How Open Source Projects Survive Poisonous People

[cypherdoc]

i would vote to follow Gavin's lead.

[interlagos]

I believe there is enough space for everybody's ideas in this community. That's why we have alt-chains.
If there is controversy over a particular issue, let's make sure that both ideas are implemented (in separate chains)
and then the time will tell who was right.

Bitcoin has been driven by Gavin fairly well so far, and I hope this will continue in the future.

[N12]

i would vote to follow Gavin's lead.

+1000

I trust our benevolent dictator.

[piuk]

Sure, let’s just wait with multisig until Bitcoin is deserted because normal people keep having their Bitcoins stolen by malware and therefore simply shouldn’t use Bitcoin, but rather stay with PayPal and online banking.

Split key is a step in the right direction, but it is not some magic solution that instantly solves all security problems. If a User has one half of their private keys on their phone what happens when they drop it in a river on a night out... If your trusting your split keys with a "wallet protection service" there is no way you can be sure they aren't colluding with other entity - or go rogue and hold you keys to ransom.

[anu]

how so? svn projects have a history that cannot be altered. there are many year old projects relying on svn! so, it's just like with the blockchain.

No single svn repository is at the heart of svn. The block chain is the heart and soul of Bitcoin. A corrupted repository can be resored from a backup and lost commits can be recommitted. Bad enough, but a corrupted block chain is much, much worse that that. Therefore extreme conservativism it the right approach to any change in the protocol. That mindset seems to be missing from a lot of people here.

[Gavin Andresen]

RE: debating ideas rather than people:  I've tried very hard to do that.

I said last year when I reluctantly agreed to function as the lead core bitcoin developer that I have zero experience leading open source projects.  But I try to do my due-diligence and learn from the experience of other successful projects.

The "aha" moment for me yesterday is the point in the video that I linked to, where the advice is to evaluate whether or not somebody causing issues for a project (whether intentional or not) is a net positive or negative to the project, and if they're "more trouble than they're worth" get them out of the project.

This isn't about BIP 16 versus BIP 17, this is about one person draining the rest of the development team with nagging, idealogical rigidity, and holy wars against "impure" ideas.

I try very hard to consider that maybe I'm wrong, but I think the evidence is clear.

You might also argue that the subversion people take the wrong approach, in which case please send me a link to some other open source project that has dealt with the issue in a different/better way.

I thought carefully about where to start this, and decided I might as well start it in the most public bitcoin discussion forum, because I think if I started it anywhere else it would eventually just appear here as "There's a Secret Conspiracy Started By Gavin To Oust A Valued Developer!"

[N12]

I really think the core dev team should decide how to go on (majority vote). It’s none of our business, we don’t code it and barely superficially understand it.

[fornit]

gavin, what they talk about in the video is one person derailing a whole project over extended periods of time. based only on information on this forum it seems to me that luke definitely had its part in not meeting a deadline that was already ambitious and even people agreeing with bip 16 didnt agree with the need to rush.
i also dont really get that part. either its really just a "different color of the bike shed"-problem. if its just a minor technical issue and doesnt matter much anyway then not just choose bip 17? on the other hand if its not so minor why rush it?

i understand why you are really pissed. but i think because of this you also overestimate the negative influence this whole affair has on the public. for me personally the bug in the encryption for example was way worse than this delay and public display of disagreement resulting in a litte bit of fud.
the fud btw would have been smaller if the other developers had voiced their opinion a little more loudly.

i also agree with the people that think a vote by the developers on the bips is a good solution. you may even vote on excluding luke from the development alltogether. but what you do is just incitement imho and therefore equally unproductive.

[casascius]

Gavin strongly has my support in this, for what it's worth.

I do not believe he is rushing things.  I agree with the sentiment that if we shelve it for three or six months, that no progress will be made on it, and we'll be back to square one with even more of a burden to get it going.

[memvola]

Split key is a step in the right direction, but it is not some magic solution that instantly solves all security problems. If a User has one half of their private keys on their phone what happens when they drop it in a river on a night out... If your trusting your split keys with a "wallet protection service" there is no way you can be sure they aren't colluding with other entity - or go rogue and hold you keys to ransom.

Probably off-topic but can't you use deterministic keys, or even a single key (easy to keep a copy) on the confirming device? Also, it should be standard for services to send users their keys via a printout by mail or an encrypted e-mail. I'd use a service which would send me a Jabber message to confirm my payments. Easy, convenient and foolproof. We need this as soon as possible (there is a bit of structure that needs to be built on this technology, which will also take months, and after they mature will it be only as convenient and foolproof as we hope).

[paraipan]

subscribing... i'm not allowed to post anything else

[Gavin Andresen]

gavin, what they talk about in the video is one person derailing a whole project over extended periods of time.

Yes, exactly. At some point you have to say "enough is enough, I'm not going to let this person derail the project any more going forward."

I'm saying that point is right now; see the unproductive, one-sided argument about BIP 20 versus BIP 21 on the bitcoin-development mailing list that is re-hashing a wiki editing war that "the rest of us" just gave up on a year ago for the latest example.

[rjk]

subscribing... i'm not allowed to post anything else

Why do your posts keep disappearing?

[kgo]

Gavin, please debate Luke's ideas, not his person.

Unfortunately, I don't think it's possible to debate Luke-jr himself, which is the problem.

