da2ce7
Legendary
Offline
Activity: 1222
Merit: 1016
Live and Let Live
|
|
December 31, 2012, 02:34:22 AM |
|
|
One off NP-Hard.
|
|
|
|
jim618 (OP)
Legendary
Offline
Activity: 1708
Merit: 1066
|
|
December 31, 2012, 11:48:35 AM |
|
Yeah that is probably the easiest.
Coming up in the next few months is more work centred around Wallet (and probably Transaction) so you'll probably get conflicts over time with those classes when you rebase against your upstream MultiBit and bitcoinj. As they are the 'centre' this is kind of inevitable but it's usually not too bad to sort it out.
I'll be interested to see what you create !
I normally track the bitcoinj head for the test code but trail it a bit for the production code (just to keep it off the bleeding edge).
|
|
|
|
jim618 (OP)
Legendary
Offline
Activity: 1708
Merit: 1066
|
|
January 01, 2013, 10:53:16 AM |
|
MultiBit uses the XChange library to get exchange rates. Currently only MtGox appears in the Preferences 'Exchange' combo box.
I have just been mailing Tim Molter and he, Gary and Matija Mazi have been working on adding: + Cavirtex + BitStamp + BTC-E
It will probably be a few weeks before the code is all debugged, added into XChange, released and then picked up by MultiBit but they are in the pipeline.
This will make the currency conversion less centralised on Mt Gox.
|
|
|
|
jim618 (OP)
Legendary
Offline
Activity: 1708
Merit: 1066
|
|
January 01, 2013, 10:48:49 PM |
|
I keep an eye on the multibit.org Alexa ranking: http://www.alexa.com/siteinfo/multibit.orgIts ranking is normally over 700,000 but lower in the US as that has always been the 'heartland' of Bitcoin (currently 338,000 in the US). Anyhow, looking further down that Alexa page it gives a ranking in Russia of 91,000. I am slightly amazed as I have never seen it ranked so low. Assuming that is real and not some Alexa glitch it would be nice to get the Russian translation complete. It is currently 93% done so is probably missing some of the most recent terms. If you are a native Russian speaker please have a look at http://translate.multibit.org and click on the Russian flag. Thanks.
|
|
|
|
giszmo
Legendary
Offline
Activity: 1862
Merit: 1105
WalletScrutiny.com
|
|
January 02, 2013, 11:54:34 AM |
|
Mulitbit being the best localized client and having some statistically relevant download figures I wonder if you could give us a chart of countries?
|
ɃɃWalletScrutiny.com | Is your wallet secure?(Methodology) WalletScrutiny checks if wallet builds are reproducible, a precondition for code audits to be of value. | ɃɃ |
|
|
|
jim618 (OP)
Legendary
Offline
Activity: 1708
Merit: 1066
|
|
January 02, 2013, 12:23:47 PM |
|
Hi Giszmo, That is a good idea. At the moment I don't have any proper server stats on the new server (just the server status which gives away too much system info to make public). It would be good yes to improve the stats by processing the httpd logs and make them available using one of the Apache log analysers. (Note: Your IP addresses won't be shown). I also need to add in a privacy policy and delete the logs after a set period too. It is a background task to improve the server offering so bear with me on this.
|
|
|
|
lenny_
Legendary
Offline
Activity: 1036
Merit: 1000
DARKNETMARKETS.COM
|
|
January 02, 2013, 02:28:20 PM |
|
It there any progress with wrong fee issue? Multibit still not working for me, it said "Seen by 1 peer. Not seen in blockchain" and my transactions actually never get out from my machine. I already described this problem few pages ago.
|
|
|
|
jim618 (OP)
Legendary
Offline
Activity: 1708
Merit: 1066
|
|
January 02, 2013, 04:13:49 PM Last edit: January 03, 2013, 10:17:30 AM by jim618 |
|
Hi Lenny,
Unfortunately no - better fees are a non-trivial bit of work and I plan to concentrate on encrypted wallets now until they are out the door.
