Blockchain: decentralisation centralised on my coin.
Hint: the core of decentralisation isn’t about having a billion copies of the same database. It’s about a billion people each having their own, unique database.
PS. In a truly decentralised system, no one makes a billion dollars. *watches all the right libertarians make for the door*
PPS. A venture capitalist would never invest in a truly decentralised system. Why? See PS, above.
@sambiohazard thanks ! I only see this message now. I had updated the keys on monday so it's strange that it wasn't updated yesterday yet. But I don't know what the propagation delays between keyservers are.
Glad it's sorted out now.
@FreePietje yea, thank you! I usually follow that but somehow missed this now
Someone build a RISC-V PC 😎
#IPFS braindump for my hosting idea Show more
So I want to provide static file hosting for others, on subdomains. My idea is to have a webinterface for the user, where they choose a subdomain, and enter an IPNS to their site root.
The server will then pin that, and create the DNSLink TXT record for the chosen subdomain. It will also get a letsencrypt certification, to serve a web gateway over https for the chosen subdomain.
Is this the nicest way to do this? What are the tradeoffs? This is based on my limited reading of the IPFS documentation
RTC is usually an external chip, and I think everyone distrusts hardware random number generation instructions enough as there's a lack of transparency, who knows the vendor inserted a backdoor 🤷♀️
@FreePietje @harding There's this recently announced RISC-V cryptography extension
(https://twitter.com/danpag3/status/1081128184794497024) which might have some impact on bitcoin use-cases (e.g. accelerating SHA256 hashing and large-int operations that are potential useful for secp256k1 curve) but it will be a while until any chips that are generally available implement this.
@FreePietje @harding There's no defined standard for this; Raspberry Pi has its 'hats', the SiFive boards similarly have their way to connect external hardware, but it's all specific to a board—not to RISC-V itself—which after all, is a processor ISA. There's also ISA extensions which expand the instruction set but I don't think that's what you're looking for
@FreePietje @harding If you'd like to use RISC-V for a wallet only, the cheap board (https://www.sifive.com/boards/hifive1, $60) might also do. It is fairly slow (Arduino-like) and has only very little working memory, but enough to run secp256k1 signing.
It has no HW RNG either, though, but lots of I/O pins that could be used to connect something to gather entropy, I guess.
There's no software for this so you'd have to build your own or adapt some open hw wallet firmware.
@harding @FreePietje Correct, the HiFive Unleashed board (Freedom U540 SoC) was that expensive. This is because it was the first run, cheaper RISC-V boards should be on the way!
It's about the same sync speed as a RPi3.
It does not have a hardware RNG, nor RTC. It also doesn't have a USB controller built-in to be able to connect any of these, only ethernet.
I'd wait a bit.
why should you write CLI & terminal software in 2019?
well, there are plenty of reasons, but a really good one is that desktop computing is in a state of relentless decline (backed by megacorp fiat that it be murdered ASAP) and the terminal will still be cockroaching along decades after it's impossible to use any of the GUI paradigm stuff that worked just fine in like 1998-2010.
In case you have connection problems in Joinmarket, you might find this brief note helpful (pls share if you know a place it might be helpful, thanks):
The year is 2019 and I can’t buy a good majority of consumer technology because we lack privacy legislation and consumer protections. Example: it’s absurd that my TV came with spyware that can’t be turned off or avoided; I had to stop it from phoning home at the network level. It also came with an arbitration clause and a clause waiving the right to a class action lawsuit.
#Debian folks: What's the recommended way of installing language packs with firefox-mozilla-build?
laanwj's first merged PR: https://github.com/bitcoin/bitcoin/pull/15139
@kekcoin @waxwing @bitcoinmerges I *think* this was due to a brownout of the github API—what the code to determine "is this the first commit" does is to check the github API "commits" endpoint and see whether the size of the result is larger than one ( https://github.com/PierreRochard/bitcoin-acks/blob/master/src/bitcoin_acks/scripts/send_toot.py#L59 ) in case of an error it might also trigger that…