Topic: Bech32 new address format and Armory (Read 713 times)
Okay. Any solution for 0.96.x would probably be temporary. 0.97 uses libbtc, with part of the idea being that it and a bit of as-yet-unsubmitted code will replace Crypto++ (Armory uses an ancient, customized version) over time. (Maybe 0.98???) It has Bech32 support. So, I would either need to backport libbtc to 0.96.5 (doable but might require Git-fu a bit above my pay grade if it's going to be clean), or backport the separate bech32 subdirectory in 0.97 back to 0.96.5 ("wasted" work but the repo will be clean in the end, unless I figure out the libbtc backport). I'll converse with goatpig. We'll see what we can work out.
Thanks guys, I'm probably not the only one who'd like this to happen 
I'll take a look when I can. Might be next week. I'm trying to finish up BIP 150 this week.
Quote
Indeed, I guess it's a case of whether it's worth it considering the timescales
If droark wants to summon his inner git-fu to somehow sidestep the conflict around subtreeing the bech32 code into testing, the rest is easy for me. Otherwise, the issue rests with how unpredictable the conflict will be.
the concern is the conflict it will create when dev is ready to be merged back into testing.
Indeed, I guess it's a case of whether it's worth it considering the timescales (0.96.5 maybe in a few weeks/months, 0.97 maybe winter/spring 2019??)
Implementing sending is actually really easy, the concern is the conflict it will create when dev is ready to be merged back into testing.
Thanks for entertaining the idea. But I can see from you're previous updates about 0.97 features that it will be an important Armory release because of your new business agreement, this question was more idle curiosity on a Sunday than anything else! Please don't waste your time if it's too awkward to implement.
Please don't waste your time if it's too awkward to implement.
I have not thought about it that way at all, I'll have to evaluate once I'm done with the fixes.
goatpig will have to answer that one. I'm not quite familiar enough with the codebase to know if a couple of lines of code would allow a functional hack or if it's more involved.
Alas, bech32 would be too much work for 0.96.5.
Not even allowing sending? (I realise bech32 addresses in wallets is a significant change whether it's the mirror type or a new design) Just error checking without correction would be a useful win in summer/fall 2018 for sending to bech32 in different wallet software (clightning, electrum, core etc all support bech32).
Most likely I'll add support for this in mirror wallets before the new wallets are out.
Updated thinking on this?
It's getting to the stage now where I'm wanting to use bech32, but Armory 96.4 rejects sending to bech32 addresses. Are we going to wait till 0.97 for the ability to send to bech32 addresses, or could it make it into 96.5? (assuming 96.5 is still planned)
0.96.5 is still being planned. We're just busy dealing with 0.97 (it changes a *lot* of the underlying code). Alas, bech32 would be too much work for 0.96.5. The code is already in 0.97. I don't know if it's used yet but it is there.
Hang in there! We haven't forgotten about everyone.
We just have a lot to do to get 0.97 ready, and not enough bodies to juggle that and other issues. Most likely I'll add support for this in mirror wallets before the new wallets are out.
Updated thinking on this?
It's getting to the stage now where I'm wanting to use bech32, but Armory 96.4 rejects sending to bech32 addresses. Are we going to wait till 0.97 for the ability to send to bech32 addresses, or could it make it into 96.5? (assuming 96.5 is still planned)
Is it possible to send coins to a bech32 address with 0.96.1 + Core 0.14? If not, is it possible with any other combination of Armory and Core and which one?
Thank you!
Thank you!
You cannot get bech32 addresses from Armory atm. This will be part of 0.97, the next major release. The code is in dev right now, but there it hasn't been plugged in the GUI yet, and dev is unstable anyways.
Hey everyone,
a new Bitcoin address format, bech32, was just merged into bitcoin-core:
https://www.reddit.com/r/Bitcoin/comments/739pjb/bech32_address_format_merged_into_bitcoin_core/
In short, an address now looks like
a new Bitcoin address format, bech32, was just merged into bitcoin-core:
https://www.reddit.com/r/Bitcoin/comments/739pjb/bech32_address_format_merged_into_bitcoin_core/
In short, an address now looks like
Quote
bc1qrp33g0q5c5txsp9arysrx4k6zdkfs4nce4xj0gdcccefvpysxf3qccfmv3
with no case sensitivity any more, different first letters (bc1) and stronger checksumming/error detection (it now detects more spelling errors, and tells you where they are).Is it possible to send coins to a bech32 address with 0.96.1 + Core 0.14? If not, is it possible with any other combination of Armory and Core and which one?
Thank you!
AFAIK bech32 is only for v0 native SW scripts. I'll get it in there at some point.
bech32 works for all segwit versions. It encodes the witness version number.I meant it as a way to distinguish from nested segwit scripts.
AFAIK bech32 is only for v0 native SW scripts. I'll get it in there at some point.
bech32 works for all segwit versions. It encodes the witness version number. Good to hear, thanks.
Would you recommend creating new wallets for this? And/or will this coincide with the migration to the new wallet format?
Ente
Would you recommend creating new wallets for this? And/or will this coincide with the migration to the new wallet format?
Ente
Most likely I'll add support for this in mirror wallets before the new wallets are out.
Good to hear, thanks.
Would you recommend creating new wallets for this? And/or will this coincide with the migration to the new wallet format?
Ente
Would you recommend creating new wallets for this? And/or will this coincide with the migration to the new wallet format?
Ente
AFAIK bech32 is only for v0 native SW scripts. I'll get it in there at some point.
Hey everyone,
a new Bitcoin address format, bech32, was just merged into bitcoin-core:
https://www.reddit.com/r/Bitcoin/comments/739pjb/bech32_address_format_merged_into_bitcoin_core/
In short, an address now looks like
Exciting, I'd love to switch to that address format.
I suppose it's just a different representation of the same keys? Like, old wallets would just show the old addresses as a different string?
Is this on the radar of Armory? Are there plans to support this?
Cheers,
Ente
a new Bitcoin address format, bech32, was just merged into bitcoin-core:
https://www.reddit.com/r/Bitcoin/comments/739pjb/bech32_address_format_merged_into_bitcoin_core/
In short, an address now looks like
Quote
bc1qrp33g0q5c5txsp9arysrx4k6zdkfs4nce4xj0gdcccefvpysxf3qccfmv3
with no case sensitivity any more, different first letters (bc1) and stronger checksumming/error detection (it now detects more spelling errors, and tells you where they are).Exciting, I'd love to switch to that address format.
I suppose it's just a different representation of the same keys? Like, old wallets would just show the old addresses as a different string?
Is this on the radar of Armory? Are there plans to support this?
Cheers,
Ente
Jump to: