Oh, clever and original...

..took a year


   You got donged!

The exchange data from localbitcoins is crazy looking:



MtGox may be interesting for historical reasons, but I think I'll remove it as a default the next time I need to make a big update.

I didn't just write a guide, I wrote the program.

https://bitcointalk.org/index.php?topic=361092.0

It's run by Nils Schneider, user tcatm on this forum. Many of the bitcoin domains were taken by satoshi, and then put into the hands of various administrators.

I just looked at https://bitcointalk.org/index.php?board=1.0 - the entire first page is mtgox trolling.


While it is interesting that money is no longer coming out of the oldest Bitcoin exchange, I think ALL that talk can be moved to some other child forum, as this Bitcoin user is unaffected.

Child forum "Currency exchange" in "Marketplace" is currently too crowded. That forum should be renamed "Bitcoin Exchanges", as the sole place where exchange service talk is not off-topic. A new child forum exclusively for person-to-person exchange requests and solicitation should be created.

Threads about limited withdraws, insolvency, and such exchange-specific topics are best sequestered to the currency exchange board, as they are mostly troll-bait.

One example where discussion would be acceptable outside of the currency exchange subforum would be the speculation subforum, if you are sharing your investment theories (or misinformation) about Mtgox and how their fate may affect pricing on that exchange or elsewhere.

The topic of transaction malleability, Bitcoin protocol, and how to not fall into the trap that Mtgox did would be suitable for service discussions or general Bitcoin. Creating a new thread when there are a thousand others on that topic will wear on the nerves of other members though.

A method I have just devised for a daily jackpot:

First block with UTC timestamp after specified time determines winner,
Users can supply or transaction ID generates their "random" number,
Lowest result of SHA256(block hash + user nonce) wins.

This still allows the house to cheat though, as they can enter their own drawing for free.

That would be my thinking; however the blockchain doesn't provide a statistically constant rate of blocks, because mining isn't constant. If the goal is to generate on average one "event" per day or per week, it will still take manual intervention if the Bitcoin hashrate is increasing or decreasing, or above or below the difficulty; the actual hashrate cannot be determined to make it fair.

In addition, a criteria that may happen on average once per week (such as block hash mod 1008 = 0) has a not-insignificant chance of taking over a month, which may make players restless.

A better solution would be a satoshi-dice style provable, where the random choosing method for transactions sent into the jackpot pool is transaction hash plus site's secret which is pre-published as a hash. For example, if one in 16 bets may win the whole jackpot, etc... If I need to explain more, you shouldn't be running a gambling site.

Deepceleron's Razor #8: If everybody brings more to a pot luck dinner than they can eat themselves, there will be leftovers.

The largest bandwidth one is likely to see besides a DDOS attempt is if a new user needs to download the whole blockchain, and randomly has picked you as the source of the blocks. Bitcoin doesn't "torrent". You are only an awesome node if you can provide these new blocks to a client at an upload rate of over 30-50MBit/s, correspondent to other's broadband download abilities.

Bitcoin isn't really that bandwidth-hungry once there is a full mesh of well-connected clients. Consider the ~7 blocks per hour - a node only has to get a new block from another upon receiving the announce. If you have 1000 nodes, with 999 connections each showing in the client, each node still only has to provide one copy of a block to the network on average, because there are only 999 "give me the block" requests from clients in total. A default block is less than 350kB.

Nobody cares about the status. Newbie, by name, no longer restricts anything you can do.

You cannot accelerate your "activity" rating any faster than to post at least once every two weeks.

We assume some version of Windows...

Install Python 2.7.6 (32 bit): http://python.org/download/

Below, download the 32 bit build of the python 2.7 package:
-install zope.interface : http://www.lfd.uci.edu/~gohlke/pythonlibs/#zope.interface
-install twisted : http://www.lfd.uci.edu/~gohlke/pythonlibs/#twisted

Download and save the pywallet.py script to the directory C:\python27 (just so we know where it is).

Press "windows key" + "R" for run, and paste C:\Python27\python.exe C:\Python27\pywallet.py --web


...I edited myself a version of pywallet where --web is the default and it launches a web browser with the URL for you.

The last post of that thread in the first post will probably be the most helpful.

Start Bitcoin, and google for "what is my IP address" and write it down.
24 hours later when Bitcoin has eight connections, google for "what is my IP address" and compare.

Likely you have crappy DSL using PPPoE that is forcing a new IP on you regularly.

You can't know how many bitcoins Mtgox has from independent investigation. However Mtgox used the blockchain to prove they weren't insolvent and didn't have all the coins on deposit stolen after the big hack in 2011, by announcing and transferring 424242 bitcoins. If they had the ability then to prove they weren't insolvent, I don't see why they wouldn't now of all times.

