randomblast

You know how the distribution of 2-value vs 4-value for new tiles is a weighted random function? And sometimes you get an unlikely 4 that really screw you over? Have you thought about adding a mode which always creates the worst option of the 2?

You could call it “God does play dice with the universe. They're loaded and he hates you.”

I used i3 for a few years when my main dev machine was Linux. I too am frustrated by the macOS builtin WM's shortcomings. Initial feedback on AeroSpace though: it's utter shite.

I started it when I already had my usual number (many) of windows open across my usual number of spaces (lots) on my usual number of displays (3). First it spent 30 seconds trying to give me a seizure, and when it had finished I'm left with a total mess of a layout. Some windows 30px wide. Some windows not-quite-fullscreen. Some windows just randomly floating without any kind of recognisable pattern.

Now I can't navigate around because dragging a window sends it flying off to the corner of a different display. None of the default keybindings seem to do what they say they should. The focussed window disappeared entirely when I tried resize mode, and I can't find the way out of it because that was the window with the instructions.

I think I'll try something else.

Is the same true for Azure Pipelines? Given Actions forked from Pipelines I'd imagine it would be straightforward.

Kind and caring != understands dog behaviour. 99.999% of agonistic behaviour is driven by fear. What a normal person thinks is a kind & caring behaviour can be terrifying for a dog.

Non-pointers point. Some pointer don't.

Non-retreivers retrieve. Some retreivers don't.

Non-herding dogs herd. Some herders don't.

Most dogs play-fight. "Fighting dogs" have to be trained to fight to kill.

Breeding is clearly a fantastic determinant of physical characteristics which pair well with the desired behaviour, and does contribute a little bit towards behavioural instinct, but (as per the *peer-reviewed long term study* in question) isn't on its own a good predictor of behaviour.

The ~.001%, by the way, is termed idiopathic aggression (sometimes called cocker rage, but again is NOT actually well-correlated with cocker spaniels). Idiopathic means of unknown cause.

Correlation is not causation. There could be plenty of reasons why pitbulls feature highly in the dog attack stats (severity making attacks more noteworthy, type of owner increasing likelihood, stigma causing human behavioural changes making a self-fulfilling prophecy). There is no evidence that genetics is a direct cause of dangerous behaviour.

Did you read the paper?

They derived 8 behavioural factors from the survey questions, and there are a few which are very relevant to dog-human aggression.

#1: Human sociability #2: Arousal level #4: Biddability #5: Agonistic threshold (agonistic is the behavioural term encompassing aggression)

Those are probably the most relevant. There is an important consideration which does seem to be missing from the study, but only reinforces its conclusion:

Most, if not all of these factors can be altered with behaviour modification (training).

You can raise a dogs agonistic threshold with classical conditioning. You can increase its biddability with operant conditioning. You can lower its arousal level with both.

Does training rewrite DNA? Obviously not.

Based on a quick scan, there isn't much right with it. Just lots of people looking at pictures of dogs doing nothing and squealing about how dangerous that particular dog is - based purely on its looks.

Depends on the dog, depends on the child. As per the paper, breed is a poor predictor of behaviour. I wouldn't let most adults in this thread play with my dogs unsupervised, based on the state of the comments.

Science: “Here is some science”

This Website: “NO! THE SCIENCE IS WRONG BECAUSE IT DISAGREES WITH MY ANECDOTAL EXPERIENCE”

No. I wouldn't let a 6 year old play unsupervised with any dog. That would be incredibly irresponsible, and has lead to innumerable unnecessary child & dog deaths.

Would you leave a 6 year old unsupervised with a hungry rat? No - what a stupid question. A chihuahua can cause the same degree of injury (life-altering).

Is there a significant difference in behaviour between a pitbull and a chihuahua? No - this study confirms what has been known for years.

Breed-specific legislation is utter bullshit. It clearly doesn't work, isn't workable anyway (breed is judged based on looks rather than genetics), never had any scientific basis, and has been proven to be detrimental to its own aims time & time again.

Exactly this. Scrolling through the customisation feature list, I just see "more work", "more work", "more work" - and it's not super clear how doing all this extra work is going to help me. And I'm in the target audience for a "technical" search engine; this is a total non-starter for the 99.999% of the world we call "normal people".

The useful ideas in here is the "just my data" search, linking into external providers across siloes. Of course, Microsoft, Google & Apple already have this as long as all of your data is held in their ecosystem.

Others have mentioned the estimate of 270 derived from serial numbering. Nikon have also said very specifically in this case that they are recalling 152 pieces sold since last July. That will likely be artificially depressed by the covid situation.

> Emacs is the most ubiquitous of IDE's.

Sorry, have you been in a coma for 30 years?

You're not from the Manchester to which the article refers. Here, we are definitely Mancunian.

Brilliant quote:

The bridge trust said the proposed planning conditions would not amount to the structure becoming an overly controlled and regulated place, insisting the visitor hosts are “not police officers”. It said that while the visitor hosts would theoretically have the power to seize any banned items, in practice this would only happen with things such as alcohol.

Since when have any powers been granted on a theoretical basis?

Blocked is the wrong word. It implies that Apple had to take some action to stop Flash from working, which would otherwise be the natural way of things.

I liked the original Tower when it came out, and I paid the £50 or whatever it was since it was so much better than everything else. I don't know if it can compete with SourceTree any more though: pretty much all of its advertised new features are already present in SourceTree, which is free.

I appreciate your efforts. Have half a pint of mediocre beer.