anon1253

They take up farming, coffee brewing, bagle cooking, photography, anything that doesn't involve a computer screen directly streaming nightmares into your brain 18 hours a day.

Oh I can set it to -40º below ambient. But usually I set it to a target of -15ºC (with an ambient temperature of around 10ºC, this winter I might go a bit lower). Not stressing the PEC is gentler on the battery too. You have to be slightly careful when doing cooldown and warmup to avoid condensation and thermal shock (or at least, I think I should be), but I've never had any problems with it.

Also great shot! I feel your pain with the wind, dew and bugs... but it's still an awesome hobby.

Not that I know of. CCDs don't generate a lot of heat, so the cooling only has to be relative to ambient, which peltier can do fine. I'd worry more about things like thermal shock when pouring liquid nitrogen over a CCD than the gains in Signal-to-Noise ratio. Although, there are scientific instruments that get cooled way further into super conducting state.

Not quite sure why this hits the front page. Most serious astrophotographers have been using similar setups for a while now, or ideally peltier cooled CCDs (like Starlite Express, Atik or QHY cameras). More recently cooled CMOS sensors (like the ZWO1600MM-Cool) have also gained popularity. It's a bit of niche market, but for long exposure photography having a cooled sensor greatly reduces thermal noise. And since the signal is very faint you often do a lot of "histogram stretching", bringing out the signal, but also the noise. In addition to having a low noise camera you also often do things like bias, dark and flat callibration, reducing impurities further. Dark subtraction is often known as "long exposure noise reduction" in consumer cameras, but astrophotographes like to do it manually. There are a lot of details to consider when doing astrophotography, but I'm always happy I can set my CMOS or CCD (own both the QHY10-OSC and ZWO1600MM-Cool) to a deltaT of -40ºC.

I thought no, and wrote about it [1]

1: https://joelkuiper.eu/change

Been running Hackintosh as my main development machine for over 5 years. Yes, it's been frustrating. But if you pick your parts right and are not too eager about installing updates it really can be a stable solution. Right now everything is working, from USB 3.0 to audio, to Bluetooth. Even iMessage and Hands-off work flawlessly. Although I'll admit that this has not always been the case, and I've spend many nights in frustration fiddling with obscure parameters and kexts. For me the main motivation is that I really like Unix-like systems, but I also need Adobe software. That means both Windows and Linux are out for the time being. I've tried various permutations of virtual machines, but they never seem to work (especially not on "retina" displays ...). I could buy a Mac Pro, but the hardware is just too overpriced, and since I've started doing some Deep Learning stuff I would really like an Nvidia card (which are oddly still supported, albeit unofficially, by Nvidia on Hackintosh).

Just of out curiosity, how hard would it be to write a compiler that takes JAGS or STAN model files and compiles it to the s-exps needed to use the library?

Yep. Still useful for things like fluorescence microscopy or astrophotography though. There you're usually interested in just the specific wave length, like the (Green) Fluorescent Protein emissions for microscopes, or the Hydrogen Alpha emissions in astrophotography. Rather than working with filter sliders, you'd essentially need to use a different lens for each wave length though …

Interesting. It has been known for a while too that it only takes about a 100 SNPs (single-nucleotide polymorphisms) to identify a person from a large biobank. I once wrote a paper how to potentially combat this https://files.joelkuiper.eu/f/preprint/biobank2015_final.pdf (pdf)