timmb

What does spearfishing mean?

That was the scariest term they could come up with? The romantic swashbuckling seafarer, antagonising empires and declaring his own rules? Surely they could have done better.

'software fraud', 'software theft', 'forgery'... hmm well there's a reason I don't work in marketing obviously.

By stripping and selling off as many assets as you can and borrowing as much as possible leveraged on the income from future fares. There is no incentive to spend this on long term investment or contingency planning. The railways are an essential service so the government will always have to step in when the private company reaches breaking point, so no money is saved by the state in the end. It’s just disguised debt with terrible interest rates.

I don't know if relevant for Wikipedia, but in CC BY-SA 3.0 where the entire licence was invalidated if you made a mistake with the attribution.

I was reading of an exploit where an organisation is publishing photos under the 3.0 licence, searching for minor infractions and then extorting random bloggers and businesses for infringement. I believe the 4.0 licence addresses this by allowing users to fix a faulty attribution within a reasonable time.

Pretty sure I’ve heard people use “he” more than “she” in parts of England, especially for tools and, I think, furniture (“see that table - squeeze him in there”). I would guess there is a gender role being attributed to the object.

Sounds like a cause for impossible to find bugs if the wrapper decides to reuse the callback in rare cases likely only existing in production. Might be better to let the end user handle this scenario in case they have side effects in their callback, like std::move.

I guess with email you can at least use your own domain with a third party provider because it has a DNS record, which gives you somewhere between self-hosting and a federated community. Of course, if you lose your domain you're out of luck.

I feel the real problem with Mastodon is that there's no way to reclaim an identity if a server goes down. The idea of a fediverse is to be robust against instances collapsing, but it seems the only way to move server is to leave a pointer on the old server.

Instead, you could put a public key on your profile and keep a local backup of your data and private key. Then if your instance dies or you get booted out (there will always be tyrants and schizms) you can still set up shop somewhere else and ping your old followers with a signed message.

Right now it's like a city with a lot of bars which is great except you can't leave and go to a different one with the same face.

My understanding is limited, but I'm guessing it's to do with the algebraic properties of the embedding vectors. I'm familiar with embeddings that you can add and subtract, which may reveal concepts existing as linear directions within the embedding space.

Here, they're talking about multiplying, dividing and permuting vectors. Multiplying combines concepts, adding creates a superposition.

They also mention randomly selecting embeddings for the concepts in mind. So my guess is that instead of one-hot encoding the classifier, they instead use random encodings on the output, and are working to give those encodings desirable properties.

I would also hazard a guess that the random vectors they choose are close to zero in most components.

I misread this as Nutshell. Slightly disappointed that wasn’t its actual name (perhaps it’s already taken).

Obligatory xkcd: https://xkcd.com/1296/

"that AI will be the death of learning & so on; to this, I say NO! My student brings me their essay, which has been written by AI, & I plug it into my grading AI, & we are free! While the 'learning' happens, our superego satisfied, we are free now to learn whatever we want"

Attributed to Zizek by https://twitter.com/LuminanceBloom/status/160059800339126681...

For cooking, non-virgin coconut oil is cheap and flavourless. The expensive virgin stuff is flavourful and nice on toast but not good for cooking.

In the uk I often find non-virgin coconut oil hidden in the Caribbean section of a big supermarket.

Washington Post not the NYT no?

Just curious but is there really a risk on public WiFi if you’re using DNS-over-HTTPS and connecting to a site over https?

But software is faster to write and so in theory cheaper to buy.

Had a similar experience with AnyDesk, who quietly decreed that they need 3 months notice before the end of the year otherwise you're tied into another year. And there are zero options to make contact with them.

It's a shame because I occasionally need something with that functionality and would otherwise have happily renewed when I need it.

Most chilling for me was the majority who would divert the trolly away from a good citizen to kill a citizen who litters.

Well, except the obvious example of the time for which nobody seems to seriously want to move from base 60.

I was contemplating this the other day. We coders often bang on about the joy of base 16 as pure 2s but what makes 60 so useful is it has so many factors: 2,3,4,5,6,10,12,15,20,30. It means you rarely need to split a minute when you’re dividing up time for everyday use.

