[ANN] Bitcoin v0.3.23 release candidate available

[jgarzik]

There will be a lot of fee-related UI changes in the next version.  Matt C already has a 'better fee UI' pull request at github.

This version is just to triage the P2P network, until we get to that awesome, big next version with wallet encryption and other useful things.

The point of v0.3.23 is to release it fast, fast, fast...

So people: test, test, test! 

[LightRider]

Does switching from a newer version to a previous version mess up the wallet.dat file? Is it possible that the previous version might have not handled full precision transactions properly?

[jgarzik]

Does switching from a newer version to a previous version mess up the wallet.dat file?

Probably not, we've taken steps to make sure that works.  But if you jump back too far in time, that could create a problem.

Is it possible that the previous version might have not handled full precision transactions properly?

Define previous version.  All recent versions handle full precision correctly, though the UI might not display it or other cosmetic details like that.

[LightRider]

Mystery solved, I apologize for the confusion. Apparently the transaction took a very small fee that I could not account for until I saw the transaction page on Blockexplorer. Also, I neglected to look for it in the transaction details panel in the client. I feel very silly now. Thanks for your help.

[Mike Hearn]

Awesome turnaround guys! I tested the Mac version both with and without port forwarding, it was able to establish a connection to the network within a few seconds in both configurations (tried a few times). This was with UPnP off.

I also tried the UPnP option and it worked fine here.

[Dobrodav]

   Is there something done to prevent sending BTC on unknown address when backuping new wallet (without  100 new address created) - or we will again see something like -  " OMG  - i am followed instructions on backup, and my xxxx BTC  gone in unknown direction - where is nearest window to jump in ?"

   Maybe some steps to send back unclaimed tx`s back to sender in 6 month ? Just to avoid that unnecessary windows jumps.
   
   How about choose clearly one address to send (not all of us  buying drugs , your know, - at least not every time) .
   Easy consolidation of wallet amounts on one address still require some focus-pocus actions with fee free nodes ?
   Clear representation of BTC`s on each address in wallet ?
   Floating point operations ? - that is annoying question, but still important.

Anyway, i am believe you are done great job. Will test it.
 

[Maged]

  Is there something done to prevent sending BTC on unknown address when backuping new wallet (without  100 new address created) - or we will again see something like -  " OMG  - i am followed instructions on backup, and my xxxx BTC  gone in unknown direction - where is nearest window to jump in ?"

The keypool is now created immediately in this version. However, it could use some testing.

[Dayofswords]

no zip version?

[[Coins!]]

One thing that might make the display of the wallet less confusing (maybe) would be to show all the properly available decimal places.

1.0000000000
for example.

or 2.500000000
not just 2.5

Why truncate the display at all?

Thanks for your hard work!

[andrew_jacksun]

One thing that might make the display of the wallet less confusing (maybe) would be to show all the properly available decimal places.

1.0000000000
for example.

or 2.500000000
not just 2.5

Why truncate the display at all?

Thanks for your hard work!

ditto. and, why was the 'generate coins' option removed? will you bring it back or is that only for elitists who can understand setting up scripts?

[gigabytecoin]

One thing that might make the display of the wallet less confusing (maybe) would be to show all the properly available decimal places.

1.0000000000
for example.

or 2.500000000
not just 2.5

Why truncate the display at all?

Thanks for your hard work!

ditto. and, why was the 'generate coins' option removed? will you bring it back or is that only for elitists who can understand setting up scripts?

I somewhat agree. Theoretically a 10mhash CPU could generate 0.35 BTC over the next month if the difficulty stayed stagnant. It won't, but still that is a great way to distribute small amounts of currency to new interested users. Sure, flash a warning perhaps that it might deplete their CPUs lifespan and increase their electrical bill... but don't restrict them from the possibility that was once there. A lot of people don't want to have to install more than 1 program in the course of a their days. One is more than enough to frustrate many of my friends.

[foo]

ditto. and, why was the 'generate coins' option removed? will you bring it back or is that only for elitists who can understand setting up scripts?

I somewhat agree. Theoretically a 10mhash CPU could generate 0.35 BTC over the next month if the difficulty stayed stagnant. It won't, but still that is a great way to distribute small amounts of currency to new interested users. Sure, flash a warning perhaps that it might deplete their CPUs lifespan and increase their electrical bill... but don't restrict them from the possibility that was once there. A lot of people don't want to have to install more than 1 program in the course of a their days. One is more than enough to frustrate many of my friends.

FFS, just start the client with the -gen option then.

[Pieter Wuille]

If you really want to mine on a CPU, please use cpuminer or another specialized program that is efficient, optimized and supports pool mining. The miner left in the default client is only intended as a reference and for testing.

[jgarzik]

ditto. and, why was the 'generate coins' option removed? will you bring it back or is that only for elitists who can understand setting up scripts?

I somewhat agree. Theoretically a 10mhash CPU could generate 0.35 BTC over the next month if the difficulty stayed stagnant. It won't, but still that is a great way to distribute small amounts of currency to new interested users

Incorrect.  Besides being slower than other CPU miners, the built-in CPU miner has never supported pooled mining.  It only mines full blocks.  Thus, users either get zero BTC (highly likely) or 50 BTC (highly unlikely).

At current difficulty, you get a block once every 8 years, at 10 Mhash/sec.  http://www.alloscomp.com/bitcoin/old_calculator.php

[ShadowOfHarbringer]

But for example in 3.22 when trying to send a very low priority tx, you were forced to pay a fee. Do you now have the option to not send with a fee even though it will likely take a long time to be confirmed?

No, you have to hack the client in order to send spam.

In the future, people are considering a checkbox to avoid the recommended fee.

Was the spam-detection algorithm improved since 0.3.21 ?
In 0.3.21, even when i wanted to send some bitcoins having 7 confirmations, i had to pay the fee. Downgrade to 0.3.20 fixed the problem, and i could send the bitcoins (and later i got 2 confirmations after 45 minutes, so the transaction was OK contrary to what client claimed).

So is this finally fixed or not ?

[ShadowOfHarbringer]

Polite bump.

Let's ask this question differently: Was the algorithm which decides what is spam and what isn't changed at all since 0.3.21 ?

[Pieter Wuille]

Let's ask this question differently: Was the algorithm which decides what is spam and what isn't changed at all since 0.3.21 ?

No, but the minimum fee required in case it is considered "spam" was decreased to 0.0005 in 0.3.22 and to 0.0001 in 0.3.23.

[ShadowOfHarbringer]

Let's ask this question differently: Was the algorithm which decides what is spam and what isn't changed at all since 0.3.21 ?

No, but the minimum fee required in case it is considered "spam" was decreased to 0.0005 in 0.3.22 and to 0.0001 in 0.3.23.

Actually, i had the forced feee even when sending 10 BTC having 7 confirmations....
So i guess this wasn't fixed.