MIME-Version: 1.0
From: =?UTF-8?B?Sm9yZ2UgVGltw7Nu?= <jtimon@jtimon.cc>
Date: Thu, 25 May 2017 21:53:29 +0200
Message-ID: <CABm2gDrZV-dwsXc++NL6T6GgBsuVUsC9iTsXOjTbbsnfa9btVw@mail.gmail.com>
To: Bitcoin Dev <bitcoin-dev@lists.linuxfoundation.org>
Content-Type: text/plain; charset="UTF-8"
Subject: [bitcoin-dev] Suggested changes to bip8
Precedence: list

Hi, I didn't want to comment on
https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2017-February/013643.html
because it seemed to me that thread was more broad.

I like bip8 very much as an extension to bip9, but I think it could be better.
With bip9, a bip9-ready node that sees a softfork activated that he is
not aware of will see a warning. See the implementation:

https://github.com/bitcoin/bitcoin/blob/master/src/validation.cpp#L1832

But with bip8, if a deployment is made at the end of the period
instead of through 95% signaling, nodes that implement bip8 but don't
implement a certain deployment that is activated can't receive such a
warning.

The solution that comes to mind is to reserve one of the nVersion for
the specific purpose of requiring that the bit is active for one block
when a deployment is locked in in this way (or maybe also when it's
activated with miners' signaling too, maybe that can be used to
simplify the way the current warnings are checked).

I expect the code changes to do this to be simple, and I'm happy to
help with it.