MIME-Version: 1.0
In-Reply-To: <20170301223101.GA17022@savin.petertodd.org>
References: <20170224010943.GA29218@savin.petertodd.org>
	<CA+KqGkrOK76S3ffPJmpqYcBwtSeKESqN16yZsrwzDR6JZZmwFA@mail.gmail.com>
	<20170224025811.GA31911@savin.petertodd.org>
	<CA+KqGkq7gavAnAk-tcA+gxL2sWpv3ENhEmHrQHaPdyAsKrLjGg@mail.gmail.com>
	<20170224031531.GA32118@savin.petertodd.org>
	<CA+KqGkrfhg3GnbWwvKXHQ2NWuCnfzYyTPUxRhzYMuDBiNQR4eA@mail.gmail.com>
	<20170224043613.GA32502@savin.petertodd.org>
	<CA+KqGkpi4GvgU-K6vt-U5ZN4AkpjZ0rruzddoJS4-V0TcnyqUQ@mail.gmail.com>
	<20170225041202.GA11152@savin.petertodd.org>
	<CA+KqGkqs8F1hK6y-JnLFRpqhQ5i8i+MXVmtGUQBYmE5d1OCAAg@mail.gmail.com>
	<20170301223101.GA17022@savin.petertodd.org>
From: Bram Cohen <bram@bittorrent.com>
Date: Fri, 31 Mar 2017 13:38:16 -0700
Message-ID: <CA+KqGkoKy4spzi4mHBEfqXV8e9xwf8GA8oWN-Nqv7kNH_yPqpg@mail.gmail.com>
To: Peter Todd <pete@petertodd.org>
Content-Type: multipart/alternative; boundary=001a1144795418427b054c0ccada
Cc: Bitcoin Protocol Discussion <bitcoin-dev@lists.linuxfoundation.org>
Subject: Re: [bitcoin-dev] A Better MMR Definition
Precedence: list

--001a1144795418427b054c0ccada
Content-Type: text/plain; charset=UTF-8

On Wed, Mar 1, 2017 at 2:31 PM, Peter Todd <pete@petertodd.org> wrote:

>
> A better way to present your work would have been to at least explain that
> at
> the top of the file, and perhaps even better, split the reference
> implementation and optimized implementation into two separate files. If
> you did
> this, you'd be more likely to get others to review your work.
>

I've now added explanation to the README, reorganized the files, and added
some comments:

https://github.com/bramcohen/MerkleSet

In fact, I'd suggest that for things like edge cases, you test edge cases in
> separate unit tests that explain what edge cases you're trying to catch.
>

The tests work by doing a lot of exercising on pseudorandom data, an
approach which does a good job of hitting all the lines of code and edge
cases and requiring very little updating as the implementation changes, at
the expense of it taking a while for tests to run. The advantage of very
custom unit tests is that they run almost instantly, at the cost of
requiring painstaking maintenance and missing more stuff. I've come to
favor this approach in my old age.

The proportion of code devoted to tests is more than it looks like at first
blush, because all the audit methods are just for testing.

--001a1144795418427b054c0ccada
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div class=3D"gmail_extra"><div class=3D"gmail_quote">On W=
ed, Mar 1, 2017 at 2:31 PM, Peter Todd <span dir=3D"ltr">&lt;<a href=3D"mai=
lto:pete@petertodd.org" target=3D"_blank">pete@petertodd.org</a>&gt;</span>=
 wrote:<br><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.=
8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><br>A better w=
ay to present your work would have been to at least explain that at<br>
the top of the file, and perhaps even better, split the reference<br>
implementation and optimized implementation into two separate files. If you=
 did<br>
this, you&#39;d be more likely to get others to review your work.<br></bloc=
kquote><div><br></div><div>I&#39;ve now added explanation to the README, re=
organized the files, and added some comments:</div><div><br></div><div><a h=
ref=3D"https://github.com/bramcohen/MerkleSet">https://github.com/bramcohen=
/MerkleSet</a><br></div><div><br></div><blockquote class=3D"gmail_quote" st=
yle=3D"margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padd=
ing-left:1ex">In fact, I&#39;d suggest that for things like edge cases, you=
 test edge cases in<br>
separate unit tests that explain what edge cases you&#39;re trying to catch=
.<br></blockquote><div><br></div><div>The tests work by doing a lot of exer=
cising on pseudorandom data, an approach which does a good job of hitting a=
ll the lines of code and edge cases and requiring very little updating as t=
he implementation changes, at the expense of it taking a while for tests to=
 run. The advantage of very custom unit tests is that they run almost insta=
ntly, at the cost of requiring painstaking maintenance and missing more stu=
ff. I&#39;ve come to favor this approach in my old age.</div><div><br></div=
><div>The proportion of code devoted to tests is more than it looks like at=
 first blush, because all the audit methods are just for testing.</div><div=
><br></div></div></div></div>

--001a1144795418427b054c0ccada--