Accurate economic modeling needs infinite resources, as the existence of the economic modeler needs to be taken into account, and it could be argued that the entire universe would have to be modelled 100% accurately - one atom being in a different place could cause drastically different outcomes years down the line, causing different economic conditions.
There's also a difference between bad engineering and bad assembly.
The US automakers have learned this one the hard way - some of their cars have amazing engineering. But, all that engineering was let down by poor assembly quality (of both the component parts and of the car itself.)
I've actually dealt with a D-Link USB WiFi adapter that the USB connector wasn't soldered to the board.
It's a wonder the thing even worked at first without giving the user a problem. (Five minutes later, after the user complained, it was working fine... but it didn't work for long.)
It popped up an error dialog, but it crashed it right away... I'd suspect it's a conflict between some driver installed for Virtual PC (I had the DOS VM Additions installed) and Windows 3.1. Either that, or some instruction that Windows expected to exist on an i386 that isn't properly handled by the Core 2 Duo, or maybe even a race condition revealed by the fact that I threw a 1.6GHz Core 2 Duo at something that the designers never expected anything faster than a 75MHz Pentium.
Why study MS-DOS, when it's a reverse-engineered CP/M-alike?
Novell OpenDOS is where it's at - it's already open source, and while it does have reverse-engineered MS-DOS bits, it traces its lineage directly all the way back to CP/M.
Anyway, there's something much newer than what you used, for the taskbar... Calmira. And, there was a fork updated in March of this year...
(And, speaking of NT 3.51... Microsoft actually wrote their own Explorer shell for NT 3.51, called NewShell, although it wasn't meant for general release. Installing it on 3.51 actually causes it to report as NT 4.0, as the original NT 4.0 alphas were merely NT 3.51 with NewShell installed.)
You could change a win.ini setting (or possibly a system.ini setting - not a registry setting, as the registry in Win3.1 was only for file associations) to disable the BSOD, but Ctrl-Alt-Del ALWAYS called the BSOD on Windows 3.1 by default.
However, Windows 3.1 always passed a second Ctrl-Alt-Del to the BIOS.
Win3.1 is also a hell of a lot more lightweight than 2k. Think about it, you're setting up an embedded kiosk with a Geode LX (slower than a Pentium II) and 32MB RAM. Are you gonna run Win2k (which stretches the limits of that thing) or 3.1?
Anyway, XP Embedded is probably what Microsoft's intending as a replacement...
"The system is either busy or has become unstable" is a Win9x error, no?
Windows 3.1's error of choice was the General Protection Fault... but it was better than the Unrecoverable Application Error of Windows 3.0...
But, the bigger problem in Win3.x is running out of "resources," especially in Windows 3.0... while newer versions of Windows were still susceptible (I've actually successfully run a Win2k machine out of resources - it took a very, very buggy beta version of Opera that had serious resource leaks, and 16384 GDI handles to be used up (contrary to popular belief, there AREN'T unlimited GDI handles left over, as opposed to 200 for 3.x, but I did it,) 3.x's multitasking capability was hindered not by the poor cooperative multitasking model, but by the limited resources.
Last I checked, almost all diesels (with some VERY rare exceptions) were injected.;)
It's direct injection that's the problem (where fuel is... directly injected into the cylinder.) Even "indirect" injection is more direct than most gasoline fuel injection - gasoline engines inject it into the intake air, whereas IDI diesels inject into a prechamber in the head, where it ignites and then mixes with the main intake charge in the cylinder.
And, turbochargers just increase the pressure of the intake air above ambient pressure, allowing for more fuel to mix in.
I might be buying a (VW) turbocharged indirect injection diesel in the next couple weeks or so, and it's just as good on biofuels as a non-turbocharged engine.
Oh, and Mercedes sold turbo IDI diesels here in the US all the way up until 1999. VW last sold them in 1992 (the one I'm looking at getting is a Canadian-spec 1993,) and started selling direct injection engines here in 1996. But, the 1996-2003 VW TDIs certainly aren't bad for biodiesel. (Just don't run veggie oil in those, and it's still a bad idea in any engine not specifically designed for it, even indirect injection engines.)
Doesn't Vista finally move the graphics into userland?
And, the funny thing is, Microsoft's been screwing this up for years now, starting with OS/2 1.1. It was IBM's turn to work with OS/2 1.3, and they quickly moved it out of the kernel. But, Windows NT was a fork of OS/2 1.2, not 1.3.