I think your bitcoin has some small amounts that are aggregated together (from my investigations earlier). The current MultiBit is probably not the best client to use with these transactions for exactly the reasons you state. The transactions created will be too large for the fees added so they won't propagate through the network properly.
Pragmatically you are better off moving your bitcoin out into another client like blockchain.info or Electrum until the fees are calculated better. Come back to MultiBit when the fees are implemented properly and you can use it without any hassle.
You don't even need to send your bitcoin out as you can do an (unencrypted) private key export and import the file directly into blockchain.info. (Be sure to secure delete the unencrypted key export file afterwards).
Obviously I would prefer you to use MultiBit but being practical I think another alt client would make life easier for you.
|
|
|
|
giszmo
Legendary
Offline
Activity: 1862
Merit: 1105
WalletScrutiny.com
|
|
January 02, 2013, 05:31:59 PM |
|
Hi Lenny,
Unfortunately no - better fees are a non-trivial bit of work and I plan to concentrate on encrypted wallets now until they are out the door.
I think your bitcoin has some small amounts that are aggregated together (from my investigations earlier). The current MultiBit is probably not the best client to use with these transactions for exactly the reasons you state. The transactions created will be too large for the fees added so they won't propagate through the network properly.
Pragmatically you are better off moving your bitcoin out into another client like blockchain.info or Electrum until the fees are calculated better. Come back to MultiBit when the fees are implemented properly and you can use it without any hassle.
You don't even need to send your bitcoin out as you can do an (unencrypted) private key export and import the file directly into blockchain.info. (be sure to secure delete the unencrypted key export file).
Obviously I would prefer you to use MultiBit but being practical I think another alt client would make for an easier life for you.
Wouldn't it be possible to just export one priv key that is used in the non-spreading transaction to do one transaction with it via blockchain.info? Shouldn't multibit forget its own transaction when it sees a confirmed transaction that is in conflict with what it tries to send?
|
ɃɃWalletScrutiny.com | Is your wallet secure?(Methodology) WalletScrutiny checks if wallet builds are reproducible, a precondition for code audits to be of value. | ɃɃ |
|
|
|
jim618 (OP)
Legendary
Offline
Activity: 1708
Merit: 1066
|
|
January 02, 2013, 06:04:53 PM |
|
Hi Giszmo,
You may very well be right but normally I don't recommend having the same private keys in two active wallets. It is too easy to double spend and you start banging up against edge conditions that neither wallet (that has the duplicate private keys in) has been tested against.
It's normally better if you copy the private keys from wallet A to wallet B to then only use wallet B.
Also, for the distribution of transaction outputs I think Lenny has, MultiBit just isn't very good at dealing with at the moment. Most people just want to be able to press 'Send' and be sure the transaction will be sent ok. Hence my recommendation to use a client that is better with the fee calcs.
|
|
|
|
giszmo
Legendary
Offline
Activity: 1862
Merit: 1105
WalletScrutiny.com
|
|
January 02, 2013, 06:13:07 PM |
|
Hi Giszmo,
You may very well be right but normally I don't recommend having the same private keys in two active wallets. It is too easy to double spend and you start banging up against edge conditions that neither wallet (that has the private keys in) has been tested against.
It's normally better if you copy the private keys from wallet A to wallet B to only use wallet B.
Also, for the distribution of transaction outputs I think Lenny has, MultiBit just isn't very good at dealing with at the moment. Most people just want to be able to press 'Send' and be sure the transaction will be sent ok. Hence my recommendation to use a client that is better with the fee calcs.
