Seems to be an exclusive or. "Ron Paul" or "somebody sane", which implies Ron Paul isn't in the sane set.

My work laptop has 4GB of RAM on it and Windows 7 and it runs just fine. The only thing that slows it down is when the corporate-mandated management scripts run and start pegging the hard drive with virus scans, audits and the like. More RAM wouldn't help that. Switching to an SSD did.

According to Resource Monitor, I'm using about 3GB, with 850MB of that used as cache. A bit over 1GB of that is Firefox.

So, yeah, I could see 1GB really sucking when used with a modern web browser and many tabs open (like I do). 4GB, though, hasn't really held me back much.

Apparently, at least part of Vista's memory woes stemmed from the poorly tuned "SuperCache" feature, that would aggressively try to pre-cache data in RAM. Its appetite was apparently too large. It apparently also didn't manage its disk buffers very well. (This is all third or fourth hand knowledge and so could be shaky. I've never run Vista myself. If someone has more details, pipe up!)

Thank you for the detailed, thoughtful reply, including all the numbers.

Ah yes. I knew there was a term for it. It's been a while since I took my thermo class. Specific heat is normalized to mass, though, and both granite and basalt are more dense than seawater. Granite is ~2700 kg/m^3, basalt is ~3000kg/m^3, while seawater at high depths is around 1050kg/m^3.

When you factor that in, water still wins, but only by a factor of 1.5 to 2.

Allow my naivete to shine: What's the temperature of all of the rock that water is in contact with, and what's its thermal capacity relative to the water? Could it be that it's slow to warm as you need to warm all the rock it's in contact with?

It was implemented against the Win16 API, and required a Windows runtime to operate. It was no less a Windows program than any other Windows program at the time. You could launch it from DOS, sure, but if that's your criteria, there wasn't a Windows program before Win95, the first Windows version to boot directly to Windows rather than launching from DOS.

Oh, yeah, I definitely remember the pain of bank-switching vs. segmented memory. I was there and programmed both. It stunk.

At least the Apple ][gs could directly address 16MB, although the 65816's addressing modes I hear were less than awesome. I must admit I never wrote any native 65816 code.

Oh, don't get me wrong. I definitely prefer a more style-sheet oriented way of declaring how a document should look, and then writing my document tagged in the elements of that style. Something style-sheet oriented is vastly superior to something more formatting-oriented at a lower level.

The problem is that most GUIs, in an effort to reduce clutter, make it at best cumbersome and at worst impossible to determine if your cursor is inside or outside of a formatting tag. (And then there's Word and Outlook, who add a bunch of heuristic behaviors on top of that that just make it worse. I've had to nuke whole paragraphs just because there's some weird 'hot point' in the middle of a line somewhere that keeps making the editor go gonzo. FrameMaker is not as bad, but it gets weird in other ways, especially with figures, tables, their anchors and their captions. And then there's the brokenness of nearly every browser-based rich-text editing widget ever.)

I would much rather have a WYSIWYG preview and a separate, less WYSIWYG editing mode that was more tag oriented. Heck, I'd write my specifications in TeX or LaTeX if they let me. But, alas, I'm stuck with FrameMaker and Word. At least with FM, they force us to never use local formatting overrides and stick to the style sheet. They strip all formatting overrides when compiling a book.

Ah, boundary conditions on the dates. Makes sense.

I had read that Excel had preview copies out in 1984 (which is fairly quick, considering the Mac itself launched in 1984), but didn't launch officially until 1985. And I suppose, since Windows was still deridingly referred to as "Wintendo" in some circles and generally not widely adopted on PCs, it makes sense that Microsoft would go with an "Excel for MS-DOS 3" marketing stance, even though it really was a Windows app.

FWIW, WordPerfect, my favorite word processor of the early-to-mid-90s (replacing AppleWorks and ///EasyPieces on the Apple //e and Apple ///) got subsumed in the same way as Lotus 1-2-3 did by Excel. While Word for DOS was nothing special, Word for Windows actually was a proper Windows application by the time Word 6 came out. WordPerfect 5.1 on DOS was great, but WordPerfect 6 lost it somehow. The DOS version overreached, trying to bring WYSIWYG into the character-oriented DOS world. The Windows version clung too strongly to their DOS traditions and never really embraced the GUI properly. It was a messy disaster. Microsoft Word 6, for all its faults, was at least largely consistent with itself and the Windows environment around it. (I still miss Reveal Codes though, to this day.)

I guess you never ran either an IBM PC XT or an Apple ][. Both were comparable in performance for many things. (A 1MHz 6502 pulls around 0.43 MIPS, while a 4.77MHz 8088 pulls about 0.35 MIPS.) And hipsters didn't arrive on the scene until the Second Coming of Steve Jobs.

No, Mod GP -1 inaccurate. Lotus 1-2-3 never ran on the Apple ][ family. It ran on the PC from the get-go. It was launched in 1983, not 1982.

Lotus bought VisiCalc in 1985, not 1986.

Excel didn't come out until 1985 (not 1984), and it was never ported to DOS. Its first appearance on a PC was as a Windows version in 1987. It came with a run-time version of Windows if you didn't already have Windows. Excel managed to kill Lotus 1-2-3 primarily because it was born as a GUI app and was native GUI all the way through. Lotus 1-2-3 stumbled on its way to the GUI, which allowed Excel to eventually overtake it.

Lotus 1-2-3 was born on PCs. It never ran on an Apple ][ family machine. VisiCalc, MultiPlan and AppleWorks were on the Apple ][. (Well, technically AppleWorks only ran on the //e and later with an extended 80-column card. It required 128K or more to run.)

Well, just because the missed advertising didn't affect sales yet, there's no guarantee that would continue to hold. Eventually someone would fill the advertising vacuum and they'd have a turf war on their hands.

Wow, that's some comedy gold right there!

We do something similar with our tool releases at work. The release notes indicate bugs that were filed on a previous release and closed with the current release, and if there are open issues what the open issues are. (Usually, it's something very obscure, otherwise it would be fixed.) We do something similar with chip errata. The errata document states which chip revisions are affected, and thus implicitly what chip revision fixes the issue.

Thus, we actually have a two tiered approach. There's the internal system(s) that tracks bugs against the actual development. So, if a bug shows up in a development version, developers and internal testers can file bugs on each other. All that noise has absolutely no business going outside the development team, as it's really just developer-to-developer communication. Then there's the customer issue tracking system. Customer-reported bugs get filed in that system, and they get their own ticket number, and it gets tied to a bug filed in the internal system. The customer bug reports are the ones we comment on in the release notes, along with any notable bugs we discovered in internal testing that customers may not have hit yet.

Disclaimer: My description above is a loose description of the processes we employ at work, and there is variation across teams and business units. It isn't intended to be rigorous. I'm only commenting on my team and teams I've worked closely with. The principle is the same, though. Our dirty laundry (the internal bug tracking system) stays internal. Externally reported bugs get tracked somewhat more opaquely, simply connecting the bug report to the release it's fixed in. It seems quite reasonable to me.