But, Core 2 (or, rather, the Core microarchitecture that Core 2 is based on) is as big of a leap over the ancient P6 (from 1995) that the original Core Duo and Core Solo were based on. (Core Duo essentially being two Dothan Pentium Ms sharing a cache, with better SSE support, and a die shrink, and Core Solo being the single-core version.)
Technically, wouldn't that make you a heavily multi-homed tier 2 with a crapload of bandwidth, of sorts, not a "tier 0" if I'm understanding everything right?
And that specifically is why it's plausible that it's Opera Mini, not Mobile.
Opera Mini has no support for client-side JavaScript - to do anything using JS, Opera Mini has to make a round trip to the Opera servers in Norway, and the JavaScript code runs on those servers.
On Palm OS, there's basically two legal choices for browsers (and one of them is only semi-legal.) Blazer, which sucks ass, but is a native Palm OS app, and can handle JavaScript without round trips, and Opera Mini, which requires the rather unstable IBM J9 JVM (which isn't legally available in any form, for any price, any more - I had to pirate a copy of it for my Centro,) has to make a round trip for every single JavaScript action, has trouble with loading images on Palm OS, but is fast when it's working, and can work with more pages.
So, I get to choose my browser based on which one sucks less... god, I can't wait for Palm to release Nova... hopefully they release a ROM update for the Centro so I can get away from this frankengarnet shit that nobody wants to develop for... oh, wait, this is Palm, they'll never do that.
But in Kazaa or eMule, the third party clients have options to say that you've uploaded the maximum without uploading anything at all, and your setting is trusted.
This is why BitTorrent downloads always start out so slow, because you HAVE to download the first chunk before you can share it, and nobody will provide you with much bandwidth until you start sharing.
Slashdot Mirror
Accurate economic modeling needs infinite resources, as the existence of the economic modeler needs to be taken into account, and it could be argued that the entire universe would have to be modelled 100% accurately - one atom being in a different place could cause drastically different outcomes years down the line, causing different economic conditions.
There's also a difference between bad engineering and bad assembly.
The US automakers have learned this one the hard way - some of their cars have amazing engineering. But, all that engineering was let down by poor assembly quality (of both the component parts and of the car itself.)
I've actually dealt with a D-Link USB WiFi adapter that the USB connector wasn't soldered to the board.
It's a wonder the thing even worked at first without giving the user a problem. (Five minutes later, after the user complained, it was working fine... but it didn't work for long.)
It popped up an error dialog, but it crashed it right away... I'd suspect it's a conflict between some driver installed for Virtual PC (I had the DOS VM Additions installed) and Windows 3.1. Either that, or some instruction that Windows expected to exist on an i386 that isn't properly handled by the Core 2 Duo, or maybe even a race condition revealed by the fact that I threw a 1.6GHz Core 2 Duo at something that the designers never expected anything faster than a 75MHz Pentium.
Well, I just tried it in a VM, and failed miserably - but Win32s was crashing outright even on FreeCell, which is the included app.
I *THINK* SeaMonkey will run under Win32s... great, now I need to source my DOS of choice, WfWG, Win32s, and SeaMonkey, don't I? :P
Why study MS-DOS, when it's a reverse-engineered CP/M-alike?
Novell OpenDOS is where it's at - it's already open source, and while it does have reverse-engineered MS-DOS bits, it traces its lineage directly all the way back to CP/M.
And I'm pretty sure Cardfile.exe would run in modern versions of Windows, up to Vista/x86.
I believe Windows NT 3.51 was 21 floppies...
Anyway, there's something much newer than what you used, for the taskbar... Calmira. And, there was a fork updated in March of this year...
(And, speaking of NT 3.51... Microsoft actually wrote their own Explorer shell for NT 3.51, called NewShell, although it wasn't meant for general release. Installing it on 3.51 actually causes it to report as NT 4.0, as the original NT 4.0 alphas were merely NT 3.51 with NewShell installed.)
You could change a win.ini setting (or possibly a system.ini setting - not a registry setting, as the registry in Win3.1 was only for file associations) to disable the BSOD, but Ctrl-Alt-Del ALWAYS called the BSOD on Windows 3.1 by default.
However, Windows 3.1 always passed a second Ctrl-Alt-Del to the BIOS.
Win3.1 is also a hell of a lot more lightweight than 2k. Think about it, you're setting up an embedded kiosk with a Geode LX (slower than a Pentium II) and 32MB RAM. Are you gonna run Win2k (which stretches the limits of that thing) or 3.1?
Anyway, XP Embedded is probably what Microsoft's intending as a replacement...
IE5 runs on Windows 3.1. ;)
"The system is either busy or has become unstable" is a Win9x error, no?
