Topic: Bitcoin development problems (Read 1471 times)

Completely agree Re: Mythical Man Month.

As an API developer, I would argue that the core devs goal should be to implement the fewest changes possible while still supporting and strengthening the consensus mechanisms. This is in order to reduce the possibility of introducing flaws into the protocol. I'd say that we do need a team devoted to scalability issues.

If coding was a simple exercise in resource allocation I would agree. In reality throwing more coders at a problem rarely improves a situation. (Mythical man month etc)

Depending on how the code base is organised, you probably only want a team of 3-4 on every major concern. Admittedly I don't know how the btc code is organised (bitcoind & then a couple of GUIs?) seems like 3 teams of 3-4 would be optimal here. For the core that is.

Sure if there are other concerns hanging off that might warrant additional sub teams.

As regards this particular issue, I think once the drama is peeled away it seems pretty clear that the supposed issue with malleability is trivially worked around. I would say it's safer to leave a known issue with a known workaround in place than risk changing the core and introducing some other issue that impacts on the protocol itself. Caution should rightly be the number one concern when messing with core. Not providing crutches to support developers who have written inadequate wallet implementations...

(Just my 2 satoshis)

I don't disagree. Heck my company is a good example of this, we keep it lean and cheap here. My current lead is getting almost everything done himself (though hes pretty fucking expensive, and I am paying for his health insurance and he has 4 kids  ). We do tools for animation professionals, which involves a lot of OpenCL, C++, and QT.

But there is not many software products out there with a $10B market cap which only have 10 people. Try telling Autodesk (~$11B market cap) to limit their team to 10 people and see how that goes. I feel the comparison of bitcoin to a company is fair because bitcoin has competition, and it's a money enabling tool. And to maintain their value in the long term, they need to stay ahead of the curve, whether that means features or stability (hopefully stability).

Just because bitcoin isn't adding as many features as possible doesn't mean it can't benefit from a larger team. You could argue for bitcoin it's better to have a small close-knit team. I agree, but it's not hard to keep the current team and expand on it from the testing side. Rather than adding more people with the intention of adding more features to bitcoin, I would suggest a larger team of testers and QA. I can see in their sourceforge mailing list they mention they don't have time to test every merge request.

More people doesn't necessitate constant updates or an addition of features. It could mean the same amount of updates but with more complete and secure testing, and more foresight on bugs and exploits.



I don't think anything that currently exists will beat bitcoin, not by a long shot. But if nvidia decided that they wanted to develop for litecoin, dogecoin, or some imaginary 'cudacoin', because it helps them sell GPUs, then you can bet the team will be larger than 10 people. They don't even have to own the code for it to benefit them, they only need to ensure that more people are using it, so they can sell more GPUs. I used nvidia as an example because they've done this before with other software. It's a longshot that something like that will happen, frankly I think bitcoin is going to win regardless, but I'm sure whales don't want any possibility of it, even if that possibility is slim.

So you really think that alternate cryptos have larger, smarter, and more well funded development teams than bitcoin?  And the Bitcoin protocol being real-world stress tested on a global scale for vulnerabilities for over the last 4 years, while almost all of the new alts have been around less than a year (in some cases < 6 months)?  And some of them not even based on the bitcoin code, but some dodgy custom code written by obscure developers, that hasn't even been fully tested and vetted?

Hmmm...  yeah, competing cryptos will win.  

+1

Think about it: Do you really want the core bitcoin client to constantly be updated and changed where a careless change might fork or otherwise compromise the blockchain?

The Malleability issue only becomes a real issue when it is not properly handled which in this case is directly linked to custom software used by exchanges.
People need to realize that bitcoin is distributed software and works because all participants comply to a specific set of protocols.
If you change this protocol and older versions reject elements of the new protocol you will have a very real risk of forking the blockchain into an "old"  and " new"  branch.

In this sense it is good that things are not being rushed because we sure as hell don't want or need a forking of the blockchain at this instance

lol @ all you guys that don't work in software

http://en.wikipedia.org/wiki/Mythical_man_month

I think its important for the community to discuss this issue. Ive always wondered why we dont have a huge number of devs working on bitcoin.

Think about it, the bitcoin foundation is literally loaded with bitcoins so i know funding is NOT an issue. The number of coins the foundation is sitting on from donations havent been put to good use imo. Ive tried to discuss this in the forum before but it went largely ignored.

Lets also assume the foundation would prefer hoarding the coins, i am 100% certain if the devs call for an Emergency fund to fix an issue, donations will pour in huge numbers. We are all in this together so everyone wants to see bitcoin prosper

At a 10 billion market cap, we should have atleast 30 devs working full time improving it.

The core devs are probably loaded with bitcoins, so it's in their best interest to fix the malleability issue asap,
and also help the incompetent staff at the exchanges (mostly MtGox, but not only) fix their wallets, before the
market mass panics and we'll see double digits.

It's time for a devcoin pump.

Who pays the devs? 

(I assume the answer is no one).

That really is a good question. Why isn't the number of devs growing ? I mean like 10 people are clearly not enough to maintain all the code. Or are they ?

And I think, really there is the problem. So many new people found about about Bitcoin the past year. Yet I don't see the developer base growing.

The few new people that have a good enough understanding and the necessary skills prefer to develop new alt and to get rich quick over developing Bitcoin. 

How can you make that assertion? Do you have information others don't?

Yesterday gox screams that malleability is life or death, everyone rightly tells them they don't understand the issue and to stop using broken software. The only person who benefits from this today is Mt. Gox. Mt. Gox will try and make itself out to be the hero of the malleability crisis that it created. You see, Mt. Gox can be trusted after all, everyone come back and get rejected 20x for properly filled out verification and do all their business there. Then they can avoid the real issue of a bank run when they have an unknown capital shortfall.

I completely agree, but I've also learned from my experience running a (small) software company that you can't expect the users not to fuck things up. If anything even remotely resembling a flaw exists, users will find a way to break it, even if unintentionally. Bitcoin would ideally not be in a position where they have to rely on people/exchanges not fucking up, because they will.

wait for the fucking confirmations and you will be fine. stupid ass developers deserve their exchanges and services get DOSed if they make all kinds of false assumptions.

I think the most important aspect of this whole malleability problem is not the temporary shutdown of the exchanges, but how well bitcoin developers turnaround a fix.

Think, why didn't developers fix the malleability soon after it was discovered? It's because there were more pressing issues to solve, and didn't have enough manpower to solve all problems at once. If there were a few more devs on it, we would not be having these issues today.

Think of all the money that was lost this week due to the fact they didn't have a few extra developers on this problem. Large holders of bitcoin would actually benefit from investing in developers, to help protect bitcoin from the competition (and malicious attacks). Something with the market cap of bitcoin should have a lot more people developing for it... Perhaps someone like the Winklevoss twins needs to set up a bitcoin development fund that will reward developers for successful commits to the bitcoin project.

If the devs rush out a fix and it fails, or take too long, I will have questions about bitcoins ability to compete long term, even with the head start. Bitcoins greatest enemy is not governments or a lack of usefulness, it is competing cryptos, and if a crypto exists with 10x more developers, it will win.

Keep an eye on this update rollout.

P.S. hello, first post