Out of curiosity, which altcoins?

What about the added requirement of a good CSPRNG source for every tx (not required by ECDSA), something which is more difficult to obtain in an isolated IoT environment w/o added (and unauditable) hardware support?

There's also extra non-volatile state which much be maintained to commit the tx, another weak suit of IoT devices in general, correct?

ECDSA isn't that hard, otherwise hardware wallets wouldn't exist....

Are you sure you're loading the correct wallet file (maybe you created more than one wallet)?

Are you sure you're using the correct 2FA entry in Google Authenticator?

Have you checked that both the current time and time zone in your phone are correctly set?

If all else fails, you can restore a wallet with your seed (which you do have, right?) to recover your funds. Note that the restored wallet will not be 2FA-enabled, so you'll probably want to create a new 2FA wallet and transfer your funds over to it.

Someone correct me if I'm wrong, but the Android security model doesn't (I believe) extend to files saved on a USB OTG stick. That means that any app could access the wallet-protobuf file.

Although Bitcoin Wallet for Android uses good key stretching for the PIN, it's still short enough to be fairly easy to brute-force by your average PC once stolen by a malicious app. So, not a good idea....

It looks like this was fixed 15 days ago in master here.

As a workaround, check that your default locale ends in ".utf8" by running "localectl status". If it already does, then I'm not sure what workaround might work (unless you modified the locale just for your user).

If it doesn't end in ".utf8", see here to change it.

Serious question: are you a troll, or innocently misinformed/ignorant?

In any case, what "Cereberus" has said in regards to privacy here is false. Please ignore him/her.

It's mostly compatible with Bither, breadwallet, Coinomi, Hive (defunct), Mycelium (Android and iOS), myTREZOR, and Wallet32 for Android, but there are various "gotchas".

It is incompatible with Armory, Bitcoin Core, Bitcoin Wallet (for Android), and Electrum.

If you have a specific wallet in mind, I can describe the minor incompatibilities, or you can look at this spreadsheet here for the gory details.

https://www.reddit.com/r/btc/comments/4kzrp0/multibit_is_being_sold_to_keepkey_llc/

Also, I should probably revise what I said earlier... the original MultiBit devs (who released version 0.3.0) certainly have my trust, but I don't know anything (good or bad) about the KeepKey folk. It certainly seems like they have a vested interest in keeping up with MultiBit's good reputation, though.

Mandatory developers' fees were removed in version 0.3.0 (and even before then, the devs did their best to maintain user privacy with their BRIT donation scheme).


No. Privacy is perhaps a bit better than many other desktop wallets because you need a password to simply view your current balance (many other wallets only require a password to initiate a transaction; addresses and balances for those are available w/o a password).

MultiBit HD also uses a significantly better (if perhaps imperfect) wallet encryption scheme; in particular its key stretching makes brute-forcing much more time consuming than MultiBit Classic's.


As you already noted, they certainly can via new releases (but not remotely). The old fee code was removed from MultiBit HD, and they have committed to not reintroducing it in the future. That doesn't mean they're technically incapable of doing so.

Given how long they've been involved in the Bitcoin community, I think it's safe to say they are among the most trustworthy of wallet developers IMHO, and I wouldn't worry about this possibility.

My bad, the tool I linked to appends the last element on its own, so you should enter a path of just m/0'/0.

Also, is there any chance you created the wallet with a beta version prior to the 0.1 public release? Versions earlier than 0.0.8-beta had a bug that would affect you....

Do you have the seed?

If so, you can use https://dcpos.github.io/bip39/ to generate your extended master private key (your "xprv") from your seed. (Follow the directions to use that tool offline before actually using it, though).

To verify, that tool can also generate your addresses. Select the BIP32 tab, and use a derivation path of m/0'/0/0 for your first address, m/0'/0/1 for your second, etc.

If you don't have your seed, you can extract it from your wallet file (assuming you have the password of course) with this: https://github.com/gurnec/decrypt_bitcoinj_seed.

That's not entirely true, notably for Electrum (and MultiBit Classic, and maybe one or two others with equally low key stretching).

I ran some quick-and-dirty numbers based on btcrecover's speed (which isn't really a brute-forcer, I'm certain a real brute-forcer would be a few times faster).

Assuming only lowercase letters, passwords of length nine are quite feasible for an individual to brute-force.
Add in digits, and then it drops to about eight characters long.
Add in uppercase letters, then it's seven characters. Still not too bad.

Take a character off, and it goes from feasible to downright cheap ($10 and around 24 hours or less on a single monster AWS EC2 instance).

In short, brute-forcing weaker (yet not trivial) passwords is possible, and none of this takes into account how a real brute-forcer would work, e.g. going off of good dictionaries w/good mutation rules.

Regardless, to OP: unless you're willing to try to find the password yourself (e.g. with btcrecover mentioned above), Dave has a good reputation and may be able to help (for a fee that is).

Take a look at the pseudocode in the second half of this response on StackExchange, it may answer your question.

I'm afraid I don't know enough about OpenCL to say which architecture is better... I'll give you what little I do have though.

These are P/s rates from a Bitcoin Classic/XT/Core wallet with an iteration count of 67908.

• 1x 560 Ti: 1070 P/s
• 2x 560 Ti: 2110 P/s
• 1x R9 390: 4470 P/s

(the test numbers for other wallet types are here)

It's an apples-to-oranges comparison, but I'd guess that it points to AMD architecture at being better for this... but it's just a wild guess.

For Classic/XT/Core, VRAM makes practically no difference (you just need 64 x --global-ws bytes). For Armory, it makes a big difference.


Edit: I just noticed an old thread by you.... if you have any specific questions, feel free to ask, but in the mean time:

By adding the "--delimiter *" option, "*" is now used as the delimiter between tokens, instead of spaces. This means you can use:

• %[ _] - a single space or underscore (note the space before the _, it may be hard to see)
• %0,1[ _] - 0 - 1 single spaces or underscores, e.g. a single space or underscore or nothing at all

Hope it helps.

Thanks for the update!

Hi, Jim.

I hate to be a pain, but do you (or someone at Bitcoin Solutions) plan to re-sign the Windows 0.5.18 binaries sometime in the future to fix this certificate expiration issue?

Like ladyada, who also just happens to be an apparent bitcoin supporter?

(not saying your wrong in this particular case, just that you never know....)

I stand corrected, thanks.

Well, if you think you may have made a mistake in recording your seed, you can give seedrecover a try (it's part of btcrecover): https://github.com/gurnec/btcrecover/blob/master/docs/Seedrecover_Quick_Start_Guide.md#seedrecoverpy

It's an open source seed recovery tool; it requires that you have a good estimate of what your seed is, and also that you know either your master public key, or an address generated from your wallet (preferably one that was generated shortly after your wallet's creation).

Sakarias-Corporation: Note that of these three, only Mycelium (edit: and also Electrum, my mistake) support multiple accounts (which you apparently make use of). GreenAddress.it also has TREZOR support, but I don't know if it supports multiple accounts with a TREZOR (and also it's an online multisig wallet, which may or may not be what you want).

Coinbase began supporting 2-of-3 multisig wallets about 8 months ago where you control 2 of the 3 keys. Unlike blockchain.info, there have been no known security failures (yet) with Coinbase wallets.