IIRC, Photogenics was originally a piece of commercial software for the Amiga. And rather good it was too. So good in fact, I started writing a free alternative. And therein lies our problem...
IIRC, this is how the later PC emulators worked on the Amiga.
However, it was actually slower for some things than interpreters, mainly because of program dynamically modifying their own code (Doom being a case in point)
If you're running one of those "mem cleaners" of course it will take ages to load up - every time netscape loads some data the mem program forces it to be pages to disk.
They're designed to be run those infrequently and *before* you load your program.
Well - I've never, ever, seen a BSOD on Windows 2000 and I've never had it lock up either (really! - unlike Windows Me - grrr...)
I have had the system spontaneously reboot at least twice (I think that's a hardware problem personally).
It was a minor project. Probably the IPv6 stack. Now Microsoft can claim other implementations are based on their code. Then they truly will 0wn 7h3 1Nt3rn37.
However, Microsoft will not ship a commercial version of Windows 2000 with built-in IPv6 for another two years, admits Tony Hain, program manager of IPv6 for Microsoft's Windows Networking group. Hain says Microsoft is focused on getting application developers to support IPv6 first, so the technology will be useful to consumers and businesses when it ships.
Yet another example of the chicken and the egg situation. Why can't Microsoft release IPv6 stacks for both Windows 200 and Windows Me now? I'm sure businesses would then find ways to make the technology useful. I mean it's not as if there's a lack of useless features in Windows 200 as it is...
Let's say I buy one of these games on a bootable CD. The CD includes drivers for A,B,C graphics cards. Now a couple of years down the line I install D graphics card. It comes with a Linux driver on CD-ROM which I install (Linux, of course, being supported by all vendors by then:). Now I feel like playing my game - so I put in the CD drive and reset. I wait, I wait some more. Up comes the message, "No supported graphics card found". The whole point of having drivers is modular upgradability. If we burn the drivers to a CD-ROM, how can they be upgraded?
IIRC, Amstrad allowed the ZX Spectrum ROM to be freely used in emulators. Following the lead, many companies holding the copyright to old Spectrum games allowed them to be copied to be used with emulators. I remember the C64's ROMs has a similar status until one company decided they were going to start selling C64s again in the Far East.
...if every Tom, Dick and Harry didn't want to have their own domain for their web pages.
In the good old days only the major hosts had their own domains and everyone else just had a home page off it. (eg. http://www.ic.ac.uk/~abc99)
Now companies like www.freenetname.co.uk give domains away for free (as in with catches) and everyone wants to jump on the bandwagon.
I think that naming domains along the same lines as newsgroups is good. Yes - I know it's not compatible with the current DNS system - but you could easily implement an alternative mechanism.
So - if MontaVista's kernel is fully preemptive it should be interruptible at any time. This means that there are no times when you can't serve an interrupt, or am I wrong? Therefore, doesn't this throw up problems for atomicity? What if I read the value of A into a register, then an interrupt occurs which reads the value of A, modifies some hardware settings, then changes the value of A, the interrupt returns, then I look at the value I have in the register and on that basis, I modify the hardware settings again? This is the classic problem, where atoms are required, and one standard solution is to use atomic CPU instructions which read, test and modify A in one instruction. But my question is, how are these kind of atoms implemented on platforms which don't support these kind of instructions? (Amiga, for eg., off the top of my head)
Slashdot Mirror
IIRC, Photogenics was originally a piece of commercial software for the Amiga. And rather good it was too. So good in fact, I started writing a free alternative. And therein lies our problem...
IIRC, this is how the later PC emulators worked on the Amiga. However, it was actually slower for some things than interpreters, mainly because of program dynamically modifying their own code (Doom being a case in point)
Real phyiscs = real macro scale physics. I remember this package for Lightwave that did physics for clothes and it was awful (and cost >$1500...)
Yeah - I remember this is an age old problem. But even if it was possibly to do realistic simulations, they'd be too slow for games.
If you're running one of those "mem cleaners" of course it will take ages to load up - every time netscape loads some data the mem program forces it to be pages to disk. They're designed to be run those infrequently and *before* you load your program.
Well - I've never, ever, seen a BSOD on Windows 2000 and I've never had it lock up either (really! - unlike Windows Me - grrr...) I have had the system spontaneously reboot at least twice (I think that's a hardware problem personally).
...I thought it was a hoax.
It was a minor project. Probably the IPv6 stack. Now Microsoft can claim other implementations are based on their code. Then they truly will 0wn 7h3 1Nt3rn37.
Let's say I buy one of these games on a bootable CD. The CD includes drivers for A,B,C graphics cards. Now a couple of years down the line I install D graphics card. It comes with a Linux driver on CD-ROM which I install (Linux, of course, being supported by all vendors by then :). Now I feel like playing my game - so I put in the CD drive and reset. I wait, I wait some more. Up comes the message, "No supported graphics card found". The whole point of having drivers is modular upgradability. If we burn the drivers to a CD-ROM, how can they be upgraded?
IIRC, Amstrad allowed the ZX Spectrum ROM to be freely used in emulators. Following the lead, many companies holding the copyright to old Spectrum games allowed them to be copied to be used with emulators. I remember the C64's ROMs has a similar status until one company decided they were going to start selling C64s again in the Far East.
...if every Tom, Dick and Harry didn't want to have their own domain for their web pages. In the good old days only the major hosts had their own domains and everyone else just had a home page off it. (eg. http://www.ic.ac.uk/~abc99) Now companies like www.freenetname.co.uk give domains away for free (as in with catches) and everyone wants to jump on the bandwagon. I think that naming domains along the same lines as newsgroups is good. Yes - I know it's not compatible with the current DNS system - but you could easily implement an alternative mechanism.
So - if MontaVista's kernel is fully preemptive it should be interruptible at any time. This means that there are no times when you can't serve an interrupt, or am I wrong? Therefore, doesn't this throw up problems for atomicity? What if I read the value of A into a register, then an interrupt occurs which reads the value of A, modifies some hardware settings, then changes the value of A, the interrupt returns, then I look at the value I have in the register and on that basis, I modify the hardware settings again? This is the classic problem, where atoms are required, and one standard solution is to use atomic CPU instructions which read, test and modify A in one instruction. But my question is, how are these kind of atoms implemented on platforms which don't support these kind of instructions? (Amiga, for eg., off the top of my head)