If it was working perfectly fine until they changed it with no option to use the old rendering, I would consider that a bug on their side. The GPU being able to be locked up is indeed a problem, but one that that I am surprised hadn't been a problem before, like when 10.6.x was new. In researching what the hell was happening, I found that the usual fix for other OSes is to keep one reserved thread (or whatever they're called) running on the GPU that allows it to be reset in case it gets totally fucked up.

They "upgraded" something that was working perfectly fine, without even a configuration option to go back to the old rendering?

Well it's only been a few decades, and I was mostly a Mac user back in the day. I did remember enough about VGA that as I posted, I was wondering where the hell all the color came from, because I was sure that 640x480 was only 16 colors. Oh the joys of cramming a frame buffer through a tiny chunk of a mere 1 megabyte addressing space. But at least I got the approximate CPU range right.

And FWIW, shrinking the screen down (and a coprocessor in the cartridge) was how they got it to run on SNES.

CRS-8 is not a "Bigelow Aerospace" mission, it is a NASA ISS mission which will carry the BEAM module.

But at least you still have Skylab.

The next launch of a Bigelow module will be BEAM on CRS-8 in late 2015, but it's only 4 meters across and will mostly be unused with its hatch shut, other than to check every now and then that it's still okay, then eventually jettisoned.

The "real" one doesn't go up until 2016.

SpaceX is still going to be lifting people to Bigelow's inflatable stations. They're basically already committed to manned spaceflight, NASA or not. But flying NASA to ISS would certainly help with their cash flow.

I'm pretty sure they've been ramping up production capacity this year. However, that's rocket production capacity (still important) with the current new F9 design, of which one of the goals was production capacity because of all the satellites they need to launch.

So how much production capacity would manned launches to ISS need? They do four missions a year, four launches a year, hence four capsules a year. As NASA requires a brand new capsule for every mission, they have to actually build four of them a year. (But that's also that many less capsules they need to build for non-NASA missions if they reuse the ISS mission capsules.)

Ah yes, now I remember the rest of the story... I seem to recall that a few weeks later I got a good deal on a Power Computing Power Tower Pro, which was a quite nice clone (except that 8.0 locked up all the time, so I downgraded to 7.6 until 8.1 came out), and that's what motivated me to sell the 4400. I hadn't really had the 4400 long enough to know just how bad it was. I kept the PTP as my main non-laptop long enough to get it running (actually more like walking) OS X. I long ago downgraded it to 9.x for when I need the retro (floppies and SCSI), and it's been a few years since I last powered it up.

The PTP basically let me hunker down during Apple's worst years of the PPC, the pre-G3 years.

One word: Performa

But there were still some pretty bad models sold without the Performa mark of crapulence. I once got a Power Mac 4400, one of the ten worst Macs ever, from a CompUSSR and realized how bad it was, so a few weeks later I sold it to a friend. I still feel bad about that.

The original article gets this right. You were expecting a clickbait peddler like IBT to even copypasta it correctly?

So tl;dr: Beads have a coating that attaches to bad stuff. Beads are also magnets and can be pulled out along with the bad stuff by a big magnet.

If the beads themselves are magnets (rather than just being attracted by magnets), they can also attach to each other to clump up on bad stuff better. (This is implied by the microscopic photo.)

You must be new here.