Sign a message in a p2sh segwit address.

[vit05]

Could we please make more public know that you can not signing a message using p2sh address. Apparently there are some wallets that accept or understand signatures using an address in this format. But if it is something restricted, it does not seem to be ideal. Looking at the topic on stake address, it is clear that many do not know about it. I think it would be important to be clear on this issue and the difficulty you may have in the future if you need to demonstrate ownership of that address in a simple and efficient way.

Explanation on GitHub issue #10542 https://github.com/bitcoin/bitcoin/issues/10542

[Sellingaccs]

If you are using at least a half decent wallet, you can sign and verify messages from bech32 (native segwit), N2PSH as well as legacy addresses. The problem is with user's lack of information rather than any wallet.

Signing is most probably the most easy thing to do...
It's up to user the user to educate themselves.

Also i suggest that you fix this part, as it gives the impression that you are not very informed either:

Could we please make more public know that you can not signing a message using p2sh address.

[nc50lc]

AFAIK The current version of Bitcoin core cannot sign any SegWit address.
Only Electrum and other 3rd party wallets can sign a message in a "nonstandard" way.

Means whenever the SegWit-signable update came, those signed message through Electrum can be invalid to the new standard method.
By that time, those 3rd party wallets might update to support the standard signing method of SegWit addresses making the old signed messages limited (verifiable) only to their platform.

It is widely known in the "Stake your Bitcoin Address" Thread.

In the contrary, Electrum's way could be standardized if core developers use its method (ThomasV is a core dev)

[bitmover]

AFAIK The current version of Bitcoin core cannot sign any SegWit address.
Only Electrum and other 3rd party wallets can sign a message in a "nonstandard" way.
Means whenever the SegWit-signable update came, those signed message through Electrum can be invalid to the new standard method.
..
It is widely known in the "Stake your Bitcoin Address" Thread.

I don't think this is widely known in the  "Stake your Bitcoin Address" Thread.

Just as Sellingaccs didn't knew about it, a great part of users here doesn't know.

People are making signed messages using electrum, but they cannot be verified using any other wallet, and the method used is non-standard. I doubt more than 10% of the users in that thread know that.

I agree with the OP, electrum should advert users about it. But I doubt they will change it, and probably even when Core make a standard signature verification Electrum will keep its own verification method, just like BIP 39, as they use a completely different method to generate the recovery phrase.

[Kakmakr]

AFAIK The current version of Bitcoin core cannot sign any SegWit address.
Only Electrum and other 3rd party wallets can sign a message in a "nonstandard" way.
Means whenever the SegWit-signable update came, those signed message through Electrum can be invalid to the new standard method.
..
It is widely known in the "Stake your Bitcoin Address" Thread.

I don't think this is widely known in the  "Stake your Bitcoin Address" Thread.

Just as Sellingaccs didn't knew about it, a great part of users here doesn't know.

People are making signed messages using electrum, but they cannot be verified using any other wallet, and the method used is non-standard. I doubt more than 10% of the users in that thread know that.

I agree with the OP, electrum should advert users about it. But I doubt they will change it, and probably even when Core make a standard signature verification Electrum will keep its own verification method, just like BIP 39, as they use a completely different method to generate the recovery phrase.

Are you sure about that? I signed an address with the Electrum wallet and the receiver verified that signature with a Brainwallet tool. So I thought this was sorted out. Is there something wrong with the way Electrum is doing this, because it is very important to verify the owner of the wallet, before the money is send.

Can this be manipulated?

[bob123]

Are you sure about that? I signed an address with the Electrum wallet and the receiver verified that signature with a Brainwallet tool. So I thought this was sorted out.

While it may be possible that eletrum and brainwallet are using 'the same way' to sign/check messages, there is no standard yet.
There are different approaches on how to sign/check messages with p2sh segwit addresses. It hasn't been 'agreed yet' on how to do this.

Is there something wrong with the way Electrum is doing this, because it is very important to verify the owner of the wallet, before the money is send.

Can this be manipulated?

It can't be manipulated.
If you sign a message in electrum it is verifable in electrum. But other tools would show your signature as invalid.

But there is no way for a signature being manipulated / faked.

[bitmover]

Are you sure about that? I signed an address with the Electrum wallet and the receiver verified that signature with a Brainwallet tool. So I thought this was sorted out. Is there something wrong with the way Electrum is doing this, because it is very important to verify the owner of the wallet, before the money is send.

Can this be manipulated?

Yes I am sure.
Nullius confirmed here in an old thread. Check it out.
https://bitcointalk.org/index.php?topic=2885058.msg29647827#msg29647827

I don't think it's wrong or manipulated, just non standard.  It's worthless.