Windows 3.1's error of choice was the General Protection Fault... but it was better than the Unrecoverable Application Error of Windows 3.0...
But, the bigger problem in Win3.x is running out of "resources," especially in Windows 3.0... while newer versions of Windows were still susceptible (I've actually successfully run a Win2k machine out of resources - it took a very, very buggy beta version of Opera that had serious resource leaks, and 16384 GDI handles to be used up (contrary to popular belief, there AREN'T unlimited GDI handles left over, as opposed to 200 for 3.x, but I did it,) 3.x's multitasking capability was hindered not by the poor cooperative multitasking model, but by the limited resources.
Well, they're perfectly fine when single-tasking, and being used as a GUI framework for a kiosk app. And, because they take so little resources...
Last I checked, almost all diesels (with some VERY rare exceptions) were injected. ;)
It's direct injection that's the problem (where fuel is... directly injected into the cylinder.) Even "indirect" injection is more direct than most gasoline fuel injection - gasoline engines inject it into the intake air, whereas IDI diesels inject into a prechamber in the head, where it ignites and then mixes with the main intake charge in the cylinder.
And, turbochargers just increase the pressure of the intake air above ambient pressure, allowing for more fuel to mix in.
I might be buying a (VW) turbocharged indirect injection diesel in the next couple weeks or so, and it's just as good on biofuels as a non-turbocharged engine.
Oh, and Mercedes sold turbo IDI diesels here in the US all the way up until 1999. VW last sold them in 1992 (the one I'm looking at getting is a Canadian-spec 1993,) and started selling direct injection engines here in 1996. But, the 1996-2003 VW TDIs certainly aren't bad for biodiesel. (Just don't run veggie oil in those, and it's still a bad idea in any engine not specifically designed for it, even indirect injection engines.)
But the basic OS architecture is lifted straight from OS/2.
And, technically, all that's left of OS/2 in NT is just that, the basic OS architecture - pinball.vxd was gone after NT 3.51, right?
Doesn't Vista finally move the graphics into userland?
And, the funny thing is, Microsoft's been screwing this up for years now, starting with OS/2 1.1. It was IBM's turn to work with OS/2 1.3, and they quickly moved it out of the kernel. But, Windows NT was a fork of OS/2 1.2, not 1.3.
Er, I screwed up. The Core microarchitecture was as big of a leap over P6 (or bigger) as Nehalem is over Core.
And, there's also a Solo moniker.
But, Core 2 (or, rather, the Core microarchitecture that Core 2 is based on) is as big of a leap over the ancient P6 (from 1995) that the original Core Duo and Core Solo were based on. (Core Duo essentially being two Dothan Pentium Ms sharing a cache, with better SSE support, and a die shrink, and Core Solo being the single-core version.)
Technically, wouldn't that make you a heavily multi-homed tier 2 with a crapload of bandwidth, of sorts, not a "tier 0" if I'm understanding everything right?
Bah, my bad, eight cores per chip, eight-way SMT per core.
Still, not 128 cores per box.
That's 128 threads (and they have a four socket, eight core version, for 256 threads,) not 128 cores.
Basically, their chips are dual-core, and each core is split 32 ways by SMT (what in the Intel world is known as HyperThreading.)
What if you wrote the compiler yourself, in assembly?
Then, the exploits would only be at the BIOS or hardware level...
And that specifically is why it's plausible that it's Opera Mini, not Mobile.
Opera Mini has no support for client-side JavaScript - to do anything using JS, Opera Mini has to make a round trip to the Opera servers in Norway, and the JavaScript code runs on those servers.
On Palm OS, there's basically two legal choices for browsers (and one of them is only semi-legal.) Blazer, which sucks ass, but is a native Palm OS app, and can handle JavaScript without round trips, and Opera Mini, which requires the rather unstable IBM J9 JVM (which isn't legally available in any form, for any price, any more - I had to pirate a copy of it for my Centro,) has to make a round trip for every single JavaScript action, has trouble with loading images on Palm OS, but is fast when it's working, and can work with more pages.
So, I get to choose my browser based on which one sucks less... god, I can't wait for Palm to release Nova... hopefully they release a ROM update for the Centro so I can get away from this frankengarnet shit that nobody wants to develop for... oh, wait, this is Palm, they'll never do that.
But in Kazaa or eMule, the third party clients have options to say that you've uploaded the maximum without uploading anything at all, and your setting is trusted.
This is why BitTorrent downloads always start out so slow, because you HAVE to download the first chunk before you can share it, and nobody will provide you with much bandwidth until you start sharing.