We agree that it is necessary to transmit more transactions, but to increase the block size is simply not an effective solution.

This post first appeared on bitcoinblog.de January 18, 2016 in German. The translation follows:

Posted on 18th January 2016 by Christoph Bergmann at bitcoinblog.de

Jonas Schnelli is one of the five core developers with commit access to the code. The Swiss resident is mainly responsible for the user interface. In this interview he explains why the core developers do not want to raise the block size limit, why Core has an image problem and why he would continue to work even after a 2 MB Hardfork on Bitcoin.

Hi Jonas, tell me something about your background?

With pleasure. I am 36 years old, live in Basel and work in the IT sector. I have no higher education, but have taught myself. With development I started about 26 years ago, at that time still on the Amiga and C64. Since I started working, I am self employed, and have now established three companies.

How did you come to Bitcoin?

First, an employee of mine told me about Bitcoin. At first I thought, that’s advanced for 2011, but after learning more I thought well, this looks really exciting. So I spend some time learning more and when I then saw how the price had risen massively, it grabbed me. I really dug in to Bitcoin and made the first commit in Bitcoin Core in 2012.

You’re one of 5 developers with commit access. How did you get involved?

I have pretty much worked for the last year full time on Bitcoin Core. I sure spent a lot of time. I think I am already developing quite well, but I am not as extensively qualified as, for example, Pieter Wuille. But it takes people like me to keep the system clean and keep track. My work this year is primarily on the GUI.

Have you stopped your other work to make Bitcoin full time?

Not quite. I have rather shifted from working on my other companies as much. I’m self-employed and therefore can manage my time among my companies better than if I had a boss to answer to. I pay my work on core from my gross income, of course in the hope that it pays off, but that’s not my central driving force.

What is your key driving force to work on Bitcoin?

Very difficult to say. I think I’m extremely interested in the system. I feel it is the future and I would like to participate and help Bitcoin.

Are there any known changes in the Bitcoin client you are responsible for?

What is obvious, are the icons, logos and HiDPI application, which are mine. From the next version you will be able to capture nodes and stop further access.

Is not that a blacklisting?

Hmm, yes, something like that. A blacklisting banishes only nodes that misbehave. With 0.12 you can effectively ban nodes via the IP address.

Why?

There are several reasons. For example, it has been observed that some nodes are not useful, such as scanning nodes or nodes that only consume or pretend to others that they are full Nodes.

But again back to other updates. Perhaps the most useful feature in 0.12 is the Max upload target. A Node caused quite a lot of traffic, and a lot of traffic it with information about the Upstream. Together with other developers, I then designed a regulator to set a Max upload target.

That sounds great … but let’s get to some unpleasant questions. How is it currently to be a core developer?

It goes like this, of course, I am currently busy. My opinion is that we developers have a technical discussion in which there are different opinions, but at the same time there are also differing opinions based on politics. All have the same goal – to create a decentralized system with the largest possible capacity.

An important point in the debate is that pragmatism varies among the developers and is, I think, the heart of the problem. Some people want a solution now, while others think long-term.

It would be pragmatic to increase the block size?

Yes, but I am skeptical. The solution should be simple and the problem is that Bitcoin is not scaled linearly. If you increase the block size, the load on the network increases as the square, since each node has to know every transaction. My basic problem is that the number of nodes has already reduced. Bitcoin is no more than the network of Full Nodes who already have a high exposure to traffic and CPU. The larger the blocks, the more difficult it is to operate a Full node.

I have a parallel in mind. Here in Switzerland there are a few beautiful lakes and rivers that people began to build paths around, because they are nice to walk around. Then they built roads, and now we have 2 and 3 lane roads and train lines around the lakes. One can compare in my opinion with the Block Size debate: people are not looking for the simplest alternative, namely major roads, and that prevents us from making better solutions.

What would be a better solution?

For example Segregated Witness by Pieter Wuille. If we remove the signatures, we can in a block bring in twice as many transactions as before. For this solution, we would have almost 100 percent consensus, because all developers approve of it.

I’ve heard criticism that it is complicated and requires a lot of work from wallet operators and exchanges.

Yes that’s true. It is exhausting. It is still easier to build larger roads. A parallel is with the SMTP protocol for e-mails. It has also never fundamentally changed, but simply made more complicated, with extensions such as ASMTP, SPF etc. The protocol was never adapted fundamentally with more efficient and pragmatic changes.

But e-mails work quite well?

Yes, but when you talk about e-mail, you speak often of Gmail. It is hardly decentralized if you write an email. You do not have your own SMTP server, and you’re dependent on someone else to receive and pass on your e-mail. If we want the Bitcoin protocol to survive long term in a decentralized manner, then we must be ready to make deeper interventions.

But long term there is no way around it, increasing the block size?

