News for hackers. Stuff that matters.

hjb

Re:ShipMap on Saturday April 03, 2021 @12:20PM (#26682241)
If you go to the 12 of June 2012 you can see a ship making its way across the Sahara. I assume this is a data funny.
Re:Show HN: I Wrote a Book on Data Analysis with Rust Notebooks on Saturday December 05, 2020 @07:47PM (#25319878)
I believe that catching that at compile-time in Rust would have to wait for RFC 2000 to be implemented: https://rust-lang.github.io/rfcs/2000-const-generics.html

This would allow generics that depend on const values (i.e. matrix dimensions in this context). At present generics can only depend on types.

Re:A Raspberry Pi as a decent residential proxy on Monday June 08, 2020 @02:41PM (#23459397)
That is a very retro raspberry pi :-D
Re:Oumuamua may be a hydrogen iceberg on Wednesday June 03, 2020 @04:35PM (#23408347)
The space tug is there to change the trajectory of the other thing:

1. Tug gets in same position and velocity as its target, and docks.

2. Tug performs some kinda of burn to put both it and its target on the new trajectory.

3. Tug undocks and performs a burn to head off to where ever else it's needed.

If I understand correctly, it's really just a way to avoid putting engines and fuel tanks on things that don't need to change trajectory much.

Re:McLibel Case on Thursday May 21, 2020 @02:28PM (#23262718)
This page seems to suggest that's mostly because the solicitor will have barristers they prefer to work with: https://www.barstandardsboard.org.uk/for-the-public/finding-...
Re:McLibel Case on Thursday May 21, 2020 @02:05PM (#23262314)
Yup, to the extent that the US decided that it needed to ensure that UK libel law could not be enforced against US writers: https://www.bbc.com/news/uk-10940211
Re:McLibel Case on Thursday May 21, 2020 @02:03PM (#23262251)
There's a line in there that really stands out to me:

>McDonald's spent several million pounds, while Steel and Morris spent £30,000; this disparity in funds meant Steel and Morris were not able to call all the witnesses they wanted, especially witnesses from South America who were intended to support their claims about McDonald's activities in that continent's rainforests.

It's pretty galling that the justice system, which is a public tool, being gamed in this way by a party with vastly greater assets. It's a bit tricky to see exactly how to fix this, since you don't really want to just give each side a court-appointed lawyer: you really should get to pick your own lawyer. But what to do about the fact that you might want to pick an expensive lawyer?

It seems to me that one compromise would be to recognize that you can legitimately have a strong interest in the case having a just outcome, and thus want to spend more money fighting it, while at the same time not allowing one side to outspend the other.

One way to achieve this would be requiring that each side can only pay into a common pool of money, which is then split equally between the defence and prosecution.

The practical difficulty in implementing this is that a large company (e.g. McDonalds) will have in-house counsel who are not just hired for solely this case, and it would be hard to dis-entangle just how much of their salary should be counted as being for this particular case.

Re:We built a new GPS receiver engine on Tuesday May 19, 2020 @11:39AM (#23236038)
If you dig into their website, they appear to be based in Singapore and Japan. I've no idea if those countries have similar laws.
> One of those weird copyright decisions, like the one where Owens-Corning trademarked PINK

Trademarks and copyright are NOT the same thing. That trademark would only give them rights to the branding PINK in the context of insulation. Copyright is more expansive - and I think you should be more worried about copyright than trademark.

Re:Are you living in a computer simulation? [pdf] on Tuesday April 28, 2020 @11:45AM (#23008124)
Right now I suspect that I'm living in an epidemiological model with entirely unreasonable parameters.
Re:Show HN: A MIX simulator on Friday March 17, 2017 @05:32PM (#13898176)
This is very cool. I've actually been trying to do something similar is rust, but I'm nowhere near as far along (I basically just have load instructions working at this stage): https://github.com/harrybraviner/mix

I love Knuth's books for the fact that they teach you advanced algorithm features 'from the ground up', as it were - it's great to see someone else make an emulator too.