listaddr        - print the wallet address(es)
 transfer AMOUNT UNIT to ADDRESS
 buy AMOUNT UNIT
 sell AMOUNT UNIT
 short AMOUNT UNIT
 cover AMOUNT UNIT
 add margin AMOUNT bts to UNIT
 cancel ID IDX
 html FILE
 json FILE
 market QUOTE BASE
 show orders QUOTE BASE

[vikram]

Change the bts total share to 80 trillion for what?

The exact value of the constant isn't particularly important, as your assets are best measured in bips. The share supply just needs to be scaled so that it is sufficiently divisible, but not overly divisible that it requires excessive storage. Bitcoin has 2.1 quadrillion shares; we are currently using 80 trillion.

[HackFisher]

4) We have identified a challenge in paying delegates... micro-payments from the transaction fees are 'unspendable' by normal means because the fees are higher than the output amount.   

Why not a fee market between delegate and users?

[cgafeng]

Change the bts total share to 80 trillion for what?

[sudo]

This test was very helpful and we learned some critical details:

1) We learned our wallet data storage (flat-file JSON) is subject to corruption and catastrophic failure despite efforts to write a backup, move it into place, etc.  It also didn't scale well.   So I have created a new wallet that uses a database for managing the data.

2) Internal testing revealed some network-layer issues that we have been fixing and are now mostly fixed.

3) We are doing some studies now on network propagation delay to make sure that 30 second block intervals are safe.

4) We have identified a challenge in paying delegates... micro-payments from the transaction fees are 'unspendable' by normal means because the fees are higher than the output amount.   

5) While working to solve these issues in the background I was also thinking about how we were going to handle fees paid in BitUSD, etc.   

6) The trustee client crashed and was a single point of failure.  We are working on approaches to help our trustee (and delegates) rapidly recover from failure to keep the network running without delay or manual intervention.

Overall I am pleased to say that we have an updated framework in place that should be able to handle all of our chain forks and actually enable full up DPOS with delegate signing rather than relying on a trustee.   When integrated with the crash recovery code should provide a much better experience.

 

[bytemaster]

This test was very helpful and we learned some critical details:

1) We learned our wallet data storage (flat-file JSON) is subject to corruption and catastrophic failure despite efforts to write a backup, move it into place, etc.  It also didn't scale well.   So I have created a new wallet that uses a database for managing the data.

2) Internal testing revealed some network-layer issues that we have been fixing and are now mostly fixed.

3) We are doing some studies now on network propagation delay to make sure that 30 second block intervals are safe.

4) We have identified a challenge in paying delegates... micro-payments from the transaction fees are 'unspendable' by normal means because the fees are higher than the output amount.   

5) While working to solve these issues in the background I was also thinking about how we were going to handle fees paid in BitUSD, etc.   

6) The trustee client crashed and was a single point of failure.  We are working on approaches to help our trustee (and delegates) rapidly recover from failure to keep the network running without delay or manual intervention.

Overall I am pleased to say that we have an updated framework in place that should be able to handle all of our chain forks and actually enable full up DPOS with delegate signing rather than relying on a trustee.   When integrated with the crash recovery code should provide a much better experience.