Is this ripped using the retroplasma flyover-reverse-engineering code? The retroplasma repo is a bit behind now and doesn’t work with the updated way Apple serves map data and the heif texture format.
I’ve been playing with the same thing to make RC tracks in re-volt (or rather RVGL). This game though, running the game as a webpage is fantastic!
Looks to me like a weird combination; maybe the spat is trained off ripped Apple Maps data? The surface textures all look static like traditional photogrammetry (e.g. no specularity), but there are definite splat artifacts rendered.
Actually, ripping the data might not be required with this approach; instead the splat could be trained with a screen recording of normal Apple Maps usage.
Interesting, it does definitely look like the ground is just a mesh. Also, it looks like apple maps has actually recently started using some sort of splatting - I wonder if their implementation is entirely splatted, or if it is this exact weird combination.
Do you have coordinates for where splatting is used in Apple Maps?
It does seem like an obvious next step towards realistic maps, could be difficult with weather conditions changing during capture. A brute force approach though could be 1000 drones to do capture simultaneously.
Pretty sure Fifth Star Labs’ Sky Guide iOS icon does this. Not the whole icon, just the stars in it. Hiding the app in a folder helps but I’m looking for a better alternative just because of this.
As a workaround, you can use a home screen shortcut to start the app.
But I agree that app developers shouldn’t do this, or should at least provide an option to turn it off. (Some apps allow customizing the app icon color, so it seems that should be possible.)
That’s a good idea, I haven’t used the shortcuts feature. I think this app also abuses the ability to change the app icon to provide subtle animations - but there’s no option to turn it off or customise.
What does it use?
A reason I switched was to stop localisation of searches, and the dropdown at the top of the page used to be great for that but now doesn’t seem to do much.
For example if I search “pergola” then I get local pergola providers: good.
But if I check the box next to my country flag (I.e. to turn localisation off) I _still_ get local pergola providers. Even if I change the region to another country I still get local results (though further down the page).
My country isn’t well known for pergolas as far as I know, and the effect looks bigger than the search reacting to language differences. So why do I still get local results?
You still get local results because that toggle is just one of the signals we use to get search results. Turning it off basically just says “don’t send this country as a signal."
If you selected a different country from the dropdown next to that toggle, instead of switching that toggle off, I would expect that to return results from the selected country instead of showing local results.
I would also expect that but that is not what I see. There are the ads at the top of the page which I do expect to be localised, but further down the page I get local results for my (quite small) country which I would not expect to get if I was in the country that I’ve selected as the search country.
Is there a way that I could use ddg that would return the same result for me as someone else would get? Not someone specific, but a an unbiased result so that I can know that if someone else searched with the same (lack of) parameters they will get the same results?
Some Countries and legislations radically ban reading private communications. So, there is no linear, direct assumption that the email provider abuses the role.
But there is a juridical mess, already evident in matters like the eu's "ChatControl" laws (exemption to national rules about privacy for communication providers).
An interesting twist on this game might be that spelling is enforced to be non-standard. Here are letters, here’s the meaning, make the word but it must not be spelt right.
No, they said that in Australia words are made up. I'm not disagreeing, but I'm saying that it's not specific to Australia!
Tangentially, it's funny to me that you happen to use "spelt" here; I've always felt that it doesn't look like a "real" word and choose to use "spelled" instead, but I also recognize that this is entirely my own personal thing and not reflective of anything other than my own issues. It's a great example to me though of how trying to push personal feelings about words against a wider consensus mostly just leads to frustration without any real gains, so it's less frustrating in the long to be able to live with the fact that they won't always match up.
Language is weird and wonderful and despite often being inconsistent and sometimes outright annoying, it's also extremely powerful, and it works well maybe because of the resiliency it needs to withstand the very real limitations it has, and I choose to embrace it all and enjoy it! (I also suspect is why I enjoy puns way more than average; to quote one of my former coworkers, "words are toys")
I imagine you’re wanting instead something the size of a hat, but smaller fMRI are certainly in development, for example [0] with discussion about the size at [1].
I’m experiencing the same in New Zealand. In-zone houses for high-performing public schools are more expensive than adjacent houses (I.e. out of zone) by almost exactly the amount of the local private school fees. House price scales with the number of bedrooms in the same way that private school fees scale with the number of children enrolled.
So the question is; is it more beneficial for the children to get the more expensive education or to inherit the more expensive property?
Side note, “public school” here means state-owned and “private” means “owned by private individuals”, which I have heard is the opposite nomenclature to what’s used in the UK?
The point is that non-deterministic outcomes can be a solution to the problem.
We are at the stage that Donald Knuth is releasing literate programs with LLM assisted proof. They aren't stochastic parrots. They are a useful additions to the thinking mans arsenal of tools.
The market loves it too, MSFT is up 6% this week.
There are likely other things that drive this but my hope is that the share price is increasing on the idea that these patches are so disastrous that even Microsoft should be able to see that their current strategy isn’t delivering a good product.
I’ve been playing with the same thing to make RC tracks in re-volt (or rather RVGL). This game though, running the game as a webpage is fantastic!