[Bitcoin-development] A critique of bitcoin open source community

Gregory Maxwell gmaxwell at gmail.com
Sat Oct 19 20:40:10 UTC 2013


On Sat, Oct 19, 2013 at 9:38 AM, Mitar <mmitar at gmail.com> wrote:
> Hi!
> Interesting read:
> http://courses.ischool.berkeley.edu/i290m-ocpp/site/article/nmerrill-assign3.html

Hopefully Nick will show up someplace and offer some specific pointers
to where we failed him.

The only interaction I can find from him on IRC is in #bitcoin, rather
than #bitcoin-dev:

--- Day changed Mon Sep 16 2013
11:45 < csmpls> Hi, I'm interested in contributing to the official
bitcoin project. Is there a mailing list I can join?
11:46 < neo2> csmpls, contributing how?
11:47 < csmpls> neo2 - probably start by approaching a low priority
issue like this one https://github.com/bitcoin/bitcoin/issues/2545
11:48 < michagogo> csmpls: There *is* a mailing list
11:48 < michagogo> ;;google bitcoin-dev mailing list
11:48 <@gribble> SourceForge.net: Bitcoin: bitcoin-development:
<http://sourceforge.net/mailarchive/forum.php?forum_name=bitcoin-development>;
Bitcoin-development Info
11:48 < csmpls> Great, thanks.
11:48 < michagogo> I don't know how active it is, though
11:49 < michagogo> There's also the #bitcoin-dev channel

I got involved with Bitcoin without previously interacting with other
contributors (AFAIK) and maybe things have changed in ways invisibly
to me. But I don't think so. Michagogo, who was answering there, is a
newer participant and I don't think anyone knows him from anywhere.
Certainly if things have become less welcome to new participants that
would be bad.

I can point out a number of other recent contributors who, as far as I
can tell, just showed up and stared contributing.  But I don't think
that the existence of exceptions is sufficiently strong evidence that
there isn't a problem.

The specific complaints I can extract from that article are:

"I wasn't even allowed to edit the wiki"

I'm confused about this, if he's referring to en.bitcoin.it.  Editing
it is open to anyone who is willing to pay the 0.01
(https://en.bitcoin.it/wiki/BitcoinPayment) anti-spam fee. This isn't
a policy set by the bitcoin development community, though I'm not sure
that its a terrible one. I've both paid it on behalf of other users
and made edits on behalf of people who didn't want to go to it.  At
least relative to some policy which requires actual approval the
payment antispam is at least open to anyone with Bitcoin.

"My IRC questions about issues on the github page were never answered"

Without a nick I'm unable to find more than the above, unfortunately.
So I don't yet know what we need to improve there.

"#bitcoin-dev would rather talk about conspiracies, or about
destroying other cryptocurrencies"

I've been pretty aggressive about punting out offtopic conversation
from #bitcoin-dev lately. Enough that I worried that my actions would
be the inspiration for this complaint. Much of the time discussion
like that is brought in and primarily continued by people who are not
active in the development community at all, but deflecting it to other
challenge without creating a hostile environment (or one that merely
feels hostile to new people) is hard.  Nicks comments themselves may
be a useful thing for me to show to people in the future on that
point.

"Bitcoiners are a bunch of paranoid, anti-authoritarian nutjobs"

I actually don't think that this stereotype accurately reflects the
development community. (In fact, I personally enjoy the great sport of
being called a statist by some of these aformentioned jutjobs, but
none of them are developers). On his other article Nick also asserts
"Most contributors hide their identities", but this is factually
untrue as far as I can tell. (In that same article he writes,
"Bitcoin's core code is written in Typescript, which is compiled into
C++"…)

"I looked at the many items sitting in pull request purgatory"

Many of the long standing pull requests are actually created by people
with direct commit access.  We use a model which has a relatively long
pipeline, a fact which I think is justified by the safety
criticialness of the software and our current shortages of active
review. Hopefully long term motion towards increased codebase
modularity will allow faster merging of "safe" changes.

But I suspect there will always be a backlog, at least of "unsafe" changes.

Which brings me to,

"I didn't even know what I had to do"

Above all, I think the most important takeaway from this is that we
need to have better introductory materials.

One obvious place to put them would be
http://bitcoin.org/en/development  but the IRC question makes me
believe that Nick hadn't actually found that page, it's a little
buried.




More information about the bitcoin-dev mailing list