I hope the core Bitcoin devs are reading this, so they will get an idea of who in this thread is on their level. Someone should send this post to Blockstream, so they can see they win with huge blocks or with small blocks. It doesn't matter. The only future is pegged side chains (for those of you so dumb upthread to claim that pegged side chains don't matter
).
Excellent and timely post by Kondrad! Impressive--I recall he also posted a comprehensive article on sidechains, shortly after the Blockstream white paper was published.
I was going to ask for a link to his analysis of side chains, until I read that myopic "analysis" on the block size limit.
Challenged minds require 30 pages of obfuscation in eloquisms to avoid proceeding directly to the point, because they don't have a lucid generative essence map in their mind.
Eloquisms
Posted by TheFascistMind on Mon 2 Nov 2009 - 13:17
Elegant or novel ways to say something.
Next time you do not understand something, try asking with inverted grammar to illustrate your point humorously...
"Yoda Master tells understands he you not, {something} can be
how, you clarify please asks he to."
The main and long-term salient reason transaction fees have much correlation to the number of confirmations is because different miners would have different costs and absent a block size limit the ones with lower costs would include transactions with lower fees.
One implication is that to do fast micropayments on block chain (thus small nominal transaction fees), would be forced to a pegged side chain that excluded the diverse miners who are causing that variability aka centralization.
(The huge block variance that Cypherdolt pleads for will kill the
decentralized micropayments goal he claims
)
Also that author completely ignores the centralization impacts widely variable block sizes have on centralization due to bandwidth, propagation, orphan rate (or IBLT which is also centralization).
In short this n00b and anyone who thought his anal-sis was "excellent" should STFU.
Even if you argue the other possible outcome which is centralization of mining will be good for reducing variance of block sizes, you are still admitting the Bitcoin is headed for centralization of mining with larger block size limits by some means of squeezing out smaller miners.
But I also acknowledge, even if you restricted the block size, you would end up with centralization, because the rising transaction fees would send transactions off chain (to centralized pegged side chains; or proprietary off chain, e.g. cartel of Coinbase, Paypal, etc). Thus as I wrote upthread, until someone invents a better model for the block chain which can scale, then centralization is inherent with scaling.
Those who cant build, talkOne of the side-effects of using Google+ is that Im getting exposed to a kind of writing I usually avoid ponderous divagations on how the Internet should be and the meaning of it all written by people whove never gotten their hands dirty actually making it work. No, Im not talking about users I dont mind listening to those. Im talking about punditry about the Internet, especially the kind full of grand prescriptive visions. The more I see of this, the more it irritates the crap out of me. But Im not in the habit of writing in public about merely personal complaints; theres a broader cultural problem here that needs to be aired.
The following rant will not name names. But if you are offended by it, you are probably meant to be.
I have been using the Internet since 1976. I got involved in its engineering in 1983. Over the years, Ive influenced the design of the Domain Name System, written a widely-used SMTP transport, helped out with RFCs, and done time on IETF mailing lists. Ive never been a major name in Internet engineering the way I have been post-1997 in the open-source movement, but I was a respectable minor contributor to the former long before I became famous in the latter. I know the people and the culture that gets the work done; theyre my peers and I am theirs. Which is why Im going to switch from them to us and we now, and talk about something that really cranks us off.
Were not thrilled by people who rave endlessly about the wonder of the net. Were not impressed by brow-furrowing think-pieces about how it ought to written by people who arent doing the design and coding to make stuff work. Wed be far happier if pretty much everybody who has ever been described as digerati were dropped in a deep hole where they can blabber at each other without inflicting their pompous vacuities on us or the rest of the world.
In our experience, generally the only non-engineers whose net-related speculations are worth listening to are science-fiction writers, and by no means all of those; anybody to whom the label cyberpunk has been attached usually deserves to be dropped in that deep hole along with the so-called digerati. We do respect the likes of John Brunner, Vernor Vinge, Neal Stephenson, and Charles Stross, and were occasionally inspired by them but this just emphasizes what an uninspiring lot the non-fiction serious thinkers attaching themselves to the Internet usually are.
There are specific recurring kinds of errors in speculative writing about the Internet that we get exceedingly tired of seeing over and over again. One is blindness to problems of scale; another is handwaving about deployment costs; and a third is inability to notice when a proposed cooperative solution is ruined by misalignment of incentives. There are others, but these will stand as representative for why we very seldom find any value in the writings of people who talk but dont build.
We seldom complain about this in public because, really, how would it help? The world seems to be oversupplied with publishers willing to drop money on journalists, communications majors, lawyers, marketers manqué, and other glib riff-raff who have persuaded themselves that they have deep insights about the net. Beneath their verbal razzle-dazzle and coining of pointless neologisms its extremely uncommon for such people to think up anything true that hasnt been old hat to us for decades, but we cant see how to do anything to dampen the demand for their vaporous musings. So we just sigh and go back to work.
Yes, we have our own shining visions of the Internet future, and if you ask us we might well tell you about them. Its even fair to say we have a broadly shared vision of that future; design principles like end-to-end, an allergy to systems with single-point failure modes, and a tradition of open source imply that much. But, with a limited exception during crisis periods imposed by external politics, we dont normally make a lot of public noise about that vision. Because talk is cheap, and we believe we teach the vision best by making it live in what we design and deploy.
Here are some of the principles we live by: An ounce of technical specification beats a pound of manifesto. The superior man underpromises and overperforms. Mechanism outlasts policy. If a picture is worth a thousand words, a pilot deployment is worth a million. The future belongs to those who show up to build it. Shut up and show us the code.
If you can live by these principles too, roll up your sleeves and join us; theres plenty of work to be done. Otherwise, do everybody a favor and stop with the writing and the speeches. You arent special, you arent precious, and you arent helping.
I agree with him that the block size limit is an anti-spam feature and should be treated as such, but the question is why (for example) 20 MB of spam is considered acceptable now. I see no good reason to allow 20x as much spam, when little to nothing has been done to control spam in any other way.
That is closer to a correct view, but still myopic.
Not necessarily. It could be more of a gradual process of spam being accepted into a system that isn't well protected against spam. The miners who are better equipped to produce somewhat larger blocks produce them (by including lower value, lower fee transactions aka spam), which puts more of a strain on smaller miners who drop out. As smaller miners drop out, the average block size then increases, which puts even more strain on the slightly-less-small miners, who then drop out, repeating the process.
The end result is larger spammier blocks, and less broad participation in mining.
The fact remains that as I said there is no more of a mechanism to control spam other than a block size limit than there was when satoshi added the block limit, and I don't really see how spam is 20x8x less of a problem or potential problem now. I'd support a smaller increase though, since I do see technological progress as having reduced the threat of spam by a smaller factor.
Longer term I see the 8 MB and auto doubling every two years as even worse than 20 MB. There is no rational basis to believe with high confidence that gigabyte blocks would be acceptable in 15 years.
Even closer, but still it is not just a "spam" issue.