I'm quite sure that it was the Bitcoin blockchain that was mentioned, not a web site. I am also quite sure that upper case letters are not valid parts of a domain name.

Small correction - base32-human originated with Zooko:
http://zgp.org/pipermail/p2p-hackers/2002-November/000924.html

I don't know why encoding is being re-imagined - Bitcoin's Base58 is a natural evolution of previous human-readable encodings, includes an example checksum, and is already in code. It doesn't use groupings, specifically so a double-click selects the whole thing.

I see few scenarios where one would need to communicate an entire transaction id by voice where lower-case-only would be worth the byte cost. In situations where the other party only needs confirmation of a transaction ID, reciting the starting letters in either upper or lower would allow positive identification of the transaction. I agree that reading mixed upper and lower over a phone is painful though, just having done that for generated passwords and user IDs.

The randomly mixed in upper case characters in base58 seem to make good visual place holders; when reading all-lower, it is very easy to lose your place.


ntxid
txbtsogbjjuimfqas7sgkbaqqkjxygyixk3deuxmrm1uqte8nukemm6yxujjjzbr

Base58 (no container):
86xpwm8Q5zCGUD844C2WZeWEn8tMadWbAG3pxBmfBea6

The above could be wrapped as ASCII in a container, "BTCTX" + base58 of txid and any ECC scheme

That seems like a genius plan by MtGox, don't let anybody withdraw bitcoins and offer $100 for them. I doubt it is MtGox offering the money though - it is buyers willing to gamble that they will be able to withdraw the bitcoins they bought sometime in the future. From the price we can infer the odds are 1:4.

When you aren't able to instantly get back the bitcoins you have on deposit with any service, it is time to be very worried. The only legitimate reason to delay withdraw is if out-of-channel confirmation is required for security against hackers. MtGox lost all your bitcoins with their bad txid programming though. They also lost at least $5M USD from the US government seizing their bank and Dwolla accounts.

Electrum uses a deterministic wallet. All addresses it will ever create are based on the initial seed, using a cryptographic algorithm. Therefore, all addresses you will ever "create" can be recreated by a backup of just the seed.

I'll give you some Python code critiques first:

-import time is not used
-You have mixed tab and space indents. You should pick one, four spaces is preferred.
-YouHaveCamelCase defs and variables. Both should be lower case.
-Lines too long, 80 is PEP8, 120 is tolerable.
-Inconsistent white space, compare line 45 (OK) to 85 (none)


Then some Pythonic tips:
-Your programmatically seem to be fishing for how many balls/rolls should be entered, the exact number required to exceed N (Secp256k1 max) is:
rolls = int(math.ceil(math.log(N, base)))

-I was going to say to use base64 = 'ABCD...'; list(base64) to make the same list, but it looks like you don't even need to make it a list, a string will do just fine for your existing indexing. It might be easier for you to just use base64 builtin functions though - http://docs.python.org/2/library/base64.html


And hey, I made one too...
**Dice to Bitcoin private key generator**
>How many sides on your dice?:75
Need 42 rolls of 75-sided dice. (261.610385001 bits)

Number of dice rolls so far: 0, need 42 more.
Input some dice rolling results,separated by spaces:)
>g
Generating random *insecure* dice rolls for remainder

dice used:
[7, 36, 73, 35, 68, 45, 1, 32, 72, 61, 71, 38, 14, 63, 7]
[21, 42, 47, 4, 72, 16, 25, 38, 73, 16, 9, 9, 46, 5, 61]
[14, 13, 72, 19, 38, 31, 48, 20, 32, 53]

====== Generated Address Information ============
Bitcoin Address:
  1JumBSm2g8RTJCqB5uRHAJtb4H6EoA4wju
Private Key (Wallet Import Format):
 5HxBYZeFgp2mZZ1UboHqfuz3er9NjvYSpYUm5FEZh9EwJmgtVko
Private Key (Hexadecimal 64 char) :
 11edbc7fd8e518efa64622839ecb167b7d530f9f1fab9e94a2d1c234de8ac842
Public Key (Full 130 character):
 042fd723ba3185607230be41ddcd1c5fbf1dbca2b2fad4edd5e677174c4e7f99135f3b3298ef0c4
45c64091770600adb9b3da2f0d905cb74182e79623a6c2f3dc8
---------
Bitcoin Address (Compressed):
 17Gh5qrLfhhVctNSUJCqhHb5YDZLGLJHC7
Private Key (Wallet Import Format Compressed):
 KwpZYK5RpiaoggH6pbKKpkrni3uxxTKnxqwXXjpsJBNWgxYLwjtE
Public Key (Compressed):
 022fd723ba3185607230be41ddcd1c5fbf1dbca2b2fad4edd5e677174c4e7f9913
===================================================
(Press "Enter" for more dice keys, "C" and "Enter" to close.)