I was actually more interested in whether Multibit would forget an "own double-spend" that way. Taken you know you won't reuse the key, it should also be safe to do it. Lastly, under no circumstances you should never loose coins from having any mix of copied keys. Just your book-keeping might get temporarily confused.
|
ɃɃWalletScrutiny.com | Is your wallet secure?(Methodology) WalletScrutiny checks if wallet builds are reproducible, a precondition for code audits to be of value. | ɃɃ |
|
|
|
jim618 (OP)
Legendary
Offline
Activity: 1708
Merit: 1066
|
|
January 02, 2013, 06:31:00 PM |
|
Hi Giszmo,
In my release testing I export the keys from one wallet twice (encrypted and unencrypted) and then reimport them into two new wallets. This is to test the import/export. It picks up the transactions in the blocks ok.
It's quite easy to experiment with moving private keys around but for the general user I don't recommend doing it as it is just too easy to get confused. For instance:
<complete rubbish on> Import Your Keys To A New Wallet And Double Your Bitcoin !!! <complete rubbish off>
Also, you can imagine more complicated scenarios such as: + Wallet A and Wallet B have identical private keys but on completely different platforms. + there is a spend from Wallet A that appears in a block on one fork and + there is another almost simultaneous spend of the same transaction outputs from Wallet B (on a completely different platform) that appears in a block on another fork.
What *should* happen is that both wallets follow their own fork (and hence the balances in the two wallets will be different until one fork dominates) and then the wallet that is reorged throws out the block with one of the similar transactions in and replays the blocks from the new fork. The balances should then match. This is all quite confusing for the user however so is best avoided.
|
|
|
|
jim618 (OP)
Legendary
Offline
Activity: 1708
Merit: 1066
|
|
January 03, 2013, 05:12:40 PM |
|
I have started on the work for: Encrypted wallets with a wallet master keyI've only just started it but there is a google doc with the protobuf wallet definition here: Bitcoinj Encrypted Private Key Wallets (version 4)(It is not finished yet) This is the low level structure of how the wallet is stored - once that has been completed, reviewed and agreed by the bitcoinj mailing list then I'll start coding up the wallet changes and the UI changes, as described in this document: MultiBit Wallet Improvements
|
|
|
|
jim618 (OP)
Legendary
Offline
Activity: 1708
Merit: 1066
|
|
January 03, 2013, 08:04:45 PM Last edit: January 04, 2013, 11:58:39 PM by jim618 |
|
I have added in a new look and feel, called "seaglass". It is a bit lighter and more pastel than the current ones. Screenshot: It will appear in the next releases.
|
|
|
|
lenny_
Legendary
Offline
Activity: 1036
Merit: 1000
DARKNETMARKETS.COM
|
|
January 03, 2013, 09:01:51 PM |
|
Can you please add field "Amount" in "Request" tab? So every address will have tab "Amount" with amount of BTC on it. Like in Blockchain Wallet.
|
|
|
|
jim618 (OP)
Legendary
Offline
Activity: 1708
Merit: 1066
|
|
January 03, 2013, 09:11:16 PM |
|
Do you mean:
Every unique address in the Request 'list of your addresses' has it's own amount which is stored ie remembered
?
|
|
|
|
lenny_
Legendary
Offline
Activity: 1036
Merit: 1000
DARKNETMARKETS.COM
|
|
January 03, 2013, 09:16:58 PM |
|
Do you mean:
Every unique address in the Request 'list of your addresses' has it's own amount which is stored ie remembered
?
Sorry for confusing you:) I mean Balance, not Amount. So: Every unique address in the Request 'list of your addresses' will have it's own balance displayed in new field "Balance". So there will be 3 fields: Label Address Balance
|
|
|
|
jim618 (OP)
Legendary
Offline
Activity: 1708
Merit: 1066
|
|
January 03, 2013, 09:23:14 PM |
|
It is already on my ToDo list.
I want to show the Balance per receiving address as part of the functionality to archive off old receiving addresses (like in blockchain.info)
Not that I'll get to doing that functionality soon but it is on the list! :-)
|
|
|
|
21after2
Member
Offline
Activity: 112
Merit: 16
|
|
January 04, 2013, 01:34:19 AM |
|
I know that encrypted wallets are a planned feature in future versions of MultiBit. Any idea when they'll be implemented?
|
|
|
|
|