Vanity bitcoin addresses: a new way to keep your CPU busy

[SgtSpike]

So I've got a compiled bitcoin.exe using minGW according to this thread:  http://forum.bitcoin.org/?topic=5851.msg86700#msg86700

Now, how do I go about applying this patch?

[SgtSpike]

Sorry for bumping, but I just can't figure this out on my own.  I think the command should be something like this:

patch -p1 < vanity.patch

But I don't want to just start typing things in without knowing what I am doing...

[joan]

Also interested in this.
I think the patch was done against the subversion repo so it probably won't work…
Also, it looks like the sources have been reorganized since then.

[dserrano5]

I'm unable to find the option to attach files to posts, so I'm forced to copy paste this here:

Code:

diff -urpN orig/bitcoin-0.3.23/src/main.h bitcoin-0.3.23/src/main.h
--- orig/bitcoin-0.3.23/src/main.h	2011-06-12 01:17:13.000000000 +0200
+++ bitcoin-0.3.23/src/main.h	2011-06-26 00:16:53.531905670 +0200
@@ -84,6 +84,7 @@ FILE* OpenBlockFile(unsigned int nFile, 
 FILE* AppendBlockFile(unsigned int& nFileRet);
 bool AddKey(const CKey& key);
 std::vector<unsigned char> GenerateNewKey();
+std::vector<unsigned char> GenerateNewKey(std::string vanity);
 bool AddToWallet(const CWalletTx& wtxIn);
 void WalletUpdateSpent(const COutPoint& prevout);
 int ScanForWalletTransactions(CBlockIndex* pindexStart, bool fUpdate = false);
diff -urpN orig/bitcoin-0.3.23/src/rpc.cpp bitcoin-0.3.23/src/rpc.cpp
--- orig/bitcoin-0.3.23/src/rpc.cpp	2011-06-12 01:17:13.000000000 +0200
+++ bitcoin-0.3.23/src/rpc.cpp	2011-06-26 00:14:12.042318453 +0200
@@ -6,6 +6,7 @@
 #include "cryptopp/sha.h"
 #undef printf
 #include <boost/asio.hpp>
+#include <boost/xpressive/xpressive_dynamic.hpp>
 #include <boost/iostreams/concepts.hpp>
 #include <boost/iostreams/stream.hpp>
 #ifdef USE_SSL
@@ -309,19 +310,61 @@ Value getinfo(const Array& params, bool 
 }
 
 
+vector<unsigned char> GenerateNewKey(string vanity)
+{
+    using namespace boost::xpressive;
+    sregex re = sregex::compile(vanity);
+
+    RandAddSeedPerfmon();
+    unsigned int tried = 0;
+    for (;;)
+    {
+        CKey key;
+        key.MakeNewKey();
+        string strAddress = PubKeyToAddress(key.GetPubKey());
+        ++tried;
+        if (tried%100000 == 0) printf("Vanity key tried %u\n", tried);
+        if (!regex_search(strAddress, re)) continue;
+
+        if (!AddKey(key))
+            throw runtime_error("GenerateNewKey() : AddKey failed");
+        return key.GetPubKey();
+    }
+}
+
+void VanitySearchThread(void* parg)
+{
+    Array* a = (Array*)parg;
+    string vanity = (*a)[1].get_str();
+    string strAddress = PubKeyToAddress(GenerateNewKey(vanity));
+    string strLabel = (*a)[0].get_str();
+
+    SetAddressBookName(strAddress, strLabel);
+
+    delete a;
+}
+ 
 Value getnewaddress(const Array& params, bool fHelp)
 {
-    if (fHelp || params.size() > 1)
+    if (fHelp || params.size() > 2)
         throw runtime_error(
-            "getnewaddress [account]\n"
+            "getnewaddress [account] [vanity]\n"
             "Returns a new bitcoin address for receiving payments.  "
             "If [account] is specified (recommended), it is added to the address book "
-            "so payments received with the address will be credited to [account].");
+            "so payments received with the address will be credited to [account]."
+            "If [vanity] is specified, is less than 10 characters, and is all valid"
+            "base58 characters, then an address containing that string is generated.");
 
     // Parse the account first so we don't generate a key if there's an error
     string strAccount;
     if (params.size() > 0)
         strAccount = AccountFromValue(params[0]);
+    if (params.size() > 1)
+    {
+        Array* a = new Array(params);
+        CreateThread(VanitySearchThread, (void*)a);
+        return "...searching...";
+    }
 
     // Generate a new key that is added to wallet
     string strAddress = PubKeyToAddress(GetKeyFromKeyPool());

Should apply cleanly to bitcoin-0.3.23.

[SgtSpike]

Thanks!    What's the command to apply it to a build?

[cschmitz]

I think this is inherently a bad idea, especially coming from one of the core developers of bitcoin.
Throwing away part of the network just to appeal to random nerdery is a horrible idea to implement when in the end, the core developers should really worry about using the resources of the system as efficiently as possible, after all the infrastructural components of the system are the basis of everything.
People thought  ipv4 was sufficient once, people thought 512kb ram were sufficient, etc. Being wasteful with a core resource in IT, given its short but rich history, is just wrong on so many fronts, i wouldnt even know where to start.

[TiagoTiago]

Shhhh! If people start wasting their hash/s it will get easier for the rest of us to crack blocks! <.<

[unk]

I think this is inherently a bad idea, especially coming from one of the core developers of bitcoin.
Throwing away part of the network just to appeal to random nerdery is a horrible idea to implement when in the end, the core developers should really worry about using the resources of the system as efficiently as possible, after all the infrastructural components of the system are the basis of everything.
People thought  ipv4 was sufficient once, people thought 512kb ram were sufficient, etc. Being wasteful with a core resource in IT, given its short but rich history, is just wrong on so many fronts, i wouldnt even know where to start.

it's hard, intuitively, to reason about numbers as large as the keyspace of the implicated private keys. but resource inefficiency isn't even plausibly a concern in this space. it's not at all like ip addresses.

maybe a comparison will help: if the world population grew by a factor of a thousand and then every person lived for a thousand years and generated a thousand keys per second while they lived, there would still be no issue with 'wasting' keys.

if that's still counterintuitive, you can think about it like this: if this idea were a problem merely because a developer of the bitcoin client encourages it, then bitcoin would not be secure, as the idea does not need the support of any bitcoin developers. i can generate and throw away thousands (i still cannot do billions like bytecoin because my maths are rusty) of keys a second, and nobody can stop me from doing that. if i
(or even ten thousand people acting in the same way) could compromise addresses that way, then bitcoin would be worthless from the perspective of financial security.

[joan]

Should apply cleanly to bitcoin-0.3.23.

I admit I haven't tried, but I think this would not apply smoothly to what's currently in the git repo (which I think is what SgtSpike is using?). As mentioned, the sources have been reorganized.
GenerateNewKey() for example is now in the class CKeyStore, not global to main.h / rpc.cpp. This is where GenerateNewKey(std::string vanity) should be too.

I do have some kind of patch too but I haven't worked out the threading yet. So it just hangs until it finds a match, which is not very nice.
SgtSpike, I think you are under Windows, do you use TortoiseGit? It will make things easy when you want to create/apply patches.

[SgtSpike]

I use MinGW.  I know next to nothing about c++ or compiling (web programmer here), so I was hoping there would be an easy way to include a patch (or diff?) file when I run the compile command.  If I find something that works, I won't upgrade my client until I absolutely have to in the future.

[dserrano5]

Thanks!    What's the command to apply it to a build?

You need the patch utility. There are a couple of implementations for windows out there. Just unpack the bitcoin sources into a directory, download the patch, open a cmd window and type this:

Code:

> cd \path\to\bitcoin-sources
> patch -p1 < \path\to\vanity-0.3.23.patch

(Disclaimer: that's untested). Then compile as usual. I only compiled bitcoind (headless) since that's what I use anyway.

I think this is inherently a bad idea, especially coming from one of the core developers of bitcoin.

I understand he did it for the lulz and in no way he seriously endorses its usage. After all, you have to jump through a couple of hoops to get it working.

[marcus_of_augustus]

... it seems like intentionally creating identifiable addresses goes counter to security through anonymity from the get-go ...

[TheSeven]

YAY! Pooled bitcoin vanity address mining!

[unk]

... it seems like intentionally creating identifiable addresses goes counter to security through anonymity from the get-go ...

obviously, you wouldn't choose a 'vanity' address if you didn't want it to leak some information. but just because bitcoin provides anonymity doesn't mean everyone cares to be anonymous in every transaction.

for example, many people post in this forum under their real names and have bitcoin addresses in their signatures. that involves 'intentionally creating identifiable addresses' in much the same way.

[inh]

Someone mod poclbm (simplest python miner I could find) to use the GPU for address generation. The CPU is so slow!

[marcus_of_augustus]

... it seems like intentionally creating identifiable addresses goes counter to security through anonymity from the get-go ...

obviously, you wouldn't choose a 'vanity' address if you didn't want it to leak some information. but just because bitcoin provides anonymity doesn't mean everyone cares to be anonymous in every transaction.

for example, many people post in this forum under their real names and have bitcoin addresses in their signatures. that involves 'intentionally creating identifiable addresses' in much the same way.

I wonder how many who are doing this realise they are effectively tying the value of their bitcoins to their reputations?

[FreeMoney]

... it seems like intentionally creating identifiable addresses goes counter to security through anonymity from the get-go ...

obviously, you wouldn't choose a 'vanity' address if you didn't want it to leak some information. but just because bitcoin provides anonymity doesn't mean everyone cares to be anonymous in every transaction.

for example, many people post in this forum under their real names and have bitcoin addresses in their signatures. that involves 'intentionally creating identifiable addresses' in much the same way.

I wonder how many who are doing this realise they are effectively tying the value of their bitcoins to their reputations?

Only their donations are tied (with proper caution) to them and how could you get donations without tying in your rep?

[cschmitz]

maybe a comparison will help: if the world population grew by a factor of a thousand and then every person lived for a thousand years and generated a thousand keys per second while they lived, there would still be no issue with 'wasting' keys.

its not a math issue its a conceptual issue. also lol³ at your "thousand keys per second" analogy for a system that by definition wants to be around 2033.

[EricJ2190]

The point is, nothing extra (except electricity) is being consumed by generating vanity addresses. An address that is generated then discarded may as well have never been generated. The network never has any indication that this address ever existed.

[TiagoTiago]

If mining hardware was instead dedicated to generating new addresses; how long do you think would it take till someone stumbled on an existing address that had more BTC stored than what the person would have earned by mining?