Author

Topic: Why its not just about block size (Read 386 times)

legendary
Activity: 2576
Merit: 1087
August 28, 2015, 01:36:47 PM
#6
So the point he is doing is that it is better to have better implementations for the same Bitcoin?

How can different implementations be fully compatible?

And if it is that what he is saying...why a fork? I mean, if he is advocating lots of different "compatible" or "agreable" implementations, Does not the fork make it incompatible with a non-forked Bitcoin?


I see a contradiction there.

Forking the chain is at one end of the spectrum of compatible-agreeable-fractious-incompatible spectrum.

All implementations are on this scale somewhere even differing versions of core which can also be at the incompatible end.

What happened in the past with core was that consensus was reached on chain (as opposed to on forums!) and that, I think is what will happen with XT, and that is what should happen with all implementations going forward.

I think that's what Gavin was probably saying. Though of course I could be wrong Smiley
legendary
Activity: 2576
Merit: 1087
August 28, 2015, 01:32:30 PM
#5
I think having different options is what makes this exciting and so on, but the thing is, the proposed scheme for upgrading the blocksize by Gavin would render us with an insanely big blockchain, and completely out of the market when it comes to running nodes. Nodes would end up being ran by corporations. We are already fucked enough with the apparently unavoidable fact that centralization of mining was, the last thing we need is centralization of nodes. That's why it's such a big deal to have huge blocks and at some point it was starting to look as if centralization of nodes is exactly what they wanted by pushing the XT agenda, and im not commenting on the checkpoint thing etc.

I'm not sure how you can be sure that increasing the limit automatically means there will be an insanely big blockchain. I also don't know what is considered insanely big. I accept the blockchain is likely to grow into its new shoes but I also think there is that while economic incentive thing going on for miners.

I don't dispute that any of the things you say could happen. To jump from 'this could happen' to 'this is what XT is planning' is to assume that XT deva are prescient and know what will happen. They don't you don't I don't so the o lay thing we know is that peopele are doing what they think is best given the information they have at a point in time. When new information arises we can make different choices.
legendary
Activity: 1148
Merit: 1014
In Satoshi I Trust
August 28, 2015, 10:32:39 AM
#4
Does XT make Mike and Gavin dictators?

Some people are portraying XT vs Core as “Dictatorship vs Meritocratic Consensus”. This isn’t right.

Gavin and I can’t be ‘dictators’ because all we do is write software: if we go crazy, or do other things you disagree with, XT can be forked in exactly the same way as we did.

https://medium.com/@octskyward/an-xt-faq-38e78aa32ff0
hero member
Activity: 770
Merit: 500
✪ NEXCHANGE | BTC, LTC, ETH & DOGE ✪
August 28, 2015, 10:22:19 AM
#3
So the point he is doing is that it is better to have better implementations for the same Bitcoin?

How can different implementations be fully compatible?

And if it is that what he is saying...why a fork? I mean, if he is advocating lots of different "compatible" or "agreable" implementations, Does not the fork make it incompatible with a non-forked Bitcoin?


I see a contradiction there.
legendary
Activity: 1204
Merit: 1028
August 28, 2015, 10:13:02 AM
#2
I think having different options is what makes this exciting and so on, but the thing is, the proposed scheme for upgrading the blocksize by Gavin would render us with an insanely big blockchain, and completely out of the market when it comes to running nodes. Nodes would end up being ran by corporations. We are already fucked enough with the apparently unavoidable fact that centralization of mining was, the last thing we need is centralization of nodes. That's why it's such a big deal to have huge blocks and at some point it was starting to look as if centralization of nodes is exactly what they wanted by pushing the XT agenda, and im not commenting on the checkpoint thing etc.
legendary
Activity: 2576
Merit: 1087
August 28, 2015, 10:03:58 AM
#1
At first I thought it was, and I've come full circle on this.

It was my opinion that block size should be bigger (I accept others feel differently - thats cool). The only way for me to do anything about it, other than arguing oti, was to switch clients.

Then I read all the bad stuff about Mike and Gavin and what they were doing and I was a bit worried. I read about how they were putting nefarious changes that track IP addresses. Blocking nodes etc

That made me worried too. I don't like creeping censorship as much as the next man so tools that allow it should be regarded with suspicion. Then I read what the change did and saw that it didn't seem (to me) to be half as bad as people were making out.

Then I started to read more about why XT came about I'm not going to link to it all because everyone should do their own research. My understanding though in summary is that the core dev environment seemed to be pretty toxic. This is a pretty bold claim so I started reading through some past posts on the developer list, pull request discssions etc to try and establish how legitimate this claim is. To me it seems to have some legitimacy, but I would encourage you to look for yourselves and see what you feel. I won't post links because I don't want to be accused of leading you.

Now I'm not here to tell you Gavin and Mike are the good guys and core devs are evil. Nor am I here to support those that think XT is the end of times and the core dev team are the knights in shining armour that are going to rescue os from the defilers that sullied out land with a fork. The devil carries a fork. Therefore XT is the work of the devil.

I'm here to say that I think its important that people look around at what is going on and have a real honest think about the situation. Don't believe any of what I wrote above check it all out. Have a look at the pull requests that have been submitted, the discussion that follows. The manifesto posted by XT and the articles written explaining there philosophy. The actions that have been taken by the various participants.

When I did that I found that I no longer supported XT for the block size limit upgrade, but I supported it because it felt like it was right that there should be diversity in clients. That hard forks are almost unavoidable so better that we learn to love the bomb.

I will paste one quote though because I can't spend this whole post being entirely ambivalent, it was by Gavin and whether you like him or not I think he makes an interesting but probably contentious point, which is based on some facts as opposed to being pure conjecture, or open to interpretation. I'm not going to try and assert this is some ultimate truth and anyone that disagrees is a traitor or anything like that either. None of the discussion like that is helpful. There are some smart people on these boards and if they can all discuss instead of bickering maybe it can set an example. I'll do my best, I hope others will too.

Quote from: Gavin Andresen
Yes, I think having lots of diverse implementations of the Bitcoin protocol is a very good thing. I know Greg Maxwell has been pretty vocal in the past disagreeing with that, and I understand his point of view: it is really hard to get two or three or eleven different implementations to accept and reject exactly the same sets of transactions or blocks. Consensus is hard.

But it is hard to get EXACTLY THE SAME implementation running on different hardware or just running at different places on the network to accept and reject exactly the same sets of transactions or blocks. Bitcoin Core versions 0.7 and earlier could "self-fork" -- exactly the same code running on identical hardware could disagree about what chain was valid because of a bug.

And it's now safe to talk about the REAL reason for BIP66, which was to prevent a fork between 32-bit machines and 64-bit machines.

One of the principles of security is to eliminate single points of failure. Another is to try to minimize the damage done by failure of any one component. A diversity of implementations running on the network helps both of those aspects of security, and that is why I'm happy to see other implementations like Conformal's btcd or Coinbase's Toshi, and why I think multiple forks of Satoshi's original codebase is healthy for the long run.
https://groups.google.com/forum/#!msg/bitcoin-xt/PBjK0BuB7s4/8LREpcaNBQAJ

Polarising debate isn't helpful I can see advantages and disadvantages to all. Stubbornly setting up in one camp and arguing beyond reason that yours is the one true solution isn't good. You may say that is what XT has done but this can be levelled equally at core. XT is a choice, so is core. Lets have more choice and less 'Conscription'.
Jump to: