Alpha channel is big advantage (worth it in its own IMHO), but even a small size decrease adds up to a lot of bandwidth for high traffic sites, especially any with a lot of image content.

But it does negate the suggestion that we shouldn't replace JPEGs with something smaller just because JPEGs are already "small"...

The BBC news site gets 40 million unique users per week and their homepage contains around 400k worth of JPEGs.

If BPG reduces the size of those images by 100k and If each of those users loads the homepage just once, that would save them 570 gigs of bandwidth per week.

Not to mention the saving for users with bandwidth caps on their connection...

Lossless compression works terribly on photographic images, which limits the kind of images you can practically use with an alpha channel in web browsers today.

A DSLR has managed to detect a large planet in a fast orbit around a small, close star. Kepler is sensitive enough to detect earth-sized planets orbiting G-type stars at 1AU, A DSLR (or even conventional telescope) can't replicate that.

I suspect most (all?) of the transiting planets that today's DSLRs could detect have probably already been detected by sky surveys anyway.

But also of observation (that bacteria were killed by mold) and methodical experimentation (isolating the mold, extracting the antibiotic chemical and performing control trials on animals). Use of Penicillin wouldn't have happened without those further steps.

But your example is about fitting various pieces of evidence together to come up with a theory that challenges previously held beliefs, our AC friend at the top there seems to have missed that bit out. It doesn't matter how true something is, if there's no evidence for it then it's not scientific.

Most archaeologists seem to have come to the conclusion that the Baghdad battery probably wasn't a battery after all...

"shite"? That's far too British!

But if you're going to talk about worthwhile spending then maybe not spending ~$700 million per day ($100 billion every 140 days) in Iraq on a war that increased global terrorism is a better place to start?

But from what I've read it works with the same photosensitive chemicals that are used for etching silicon, so could both be used instead of a mask (I'm assuming they still have to use masks to etch the silicon?) and maybe add polymer parts to the silicon components.

What about shielding? The missions that used SNAP-19 power sources kept them on 3m booms to prevent their radiation from interfering with the scientific equipment.

Given the sculptures were made using lithography I can't see why the same technique couldn't be applied to MEMS.

Too heavy though, the weight budgets for space are brutal.

SD cards can't impersonate a keyboard, so anything like the USB firmware hack you linked to is impossible. There could be malicious files pre-installed on the drive, but then that's happened to big name suppliers plenty of times too.

As far as I know Android has no facility to run code directly from an SD card anyway, and if you're using an antivirus package worth its salt on your PC it would block any autorun attempt.