Gold collapsing. Bitcoin UP.

[jbreher]

SegWit is actually a good change if you take the politics out of it.

SegWit is one of several potentially usable solutions to the malleability issue. All things being equal, malleability is an issue that should be fixed, though it's never caused me any heartburn, and does not seem to cause the system any problems on the whole. Is SegWit the best malleability solution? Exigency has not yet forced me to an in-depth examination, but from an initial look, FlexTrans may be a better solution. Maybe.

Other than that, the only thing I can get behind -- even in principle -- in The SegWit Omnibus Changeset is the resolution to the quadratic sig hash issue. Another issue that should be fixed, all things being equal. But one that is not currently causing systemic issues, and one for which the protocol already has incentives to route around. Is it the best fix for such? No idea, frankly.

But on the whole, The SegWit Omnibus Changeset has the impression of something that was not properly vetted in the first place. It feels like a tool Core blundered upon that they could employ as a stalling tactic more than anything.

 

[AdrianX]

Fixing malleability is like fixing your cat.

It's the owner that gets to define the problem and it's the cat that gets fixed. (The cat works just fine with it's manhood intact)

Malleable transaction haven't stopped Bitcoin from working. In fact malleable transactions prevent types of applications that move fee paying transactions off the Bitcoin network and onto other networks like the Lightning Network - resulting in less fees to pay for bitcoin security

I like @awemany's aproach to transaction malleability https://bitco.in/forum/threads/gold-collapsing-bitcoin-up.16/page-877#post-33536

 

[freetrader]

I think @SysMan is gone from here, but I'd like to thank him nevertheless, otherwise we wouldn't have such entertaining articles:

http://hackingdistributed.com/2017/02/27/smallest-first-bad-for-bitcoin/

@jbreher : If you are sold on the quadratic scaling promised by SegWit, avoid reading this article because you may end up disappointed with what you would get:
http://www.deadalnix.me/2017/02/27/segwit-and-technologies-built-on-it-are-grossly-oversold/
However, in the interests of seeking truthful information, I must recommend reading it.

 

[jbreher]

@freetrader - I really thought I was conveying that there are only two features within The SegWit Omnibus Changeset that seem like a step forward to me, and that these features both:
- addressed annoyances rather than systemic problems; and
- may be better addressed using alternative designs.

 

[go1111111]

Malleable transaction haven't stopped Bitcoin from working.

They haven't stopped it from working as digital gold or cash for relatively large purchases, but maybe they have stopped it from working for machine to machine payments.

Just because Core is wrong to force layer 2 stuff to be the only path forward for Bitcoin doesn't mean layer 2 stuff might not be awesome. We don't know exactly how Bitcoin will be used in the future. Enabling layer 2 allows for a lot more types of experimentation.

 

[Zangelbert Bingledack]

I feel like overwhelmingly the push to fix malleability is to enable LN, "our savior," because of course increasing the blocksize cap significantly is "off the table."

 

[AdrianX]

They haven't stopped it from working as digital gold or cash for relatively large purchases, but maybe they have stopped it from working for machine to machine payments.

Payment channels like LN and 21inc or other layer 2 options are a good solution for machine to machine payments.

I am pro layer 2 solutions but not at the expense of moving fee paying transaction off the bitcoin blockchain, and opposed to implementing protocol changes that are designed to do just that - layer 2 solutions need to compete on merit in a free market. -

It looks like more research is needed, lets give it 5-10 years so we know exactly what it is that we are doing when address transaction malleability.

the first step in growing bitcoin is increasing the block limit - without removing the block size limit the reduced capacity could cripple bitcoin's ability to compete - it will force fee paying transactions off the blockchain and degrade long term security.

@Zangelbert Bingledack LN as designed needs the segwit malleability fix. LN can work without it but with without a malleability fix as I understand it each LN channels needs to open bitcoin node to bitcoin node. this is how I think LN should be introduced. - not as a scaling solution.

 

[albin]

Since new year i've found myself watching the various games going on with hash rates, rather than the price. Strange since we're at an ATH, yet i've hardly noticed. Understandable?

