It'd be nice to know if Mint customers will get the FULL T-Mobile coverage area once the merger goes through.... Currently, Mint does NOT offer ANY coverage in Alaska, Canada, or Mexico, while T-Mobile does... It would be nice to at least get Alaska. You can see this one their respective coverage maps.

That comes to $1377 per household they want to serve (642,925 of them).. They could send an install kit ($499), and pay for a discounted first year of service ($877/yr, instead of $1188/yr) for EVERYONE of them. So everyone of those households would have free high-speed internet for a year...then have to pay prevailing rates after that.... That's SO much better than ANYTHING ANY cable/cell company has done.... Historically, they have used the money to build-out in near-to-suburban areas, then charge out the nether-regions for any service (even at an unreliable 2-5mbit/sec service).

Has NASA done any in-space studies of artificial gravity from centripetal force (or is it centrifugal?)? Many books and movies have done real-studies on rotating sections (2001, Rendevous on Rama, O'Neills space colonies, etc...). Even some pretty good math support on how long the lever arm has to be vs rotation speed.... I don't know how they would attach it to ISS, but it sounds like some kind of testing should be done.....

Nah. I'm in Portland, OR, land of rain and 'sun breaks'...and I just got a 7.5Kw array put in, on multiple sloped surfaces on my roof, only 1 of which faces south), and I can still generate an average of 43-46 KWh per day (obviously will probably do half that this next winter). Yes, I still suck from the utilities teat at night, but I COULD replace that with batteries, but not worth it yet. I EXPECT to be 'net-zero' over a years time (no winter data yet)... Putting it all in a 'single point of failure' just makes it less efficient (from both a technical POV....getting all that juice transported to the rest of the country, and from a security POV...it's all in one place for when the meteor/nuke/MOAB/WMD/terrorist plots take out the distribution/transportation hub)...... I'll take the inefficiency of 'distributed solar' any time.

Staying? Why would you....As a loyal Sprint current customer, you DON'T GET this plan.......yoyu can't switch to it, only it's non-discounted version (for $160, as quoted above)....

BIOS isn't really the problem......it's EVERYTHING else.....

Sure, the BIOS may take 5-10 seconds to get through the memory setup for 32GB of memory, and another 5 seconds to run the pre-memory part of TPM (and all the rest of the TPM measurements at various points in the BIOS POST processing), and another 3 seconds to do a keybaord reset (to keep the old PS2 controller capability, then another 3 seconds to setup and discover all the USB devices behind all the hubs they are putting in the front panel and back panels....that ain't the frustrating part....

it's the LSI Raid controller taking 45 seconds to find the drives it knows it controls (try a PERC 5 on a DELL server to watch that one).....

it's the NIC ROM's that want to tell you they can boot or get into their config ROM, and wait 5-seconds to tell you that EVERY boot, (even though most modern systems have a net boot hotkey...even most servers),

it's the Management engine's option ROM that takes 5-7 seconds doing 'stuff' before asking you if you want to get into it's 'info' screens for another 5 seconds...

it's all the stuff you have to do to setup ACPI structures for the numerous various flavors of OS's particualr 'idiot'-syncracies. They all want 'their' versions of ACPI structures, so you do all of them (and special case every darn one of them).

Then you throw in EFI, and it's OS-like layered-to-inifity-and-beyond 'driver' structure that wants to hot-load EVERYTHING, and the fact that FEW card vendors are even going to support EFI option ROM's for their cards in the near future (and how long are THOSE ROM's going to take to run...45 seconds to find my drives is bad enough).....

We need to hold these vendors feet to the fire to give us fast booting machines (heck, my DELL WS670 with 8G of memory, and 2 Dempsey Xeon's is done with POST, and waiting for my 2 - 1T drives to finish spinning up (in 11.8 seconds from power on) before launching any OS I want, barely beating out my LCD panel...)
But it has to start with EVERYONE involved, not just the system BIOS vendor....until option ROM's are given a 100ms limit (like the big push years ago on video cards) to run, it isn't going to get better (especially with EFI).

I'd love to push it all into the OS (as some suggest), but the work for the 'boot' device has to be done before that happens, and all that setup isn't trivial, and once there, the OS's aren't good enough yet to handle it all....

Someone has to do the work, and I don't trust the OS (considring all the rest of the stuff it tries to do) to do it any better than a bunch of dedicated BIOS programmers that are making sure it works with THIS HW, not some generic OS code that HAS to work on everything from a Compaq luggable (circa 1991) to a quad-core quad-socket server...

Expecting a bunch of C coders to write EFI code efficiently in the barely-debuggable EFI infrastructure is not realistic...

The GOOD BIOS coders have to come from somewhere, and they need to be focussing on more than just the system BIOS. The peripherals make it worse. The legacy leftovers hold us back, but that's not the real problem.....

Off my soapbox.