I first became exposed to his debating style on bitcoin-otc.  As we know, OTC stands for Off-Topic Channel and I don't take discussions there too seriously.  But he'd get into debates with people, and respond to well thought out points with "No."  Or "You're wrong." Without providing any evidence, without saying he's already discussed that, etc.  Other responses proved equally obtuse.  How do you debate with someone who simply cuts you off with "No."?  With someone who refuses to take your opinions and even evidence into account?

At first I thought he was just trolling, since these weren't the most important conversations in the world.  But since then I've seen a pattern across the board, not just -otc, where he thinks in absolutist terms, and refuses to even consider other opinions or evidence once he's made his mind up.  You'll see the same thing in serious discussions about serious matters.

And it really is unfortunate.  Luke has obviously contributed a bunch of stuff to the bitcoin code itself and the community.  But I can see why Gavin thinks the problems outweigh the benefits, and why it's difficult to believe that Luke would 'reform' or moderate his behavior.

[BadBear]

subscribing... i'm not allowed to post anything else

Why do your posts keep disappearing?

Off topic split here
https://bitcointalk.org/index.php?topic=62054.0

[rjk]

But he'd get into debates with people, and respond to well thought out points with "No."  Or "You're wrong." Without providing any evidence, without saying he's already discussed that, etc.  Other responses proved equally obtuse.  How do you debate with someone who simply cuts you off with "No."?  With someone who refuses to take your opinions and even evidence into account?

At first I thought he was just trolling, since these weren't the most important conversations in the world.  But since then I've seen a pattern across the board, not just -otc, where he thinks in absolutist terms, and refuses to even consider other opinions or evidence once he's made his mind up.  You'll see the same thing in serious discussions about serious matters.

His memory is a PROM, not an EEPROM Flash it once, set in stone.

As we know, OTC stands for Off-Topic Channel and I don't take discussions there too seriously.

Lol. Over The Counter is what it stands for, but I can see how it is definitely more of an Of Topic Channel by what actually goes on.

[MoonShadow]

RE: debating ideas rather than people:  I've tried very hard to do that.

I said last year when I reluctantly agreed to function as the lead core bitcoin developer that I have zero experience leading open source projects.  But I try to do my due-diligence and learn from the experience of other successful projects.

I don't speak for anyone but myself, but I am of the opinion that you are the legitimate lead developer for the current state of the standard bitcoin client.  Even the most openly sourced, and collectively developed, projects have their benevolent dictator.  The GNU/Linux kernel development over the past 17 years being the most prominant example.  If you are honestly certain that you are correct in your assessments, make your decision known and run with them; however, if you are not certain let that be known as well and let us has it out for ourselves.  If you force a code split, so be it.  We all know that it will work itself out eventually.  There has already been a number of splits among this community thus far, and such code splits don't necessarly spell the end of an open source project who's time has come.  This community is no longer small and weak enough to be destroyed by such a division.  I won't even promise that I will personally upgrade without my own due diligence, but the details that really matter will filter out over time.

[makomk]

What LukeDashJr did to coiledcoin, without his users consent, or permission, is proof positive of why he needs to be severed.

Perhaps ironically, the entire reason I launched Coiledcoin in the first place was because I saw that P2SH was turning into a fiasco and didn't expect it to launch any time soon in any form - and it would probably have failed to do so even without Luke's meddling. Gavin was doing the equivalent of a hairpin turn in a supertanker and his code kept turning out to have bugs that would cost miners money, and none of the reasons for the last-minute change or the rushed schedule were terribly convincing.

It's actually gotten worse since then. gmaxwell keeps accusing anyone who dares to tell end-users about any of the technical details of spreading FUD by drawing the users' attention to something that's none of their business and is trying to shut them up, whilst spreading his own fricking FUD about Luke Jr's proposal (no it doesn't reverse the bugfix that evaluates scriptSig and scriptPubKey seperately rather than appending them, nor can it cause the kind of security vunerabilities that split protects against). Gavin claimed that Deepbit was somehow attacking Bitcoin by not immediately adopting his buggy code, despite the fact that hardly any of the other pools had either, in an attempt to force Deepbit to do so and use their hash power to force all the other pools to follow suit. And so on and so forth. Luke's actually been polite by comparison and frankly that's scary.

[rjk]

Luke's actually been polite by comparison and frankly that's scary.

He's plotting something! Everyone go down to your tactical nuclear bunkers and wait out the storm!
/thread

[FlipPro]

Thanks to midnightmagic who directed me to this very helpful video about identifying and then dealing with "poisonous people" in open source software projects:
  http://youtu.be/ZSFDm3UYkeE

Advice on disinfecting:
  http://youtu.be/ZSFDm3UYkeE?t=34m22s

Watch the whole thing for examples of "poisonous person" behavior, like repeatedly flooding mailing lists/forums with their opinion or comments, not listening to the opinion of others, or making sweeping "the world will end if..." statements about the project.

I'll be blunt:  I think Luke Dashjr fits the definition of a poisonous person, and I think Bitcoin would be better without him. At the very least, we wouldn't be creating two BIPs for every technical issue, one for Luke and one for the rest of us (see BIP 16/17, and now we have BIP 20/21, too).

+1 Gavin,

100% agree with your analysis on Luke-Jr.