I can appreciate where you're coming from here, because I don't really think the price at the moment indicates that we've gotten much of anywhere, it just really validates to the outside world that maybe we all weren't really total morons three years ago!

 

[Mengerian]

@Peter R on Epicenter Bitcoin!

 

[steffen]

1. saying that open source quality is not up to the same level of quality as proprietary software is certainly an oversimplification.

My main exerience with open source software is that I used to run Windows on all my computers. I have gradually switched to Linux since 2012 or 2013. The Snowden revelations made my determination to switch higher. I now run Linux almost exclusively but I can see that for example Linux graphics drivers and other drivers do not match their Windows counterparts. I miss some features and have errors which do not occur on Windows. I can see that on dual boot machines. I regularly read https://phoronix.com and that site also gives me the same impression. Linux might be catching up but is still far behind on drivers.

In the case of your model, once the developer provides the solution, the bug reporter can refuse to acknowledge the fix but use it without paying up. So it's still a system requiring trust.

My idea is that the donor locks his bitcoin in some smart contract and cannot technically run from his promise. The one who determines the status of the bug report is also automatically the judge who determines the outcome of the smart contract:

1. Mark the bug as solved => release the locked funds to the developer who solved the bug or
2. Close the bug report without a solution or the acceptable time frame set by the donor for solving the bug has expired => pay back the bitcoin to the donor who made the conditional promise.
   
3. Do nothing => wait until either 1 or 2 happens.

[doublepost=1488324573,1488323946][/doublepost]

It means "unknown". Normally the checkin hash would be there. So this means that you built BU yourself and had a local modification to a file.

Thank you for answer.

I followed this procedure from https://www.bitcoinunlimited.info/download to install version 1.0.0:

If you're running an Ubuntu system you could use our official repository. It supports four archs - amd64, i386, arm and arm64 - and 3 Ubuntu versions - 14.04, 16.04 and 16.10. To install the binaries from the repository execute these commands from a console:

sudo apt-get install software-properties-common
sudo add-apt-repository ppa:bitcoin-unlimited/bu-ppa
sudo apt-get update
sudo apt-get install bitcoind bitcoin-qt

A few days later I received an update to version 1.0.0.1 together with other Ubuntu updates.

I don't know if that falls in the category "this means that you built BU yourself and had a local modification to a file."

 

[lunar]

Great talk @Peter R

Very eloquent and approachable speaking style. The epicentre guys are killing it with these great, highly informative interviews. Keep it up gents.

Intended to be constructive here, but I feel you missed a trick when the conversation turned to Classic. The focus drifted to their past issues. Perhaps it could have been steered to mention how they adopted BU block scaling ideas and the good working relationship between the two teams, towards fostering a multi implementation ecosystem.

Good stuff....... share share share

 

[Richy_T]

I think since no one else is bothering to do it (yet they talk about it all the time), we should come up with a reasonable definition for decentralization and how to measure it. I think this kind of thing is Peter R's field and I'd like to hear his thoughts on the matter.

 

[freetrader]

@steffen:

Linux graphics drivers and other drivers do not match their Windows counterparts.

The problem here used to be this:

At first, manufacturers of graphics cards didn't release open-source drivers at all, nor sufficient documentation on their hardware for open-source developers to write drivers exploiting all their features. This meant that even if open-source devs wrote a clean-room driver that worked, its performance would be below that of the proprietary, closed source driver.
This is not a "software quality" problem with open-source, but proprietary software and hardware vendors exercising their powers to compete.

Nowadays some of the vendors do provide open-source drivers, but these are still limited on purpose, because large companies provide incentives to have better drivers on their platforms.

Graphics drivers are a very ugly business when it comes to "quality" - it's a cut-throat business tied to selling particular hardware in a very competitive market, and the manufacturers basically use intellectual property as a shield for all sorts of shenanigans such as implementing high-performance code which is specifically optimized to make some benchmarks look good, or optimizing for certain games where there is a contractual relationship with the game producer.

This has resulted in hugely bloated driver packages where an update is hundreds of megabytes, and a kind of 'payola' business model where if you want to have good performance for your application, you need to 'work closely with the vendor'.

