@alcinnz I would love to see a healthy ecosystem and good UX around feed subscription management. As far as I can tell, there is no feed reader that works the way I want (starting with the lack of RFC 5005 implementations in the wild and corresponding UI for full-history feeds), so I think it's important that people be able to easily try different options so we can collectively iterate on what these things ought to look like.
TIL it is not only possible to have a database where the database servers can't tell which records clients are accessing (this is called Private Information Retrieval and is already pretty cool)... not only that, but it's possible to do that while splitting the database across multiple servers in such a way that the servers can't learn anything about any data in the database! Also you can efficiently identify servers that give you wrong answers, and tolerate faults. https://cypherpunks.ca/~iang/pubs/orpir-usenix.pdf
@ajroach42@Shamar By contrast, Taler exchanges are less trusted than their equivalents in Bitcoin (where currency exchanges also know who you are but transactions are public record); and also less trusted than traditional payment systems including ACH, credit cards, and PayPal, which make no effort at anonymity and may even use transaction data for other commercial purposes.
I just learned about GNU Taler and wow, it's an interesting different direction for secure micropayments. Superficially it might seem kind of like Bitcoin, but very different in details: No blockchain, computationally efficient, a payment method rather than a currency, more anonymous for buyers, not at all anonymous for sellers. Not an anarchist or libertarian system: it relies on government regulators and supports taxing and auditing merchants without revealing customers' identities. 🤔
@enkiv2@clacke@ajroach42@jk I'm given to understand that Canada does require professional licensure for everyone using the job title of "Software Engineer"; universities offer an alternate track called "Computer Science" under the math department for those who want a similar degree but don't want the professional license. I don't know, and would be curious to find out, what impact that's had on their software industry.
I don't think I agree that this is a thesis of that work, though; I'm not aware of CCN literature having much of anything to say about content discovery at all. (I'd welcome pointers though!)
Instead I'd say CCN forces thinking about discovery because it invalidates our existing approaches.
@djsundog I've been thinking for a few years now that indexing/search, in conjunction with chat/e-mail, are going to be the important ways for people to find whatever content they're looking for in the decentralized (mesh or otherwise) future. Globally consistent name systems like DNS can't work without arbiters making essentially arbitrary decisions, but if we can pass around names that are unique but not human-meaningful without breaking usability, that opens so many more options.
@garbados@mairin I've been meaning to dig up something I vaguely remember reading years back: Van Jacobson, one of the big names from the development of TCP in the '80s, got enthusiastic in the '00s about what he called "content-centric networking". Here's an interview transcript with him about it (http://queue.acm.org/detail.cfm?id=1508215) or he gave a Google Tech Talk in 2006 (https://www.youtube.com/watch?v=8Z685OF-PS8).
I haven't re-read/watched those but ISTR he gave good motivation for the idea at least.
The first version was an Excel spreadsheet where once you entered the design parameters, one cell computed a series of commands that, when copied and pasted into Eagle, would construct the patch antenna layout.
Real antenna engineers quoted us $10k per antenna to design those, and we needed 3 of them. So we spent a year figuring out the math instead. Now it's a Python script and $100.
@garbados My current thinking is that Liberapay needs to stay focused on inexpensively aggregating financial transactions―so that no matter how many people you're backing, or how many people are backing you, each person only does one financial transaction with associated fees per month or so.
At the same time, I think they should provide an OAuth API that allows backers to prove they're supporting you to gain access to backer-only content you host elsewhere. Let others build that piece on top.
As a teenager, I got to be heavily involved in building my own bedroom when my parents bought a house for the first time in my life. We couldn't have gotten hold of real telco equipment, but I used an electronics prototyping board in my bedroom closet to wire together the house's phone jacks. The one thing I always wished I had was a professional punch-down block. #geek 😅
@clacke I am 33 years old now and I have spent much of this year being too depressed to get anything done. Thanks for the reminder. 😅
Anyway, sure, some people find just the right twist on an idea to have an enduring impact on how humanity thinks, and we consider them heroes (or sometimes villains)… but let's not forget to celebrate all the other people moving the world forward in less visible ways.
I remembered today, and verified in my chat logs, that I was using the phrase "federated GitHub replacement" in 2010. Maybe in 2017, the year of federated everything, this is a thing we can have? At core I just want to be able to fork a repo from another site, have upstream be able to discover which forks exist, and perhaps be able to create cross-site pull requests.
I was negotiating a contract since February to support my work on my open source C to Rust translator (https://github.com/jameysharp/corrode), but last week I learned that the client painted themselves into a corner in which they couldn't legally let me do work for them in the open. (Don't ask.) Hell no.
I really want to work on Corrode. Should I try Patreon or similar? Any rewards I can think of are things I'd do for free. I just want a recurring tip jar for those who can afford to chip in…