What the Crypto Community Should Learn from GitHub's Acquisition

https://www.buildblockchain.tech/blog/crypto-learn-from-git

by benmdi • 8 years ago

29 33 8 years ago

qsymmachus• 8 years ago

This is the thousandth time I've clicked on a "crypto" article expecting it to be about cryptography, and instead it's about blockchain/cryptocurrencies.

Is this a losing battle?

benmdi7• 8 years ago

lol, sorry. I actually used to find this frustrating myself. At some point, I just gave up and it switched in my brain. Guess I'm part of the problem now....

recursive2• 8 years ago

So how do you refer to cryptography? Just fully spell out "cryptography"? This is aggravating to me as well.

gsich• 8 years ago

Haven't heared of a "cryptography community", so I'll assumed cryptocurrencies.

pejrich• 8 years ago

There was a cryptography community before there was a cryptocurrency community, by definition.

eesmith2• 8 years ago

At one point about 1/2 of the "cryptography community" was employed by the NSA. Anyhoo, here's some references over the last few decades:

"But now, in light of the leaked NSA documents, NIST is reevaluating the cohort of random-number generators that includes Dual EC DRBG, and it has opened a public forum on its website where the cryptography community can raise concerns." - https://spectrum.ieee.org/telecom/security/can-you-trust-nis... (2013)

"Anyone possessing that second set of numbers would have what’s known in the cryptography community as “trapdoor information” – that is, they would be able to essentially unlock the encryption algorithm by predicting what the random number generator generated." - https://www.wired.com/2013/09/nsa-backdoor/ (2013)

"It was declassified on 24 June 1998, shortly after its basic design principle had been discovered independently by the public cryptography community." - https://en.wikipedia.org/wiki/Skipjack_(cipher)

"The cryptography community has long known that generating random numbers requires great care and is easy to do poorly; Netscape learned this lesson somewhat painfully." - http://www.ddj.com/windows/184409807 (1996)

"Would the cryptography community be better off if only the NSA did patent reviews?" - https://pdfs.semanticscholar.org/216c/216af763d36f3f35dc9f33... (1988)

mxuribe• 8 years ago

I happened to notice the inclusion of "blockchain" in the destination domain name...so got the clue before clicking through. But I don't even think this piece focuses much on cryptocurrency anyway...its more about the hazards of centralization/decentralization with respect to protocols and platforms that host services using said protocols. Don't get me wrong, i actually really liked the article...but a discussion on cryptography or crypto-currency its not.

sorokod• 8 years ago

In one of Neil Stephenson's books he mentions that in the 19th ( ? ) century, people who performed (manual) calculation where called "computers".

shpx• 8 years ago

Do you also get frustrated when people say "decimate" when they don't mean "remove 10% of something"? Words change meaning, and that's okay. There's way more people talking about bitcoin than AES, it would be more efficient for them to get the shorter word. Cryptographers shouldn't get to keep a prefix forever just because it makes more sense etymologically and they used it first.

You probably understand all of this and this is you exerting social pressure against change you disagree with. I say best to just get used to it. Wait until these get-rich-quick schemes stop working and we can have our abbreviation back. I wouldn't be surprised if we end up waiting decades or longer though.

https://xkcd.com/1108

qop• 8 years ago

Man that's a comment I really resonate with.

I wish I had sql or something hn submissions. If it has crypto in the title but block chain in the article, I'm not interested....!

strkek• 8 years ago

Worse of all is that this abbreviation must have come from someone who had no idea what cryptocurrencies are; someone who thought it was "magical WoW money but better".

Because I just can't wrap my head around the fact that anyone who knows what a cryptocurrency is, is somehow unaware of all the massive prior art indicating that crypto=cryptography.

It's like calling light bulbs "glowing pears" (jk, I love the German language <3).

crtasm• 8 years ago

When the sitcom Silicon Valley recently used it to refer to cryptocurrency I considered the battle lost.

twblalock• 8 years ago

I expected this to be dumb, but it's a very perceptive article that makes a very good point: There are a lot of benefits to centralization, and people who want to create decentralized systems need to recognize that and take the necessary steps to ensure their systems do not become re-centralized.

benmdi7• 8 years ago

Thanks! Glad to have surprised you. That's pretty much it, yup!

mLuby2• 8 years ago

This happens because `distributed < distributed + centralized` nearly always.

For example, git is great, but git plus an (optional) issue tracker has more value, so that's what people use.

Coincidentally (or perhaps causally?), FOSS < FOSS + closed source.

The only way I can see this changing is if the distributed (or free or open-source) benefits of the system fail early and painfully when centralized (or paywalled or close-sourced). I can't think of any such systems off the top of my head.

benmdi7• 8 years ago

Yeah, sometimes I worry this pattern is inevitable as well. At the very least, though, I'd like to see us paying attention to the issue this go-around with crypto/blockchain

davesque• 8 years ago

This was actually a great article. So much crypto centered content nowadays is garbage. Very pleasantly surprised by this. I think there are even a lot of interesting related questions that were not explored in this relatively concise post.