As for error messages on Windows - my experience has been that when errors happen with some software on Windows, it is usually MUCH harder to find a log file entry, and in many cases there just simply isn't one at all.

Of course I could never trust keeping high-value personal & confidential information, let alone a wallet such as Bitcoin on an Internet-connected Windows system. Windows which used to come with a backdoor key for the NSA, and whose intrusive snooping on the user has only become worse with every recent iteration.

 

[Zarathustra]

Coblee promoting UASF (Upper Altitude Science Fiction).

 

[freetrader]

@Zarathustra : Wait, it's not Unanimously Agreed SegWit Forcefeeding?

ノʕ•ᴥ•ʔノ ︵ ┻━┻​

In other news, nowadays if you contribute any transactions to the full mempool you are apparently called a spammer and told you are abusing the protocol rules.

 

[albin]

That's interesting, if you look at the votes in that interchange, it's very obvious that Charlie Lee is followed everywhere by two pocket vote brigaders.

Honestly if I were a pre-existing dev in the Bitcoin space prior to the recent (since spring 2015) kerfuffle, I'd very likely be spitting Core poison every chance I get, because of a twisted Pascal's wager.

There's one side that will systematically purge you from Bitcoin if you speak out against their ideas, and there's another side that will do nothing bad to you at all.

 

[chriswilmer]

How on earth do you get from "there must be extreme consensus" to "we need to force segwit via executive order"?

 

[sickpig]

A few days later I received an update to version 1.0.0.1 together with other Ubuntu updates.

I don't know if that falls in the category "this means that you built BU yourself and had a local modification to a file."

it doesn't fall in that category, the fact is that we were in a rush to put out a new release cause we found a bug in the previous one.

In doing so I messed up the production of the deb files for the Ubuntu ppa repo. And this is the reason why you see such version in your log.

I'm rebuilding the packages to remove this problem, what's your Ubuntu version by the way?

 

[steffen]

@sickpig

I'm rebuilding the packages to remove this problem, what's your Ubuntu version by the way?

Ubuntu 16.10

 

[Christoph Bergmann]

Hei ...

UBUNTU: After my windows broke and I had no key or something else, I installed Ubuntu. This is around 1.5 years ago. I love it. It is faster and has not the windows-tendency to babysit the user with unhelpful help and baby options. With Win10 it has become horrible, and I'm really upset about how they spy on you and how they suggest you to use their shitty searchmachine. Had an old laptop, which needed ten minutes to load with windows and everything was soooo slow. Installed Lubuntu on it, and now it is working fine. On the other side, yes, I see myself all day searching the internet how to make things run on ubuntu. Recording music becomes a major problem, when a recording software on windows just finds the right channel, on ubuntu you have to google what the right channel is to select. And this was after I installed three programms with all the subpakets and so on which just didn't work (or I lack the knowledge to make it work), but left randomly installed pakets and software on my system.

Same with gaming. Steam for windows is fun, steam for ubuntu is a desert. And games that work, are incompatible with multiplaying. On windows you just install a game and play, on ubuntu you install wine, read through a dozen of forum, install the game with wine, try this, try that, fail here, fail there, and in the end you just plug in your windows 10 harddisk to play a game ...

FORCED SOFTFORK: This is madness. The arguments of core and its fans have really lost its sense, and things have become crazy ...

@albin nailed it, I guess:

Honestly if I were a pre-existing dev in the Bitcoin space prior to the recent (since spring 2015) kerfuffle, I'd very likely be spitting Core poison every chance I get, because of a twisted Pascal's wager.

There's one side that will systematically purge you from Bitcoin if you speak out against their ideas, and there's another side that will do nothing bad to you at all.

Yeah. You forgot, that the one side not only doesn't troll you with an army trolls, but gives you a 6-digits job from home-office for being with them. If you dare to go to the other side, you get banned, you loose the ability to participate in any important development channel, your commits will likely ignored, and on top trolls will spin their crazy stories about you and repeat it so often, that everybody thinks you are a brainless maniac trying to destroy bitcoin and unable to write 3 lines of code without 2,000 errors ... World is bad.