HI version is available. Content is displayed in original English for accuracy.
Advertisement
Advertisement
⚡ Community Insights
Discussion Sentiment
72% Positive
Analyzed from 1423 words in the discussion.
Trending Topics
#companies#data#bluesky#solid#more#banks#need#pds#built#infrastructure
Discussion (97 Comments)Read Original on HackerNews
It immediately makes me think of sovereign citizens and I get twitchy.
The catch is that being government contract you, the guy doing the actual work, are beneath three or four layers of companies and bureaucracy and you get over engineered yet somehow too vague specs and projects that take 6 months just to get approved. But hey, the pay is good, and it’s for one of the better causes.
My other EU client, a much smaller non-tech company for whom I host their servers, has recently wanted to know if we depend on any US services, to reduce their exposure.
I believe you can get decent work just by advertising yourself as an expert in migrating code and data out of the US.
That said, the job and economy situation is a big question mark and appetite to invest has lessened dramatically so YMMV
Could you elaborate perhaps a bit more on this on actually why the appetite for investment has lessened? I'd be curious to know more, thanks!
Fair to say investments and new projects are a bit harder to come by.
The only exception to this rule I would say is AWS GovCloud, which also might be one of the only chill teams to work at across Amazon. It turns out having "only one way to do it", a system proved through a rigorous vetting process and a thoroughly worked-through contracting process leads to a pretty fantastic work environment for practitioners.
Trying to reimplement that piecemeal is for tougher men than me though. I think I'd rather sit on hot nails.
https://xkcd.com/705/
AT Protocol achieved what Solid envisioned without the inane complexities of rdf and json-ld, which were the biggest learning blockers to people actually adopting Solid.
"To use this system, you must understand that we cannot make any guarantees regarding the security and privacy of data that you may store in a solidcommunity.net Solid pod, or concerning the system's functionality and availability."
Related:
ATProto Permissioned Data Proposal Draft - https://news.ycombinator.com/item?id=48651727 - June 2026 (4 comments)
Unless I'm missing something else...
(They are not self-hosting; Eurosky is doing it.)
decentralization is not about the number of app instances but how easy it is to switch from one to another. on that metric bluesky is already better than fediverse.
I've been meaning to move to my own PDS for a few months now. Still haven't. Whenever I decide to get around to it, it'll be fine.
(Think of it this way: "I am following <username>" is a record stored in my own database, so it doesn't matter which app I click the button on that writes that record.)
But I do think it's always worth pushing back a bit on this idea:
> "The way Bluesky is funded is at odds with the idea of decentralisation because the platform relies on venture capital and operates under a shareholder model."
Large decentralized infrastructure like the internet, DNS, email, and the web was largely built by VC-backed companies.
The most important open source project, Linux, is funded by major tech companies through the Linux Foundation, with $311 million last year.
Corporate incentives do create conflicts, so it makes sense to be paranoid and skeptical. But the idea that companies can't contribute to open and decentralized systems is exactly the wrong lesson to learn.
We want more VC-backed startups working on open social networks and protocols. It would be great if many of them were in Europe.
The poor need the rich to start a company as banks are prevented (by the rich) from lending to them.
The rich like VC as it's a tax write-off, they invest in VCs and get even more richer.
Most startups fail, the VC's investors get any leftovers and poor founder walks off empty.
>What about when things go wrong?
In general, if you lose money on an investment, you can offset that “capital loss” against a capital gain you have from something else.
https://www.venturesouth.vc/write-offs
no. the banks hold the poor's money, and it needs to do so without risk because the poor need their money. lending money to start companies that are completly unsecured is too risky for banks, they lend money to buy houses which is secured debt.
Banks often lend at low LTV ratios because the prices are inflated so people on normal salaries can't actually afford to put down a large deposit, which means a slight drop puts them into negative equity but the banks are not concerned as they are protected.
If the state chose to underwrite startups in the same way...
Really ?
It turns out that commercialization is most of the work of creating a globally decentralized system. Which doesn't mean the non-commercial work wasn't critical.
What organizations do you think created the switches, routers, servers, software, fiber optic backbones? Who created the new protocols?
It was companies like AT&T/Bell Labs, Cisco, 3Com, Sun, UUNET, Netscape, AOL, the major telecoms, and a thousand other companies we don't remember.
Something like 1% inspiration from academia and government, and 99% perspiration by people working inside companies.
Nothing, except make it more available.
This is why I often argue against (or at least want to point out the dangers of) the ATProto/Bluesky model.
It's an absolute boon for people who want heavy surveillance, government or otherwise.
The looseness and "unreliability" of protocols like Mastodon ironically make them safer.
There's another protocol in the works that should be useful for syncing private data:
https://github.com/bluesky-social/proposals/pull/94