RedOrGreen

Very nice, even with no planes overhead. But it looks like the map orientation is reversed (off by 180 degrees)? Real bug, or just me?

Oof, yes, I didn't even mention corporate compliance.

Paraphrasing: "If you don't want to, don't sideload apps, no one is forcing you to" - do people really not see the problem? It's not that technically unsophisticated users will want to sideload apps. They don't know / don't care / have different things to worry about.

But they want their Facebook. Or SnapChat, or Insta, or TikTok, or whatever.

Once other app stores are allowed, there's nothing stopping Meta (for example) from revoking their existing apps, and requiring the use of the Facebook App Installer for access to Facebook. They've paid people in the past to use their Onavo VPN app to bypass Apple's privacy controls, so this would be unsurprising.

Of course, that's just an example; replace with the next SnapChat, TikTok, whatever. If that's the thing that teens want, and the way to get it is to click a bunch of "Yes I agree" dialogs, they'll happily do it. And now suddenly some developer has access to all your family financials through your teen.

If your solution is "well, people shouldn't do that then", you might not understand teens. (Or grandparents. Or regular people.)

> If the full hundred hours was such a huge risk, why not take, like, 5-10 photos first to see if you get much of anything? Seems like they would have been able to know pretty early on whether it was going to be worth it.

This has been answered already but just to pull some threads together - the suggestion to take a few exposures first and see what comes out is precisely backwards. They already knew, going into it, that what they were looking for was at the limits of sensitivity, so nothing would show up until they accumulated enough observing time.

In the absence of systematic effects, our signal to noise ratio improves as the square root of integration time. Think of a bucket accumulating photons - the signal accumulates in linear proportion to the integration time, while the RMS fluctuations increase as the square root of the integration time, so the signal to noise ratio improves as T/sqrt(T) = sqrt(T).

So the idea was to accumulate enough exposure that the faint galaxies would become visible at the telescope performance limits - can't get to something scientifically useful by imaging only the first few exposures. It might have made a pretty enough picture, sure, but not an informative one, compared to what we could already do from huge ground-based facilities.

This is a comparison of images, not just catalogs of sources in images. If you look at the image in the linked article, you can easily spot one of the stars that has "gone missing".

Looking at the brightness of the source that went missing, it is hard to suggest instrumental error - much fainter stars remain detectable in both epochs.

And fraud, really? Who would bother to fake stars in catalogs and plates in the 1950s, and why?

We measure the energy of signals received at our radio telescopes - our units are Janskys, which correspond to 10^-26 watts per square metre per hertz of bandwidth, and we can calibrate our telescopes using noise diodes or known sources in the sky.

So for FRBs, we get the received brightness, and if we can measure the distance, we can use the basic inverse square scaling to estimate how much energy it had at the source. (Why inverse square? Because the area of a sphere is 4 pi R^2 - if you go twice as far away, the signal is spread out over 4 times the area, and so on.)

The key problem is a distance estimate for the FRBs - we have a pretty precise measurement for FRB 121102 because we identified its host galaxy and measured the redshift to that galaxy, so we know how far away it is. That's now been done for 2 other sources. For the rest, we can use the pulse dispersion, which measures the integrated column density of electrons along the line of sight, along with models for our galaxy and the intergalactic medium, to get some idea. It's not as good, but better than nothing.

Calling it a "really fast camera" elides much of the actual difficulty. We're trying to tag individual wavefronts of light at different telescopes, record them, and then play them back at a central "correlator" with the appropriate delays so that the waves come to a focus.

For a wavelength of 1.3 mm, we'd want the time tagging to be better than a quarter of the wavelength at least - say 0.3 mm. The speed of light is 300 mm/ns (a foot per nanosecond is the shorthand beloved of circuit and chip designers). So, for 0.3 mm, we're going to have to get down to a wavefront tagging accuracy of 0.001 ns.

No clock is going to get there, but if we can get ~close enough, we can use a procedure called fringe fitting to determine the clock corrections by looking at the wavefronts. (Does it line up this way? How about this way? How about now? Yes, it's as laborious as it sounds, but computers, eh.)

This is all in the calibration of data, before we do the Fourier inversion to create images - the magic of radio interferometry is that we can record the signal to disk while preserving phase. Optical photons can not be recorded and played back with phase preserved - optical interferometry has to split up the photon streams and send different parts to be correlated against streams from other telescopes, so you run out of signal quickly. Meanwhile, we can record radio waves at the 27 VLA dishes, say, and play them back for correlation on all 27*26/2 = 354 baselines, no problem. That's why radio VLBI is a thing, but not optical VLBI.

Even as a professional radio astronomer, the underlying physics is deep and almost magical.