I’m used to seeing for(;;) in C code but I disagree that it’s conceptually similar - the opposite actually. The for() syntax is already inconsistent with three expressions being interpreted as statement;guard-expression;statement.

There’s no other occasion I can think of where an expression becomes vacuously true by being omitted. If it were conceptually coherent you’d write for(;true;). I guess maybe the reason they made C accept for(;;) but not while() is because while() might confuse a compiler by looking like a function call.

Hmm sorry, I seem to have lost the link and now HN won't let me add it. I'll resubmit it in a bit.

If you don’t want to offer the freedoms of a licence then pick a different one. To choose one with an established meaning and then remove those freedoms undermines the purpose of having standardised licences. The authors are potentially setting users up for unintended infringement.

I see it as similar to organic food. Once you add non-organic ingredients then it’s not organic any more. You can do what you want but it’s fair enough if everyone else asks you to respect the labels they’ve built to let people make informed decisions.

I believe it’s the imperative (enable widget) not the infinitive (to enable widget) that is generally used.

Now, persuading people to paste your keyword dump into the HN page discussion for your site. There’s some real SEO for ya ;)

I blitzed through the author’s book 4000 Weeks (on audiobook) earlier this week and it was excellent. It addresses many of the objections to commenters here better than might fit into the Guardian article.

One of the main points I took from it is that a failure to acknowledge your limits can actually undermine what you’re trying to achieve. For example, deferring the things that truly matter to you (personal projects, time with family) until you’ve got a hold of your endless todo list will lead to a life perpetually yearning for a fantasy future where everything is somehow under control.

Similarly (my example), fixate on finding a romantic partner to the point where you make yourself miserable until you succeed and you may find your despair makes it harder to connect with new people.

The key message for me is that this day today is your life. Work hard towards your goals but remember to enjoy that journey.

Looks like there was a lot of Rothko in the training set

It’s worth reading the peer reviewed paper in the Lancet. It’s been a PR disaster which is significant when you’re talking about persuading people to take vaccines. But from what I see:

- a mistake was made on standardising measurements, but a decision was made that erred on the side of caution (diluting the batch that may have had too high a dose). I can appreciate these decisions being decided based on balancing what is needed to measure efficacy vs. the number of additional people who die due to delaying the trials. In hindsight it was wrong and something apparently simple was overlooked but the decision process seemed to have integrity and be based on what was known at the time.

- The accidentally modified dosing trial was just tested on younger people but the improvement was the same even if you only compare it against the same age range on the original dosage. This dosage I guess needs more data which will probably be forthcoming. However the original dose efficacy of ~60% is still useful given the affordability and practicalities of distribution.

- the validity of the pooling of the data was approved with the regulator well before anyone knew how it would affect results

- However a key thing looks to be that the reduced dosing increased the vaccine’s efficacy on Asymptomatic cases from ~4% to 60% - But at this stage more data is needed because to give a confident picture. This is big news as this is the only vaccine so far with an indication of efficacy against asymptomatic.

Please correct if I’m wrong in the above?

The paper in the lancet: https://www.thelancet.com/journals/lancet/article/PIIS0140-6...

There is a lot of variation among native English speakers. The only thing that sounds non native in what you wrote was “if you spot out anything”. I would always expect “if you spot anything”. But the fact that other posters picked up on different things suggests this could be regional variation (I’m in the UK). I know people who have lived and worked here for a decade and still feel they perform worse in a job interview in English rather than their native language due to the extra cognitive load. On the other hand, my partner has lived here about 10 years since she was a student and finds many things (e.g. work stuff) she can express more easily in English while other things (how she’s feeling) in her native tongue.

I speculate that you get better when you are pushing past your limit, whereas comfortable practice stops you from regressing. For example, being in a situation where you are too tired to speak English but you have no choice. Also, you could try more challenging forms of text such as poetry or older English literature. While I doubt you have need to speak like Shakespeare, forming a bit of fluency with 16th century English may prime your mind to be more aware of the subtle differences you hear in modern English. Anyway, just speculation so take with a pinch of salt!

While I have all the bloggers - please can I ask you to remember adding an RSS feed to your blog? This is what makes it possible for those like me to keep returning for the occasional new post.