The problem with offering the buyout is that not all users have equal value to Facebook. The people with the highest value - ones who use Facebook a lot and are demographically desirable - would take the $1/month buyout and those people are probably worth more than $1/month to Facebook.

All of the original Big Three car makers and many other car makers offer their own lease programs. In that case the car IS reverting to the manufacturer, or at least to its financial branch.

This time I think there is some chance that the public opinion will matter, because there is big money weighing in on both sides so the public opinion could tip the balance between the two.

There are two big money arguments in the net neutrality fight, one on each side. In favor of neutrality you have internet companies like Google and Microsoft, who argue that allowing ISPs to charge fees to service providers (beyond the normal cost of their access) would stifle innovation in network services. Against neutrality you have ISPs like Comcast and Verizon, who argue that forcing them to carry unlimited data for everybody without being able to recover their costs would stifle investment in network infrastructure. The FCC has to decide which of these arguments is the more convincing one.

The situation in past battles, such as the one about consolidation of media companies, was a very different one. In those fights almost all the money was on one side; it was a battle with big corporations on one side and consumers on the other, and money tends to win those. Net neutrality has big companies and big money on both sides.

Qubo is a broadcast network; it's on the air as a digital sub-channel anywhere there is an ION network station. (Though there are some major cities with no ION station, and some cable systems, notably Comcast, don't carry their sub-channels.) But that's a bit too non-mainstream for Gizmodo to have paid attention to, and in any case the programs are forgettable.

The storage picture on the latest crop of tablets is a lot better. Windows 8.1 can boot directly from the recovery partition, eliminating the need for a second (and larger) uncompressed copy of it in the file system. There are now Windows 8.1 tablets with a measly 16GB flash and you still have some space available to use.

Current Atoms (Bay Trail) are 64 bit capable. Many of the laptops using them run the 64 bit version of Windows, because they have to in order to use their full 4GB of RAM. Bay Trail tablets are mostly running the 32 bit version because they only have 1 or 2GB RAM; they could use 64 bit Windows but the memory and disk space requirements for the 32 bit version are a bit lower so they use that.

OS/2 also failed because of the inflated pricing of RAM at the time. There was a period in the early 1990s when RAM prices failed to decline in the usual manner of computer components. (There was a second period of that circa 2000 that eventually led to a price fixing lawsuit and settlement.) Sadly for the fate of OS/2, this period of high RAM prices coincided with the introduction of the OS, and those high prices made the adoption of OS/2 unappealing.

OS/2 did a number of things that Windows did not at the time: full 32 bit code, preemptive multitasking, a virtualized DOS compatibility box that was protected from crashing the entire system, and a technically superior file system. Windows didn't catch up until the release of Windows NT, and the initial releases of NT had problems running many existing Windows programs, while OS/2 could run them properly; that wasn't addressed until Windows 2000, and then XP finally got it completely right. Doing all that extra stuff meant that the OS needed more memory.

I really liked Ami Pro. But Samna/Lotus never came out with a proper 32 bit version of it, instead giving us the atrocity that is Word Pro. Word Pro tried to be an imitation of Microsoft Word, abandoning the things that Ami Pro did well (especially its use of style sheets), and it was horribly slow.

Multiplan was also available for the Mac for a short time. But it was just a spreadsheet that ran under the GUI, rather than one that took advantage of the GUI as Excel did. Excel added things like full GUI styling of text and dragging of column widths and row heights, capabilities that Multiplan didn't have, plus it had much better graphing features.

For many things, yes. But the 8088 had a built in way to address more than 64KB of RAM (the CPU's limit was 1MB, but the IBM PC used some of the address space for I/O so the RAM limit on that platform was 640K), which gave it an edge for spreadsheets. Yes, it was convoluted; the segmented memory model was a pain. But it was better than nothing, or than a variety of inconsistent ad-hoc methods that 8 bit microcomputers acquired before they became obsolete.

I don't miss Reveal Codes. The need for it showed that the character decoration model of WordPerfect was fundamentally broken. A word processor that got the model right, like the late lamented Ami Pro (the model there was essentially like Cascading Style Sheets on the web, but long before those came along), had no need for Reveal Codes because you always got what you expected.

Another misstep for 1-2-3 was that Lotus was focused on OS/2 development, not Windows.

But most people were editing smaller images back then. We didn't have 24MP and 36MP digital cameras yet, nor did we have cameras with more than 8 bit color depth. A 24MP image takes up 72 megabytes in standard 24 bit color and 144MB if it has high dynamic range color; a computer with a total of 64MB is not going to be able to edit that image efficiently.

Desktop processors don't have the necessary logic for synchronization of more than one CPU. (The CPUs have to talk to each other so that the on-chip caches don't cause data inconsistency between them.) Server-class processors (Xeon and Opteron CPUs designed for multiple-socket motherboards) have it, and adding that extra circuitry increases the manufacturing cost. Intel and AMD are probably overcharging for that logic because they can, but it certainly does add to the complexity and cost of the chip.

And that 2009 Mac will probably only be able to run the current OS X version for a year, until OS X 10.10 comes out and support for it is dropped. The average lifetime these days for Macs to stay in OS support is five years, though some models get a little more and some a little less.