Though I don't have any problem with them today, services like Infura are definitely analogous to GitHub and may eventually become weak points in the same way.

benmdi7• 8 years ago

Glad you liked it!! Indeed, I could have written much more, but it felt like it was getting long as it was :)

nemild• 8 years ago

I've lived through a few eras, where open and/or decentralized systems were expected to win (Internet, crypto today). For most users, the critical things are usability and utility, not decentralization.

If we care about decentralization, we have to care about these points even more, to make products competitive with what users are used to. Even though many engs I know value decentralization, most people won't choose systems for this reason alone.

Just like with Github, there was a bunch of anger before when Slack was displacing IRC. I wrote some thoughts about what we/I could learn from that example:

What Open Source Can Learn From Slack

https://www.nemil.com/musings/oss-and-slack.html

If you really want to get more cynical, Tim Wu's book "The Master Switch" is a masterful look at 20th century technologies (radio, telephone, telegraph) going through the idealism of the early days, to the inevitable frustration when it creates new anti-consumer behemoths.

benmdi7• 8 years ago

Indeed, indeed. It's a tough issue and I do sometimes slip into cynicism about it. At the very least, I hope by being aware of it we can mitigate the worst outcomes, even if we ultimately always see some degree of re-centralization.

phicoh2• 8 years ago

"With this new set of powers, we saw a Cambrian explosion of open source that coincided with git's adoption. [...] That's where GitHub stepped in, providing elegant, centralized solutions around all of these new problems."

Did the author live in a parallel universe? There was an insane number of small open source projects way before git was created. And there was a clear place where to find them: SourceForge.

The cool things about git are the ability to sync repos and the ability to handle merging branches way better than SVN.

Everything else we had. Sourceforge dropped the ball and that allowed github to take its place.

benmdi7• 8 years ago

Haha, no I was in this universe, and am old enough to have participated in that era as well ;)

I think it's pretty clear, though, that both the quantity and quality of open source software has increased drastically in the last decade or so. I think it's fair to say I may have stretched the degree of causation git had, as opposed to just correlation. But I also think it's pretty evident git, as a tool, had something to do with it.

phicoh2• 8 years ago

(Of course this anecdotal) I see no correlation which source code control software.

What I think is key for open source is the quality of internet access we have these days. You can copy any repo without really thinking about it.

The other aspect is that open source is extremely mainstream. Mobile is still a reflection of the old world where everybody tries to make a small amount of money which some closed source apps.

Lots of people realize that is it is better to collaborate with other people in an open source project then to be on your own writing your own programs.

All of that would work just as well on SVN.

eesmith2• 8 years ago

I am old enough to have participated in the pre-SF days, and can say that "the quantity and quality of open source software [...] increased drastically" with SF.

Ditto with Python.

Ditto with pretty much everything computer related.

It's as if the extensive shift in education starting in the 1980s or so, to emphasize training with computers and programming, has lead to a large number of well-trained, competent programmers.

flyblackbox• 8 years ago

I am really upset about this acquisition.. And equally or more concerned with centralized power points, maybe to an extreme that is too radical. But I do believe at it's core, crypto is different because it's core ethos is to reject that notion. So while the internet disrupted centralized broadcasters with a decentralized protocol for content distribution, it wasn't baked into the technologies ethos. So in my opinion, this difference of core mission, along with the ability to launch dApps (anonymously even, if you'd like) without fear of being shut down, will be a differentiation that protects this technology's fate from creating new centralized choke points.

Ie. if Sean Parker was driven by the mission/ethos of crypto, Napster couldn't be taken down, people would still be using it today, and the traditional music industry would be dead (as well as traditional tech companies like Spotify).

hudon• 8 years ago

This article would make more sense if crypto currencies were actually decentralized... that ship sailed maybe five years ago? Today, only a handful of corporations control the majority of the proof-of-work hashrate [0]. What’s more, development teams represent another centralized group (ie. you can count the devs that contribute to consensus code on 2 hands if you’re generous).

So my question is: in what way are crypto-currencies not already in the state the article is warning about? If I capture the lead commiter to your chain as well as the lead miner, don’t I effectively own the chain?

[0] “Both Bitcoin and Ethereum mining are very centralized, with the top four miners in Bitcoin and the top three miners in Ethereum controlling more than 50% of the hash rate.” http://hackingdistributed.com/2018/01/15/decentralization-bi...

fwdpropaganda• 8 years ago

Very insightful article, thank you.

So the cycle is:

- problem exists

- quickly centralized solution to problem appears

- new problems appears

- slowly decentralized alternative solution to old problem appears

- repeat

I would add that it seems that the big winners are the ones coming up with centralized solutions. The inventors of decentralized solutions don't have nearly as great rewards.

What the Crypto Community Should Learn from GitHub's Acquisition

Related Stories