Why? Even if we increase the block size to 20 MB, we would not be able to match Visa. With an immense block size, it might be possible to produce a global payment system for the end user, but to operate a full node in this environment is extremely difficult, even in five years.

What are the alternatives?

Around the Lightning network. It would be too simple scaling for a decentralized system, if we have any payments on the block chain. I think there has to be subsystems such as Lightning, if we want to have mass use of bitcoin.

But you do not yet allow as many transactions as possible on chain?

That’s what we do today. But core is of the opinion that the block size of 1 MB, together with measures such as Segregated Witness for the time being is sufficient. Most at Core are of the opinion that increasing the block size is not the preferred solution in the short term.

But still better than no solution?

For this purpose it is necessary to define the problem. The problem may be that transactions with a low miner fee may mean you have to wait six hours until a transaction is confirmed. In that sense it is not a problem, but it just takes longer and is more expensive for fast confirmation. With expensive I mean a transaction might cost 6 cents instead of 2 cents.

But if the transaction is permanently on the capacity, then the number of unacknowledged transactions rises but ad infinitum?

There are still many blocks that are <500kb. But if all blocks are full, some transactions are uneconomic and associated costs one dollar. We agree that it is necessary to transmit more transactions, but to increase the block size is simply not an efficient solution.

Are you worried you will lose the support of the community with this road map?

That’s already happening. The community is divided. Many arguments are wrong headed, such as ones driven by conspiracy theories about Block Stream. Scalability is a technically highly complex matter, and most proposals that claim to be an easy fix are wrong because they do not get the technical basics right.

But I do think that Core has an image problem. We are now working to present a better appearance to the community. The average Bitcoin user may think more transactions per second are good, but the long term consequences must also be considered and consensus reached.

Currently new clients are being considered like XT, Unlimited and Classic. What do you think about this development?

I think that’s great. In open source alternative applications are always good. Basically I find Bitcoin Classic and Unlimited super, if not for the necessary Hardfork that Core developers do not agree on. But we must not forget that some Code Forks are good, but these new clients come with a Hard Fork, which means that you can create a new crypto currency. If one splits the block chain without Consensus, there is an Altcoin, on one side or the other. If we assume XT would have 80% of support, then the 20% Core users would be on an Altchain. And to do that without consensus, without the full support of the community, is very dangerous and can even lead to the death of Bitcoin.

Can Core prevent it?

Yes, a compromise could have been made if we agreed to implement a small increase in blocksize. We have now had two conferences on scaling we have learned from. We have come to agree that Segregated Witness is the better solution to more transactions per second. This is more complex but technically the better solution.

But to XT I must also say that it was very much dependent on Mike Hearn and he lost the confidence of the Core team and dismissed. And Mike is not the person who can gracefully fade away, so he has also claimed Core is fragile and in danger.

And how is it with Gavin Andresen and Jeff Garzik?

For them, it’s different. They consider Classic a legitimate Fork. What bothers me, is that Classic does require a Hardfork. Even though I am concerned I am still pleased that there are companies in the background like Coinbase who are naturally very capitalist oriented. A Hard Fork is something very dangerous, and I think it’s critical that it has the support of major companies.

I suppose Coinbase wants to just create an environment in order to achieve further growth. You have also said that the block Stream conspiracy theories are nonsense …

Yes, then I should probably also believe in any conspiracy with Coinbase, you’re right.

Suppose that Classic succeeds. Would you then support them, or would that be for you the end?

No, I’m used to adapting myself, and while my opinions are important to me if the status changes I can adapt.

That would not be the absolute end for your involvement?

No, I am in love with Bitcoin, but can be separated from my own nature. I have no problem to jump over my own shadow. What I want is to make Bitcoin better.

At the end I would like to talk about censorship. There has been much criticism that Core has not distanced itself from the censorship on some Bitcoin media. What do you think about that?

Censorship bothers me also. But we are now going to distance ourselves. We now have a website and want to move away from bitcoin.org, reddit and bitcointalk. This is a result of the censorship and the moderation that went too far.

In part I understand, because the community is more politically driven than technically, which increases the risk of Hardfork, but I also see that censorship does not help. Bitcointalk for example, which is led by this infamous Moderator, has always been the most important medium for all discussion. Today it is no longer used by the Core Community. The main discussion will take place in the IRC channel.

Jonas Thank you for taking the time.

  • rodomonte11

    It’s really sad how emotions and politics can destroy any good purpose in a second, I hope it will never be an hard fork and developer can regain the trust they deserve without judgment. I hope people will have the seriousness to create their own altcoin instead on force a commit to happen using unaware people and ignorant miners.

    I’m sure, if they are just able to keep calm a logic driven solution will arrive, but if they will chose the soon and easy solution Bitcoin in the long term will be doomed. To get an immediate relief at the expenses of the future it’s the very own definition of a debt, and it’s nonsense, especially in Bitcoin.