I liked DOS on my old machines. You could do amazing things with it, and it would just keep going. Program to snoop passwords on old Netware systems? No problem. Hook up int09, wait for someone to enter 'login' and record the next 30 keystrokes. Want to make a cooperative multitasking system out of it? Took less than two weeks of coding, and basically just involved reprogramming timer frequencies and wrapping int13 and int21 to provide primitive reentrancy. Oh, memory lane is a good place to visit:-)
Win3.1 was fun to play with, but died on me way to often for my liking. Win95 was better, but started to get in the way too much...
Don't get me wrong - I like my Linux box. And my new W2K box at work. I can do fun stuff with them too. I just don't get the same great feeling of control with them, since the OS will NOT move out of the way. Hmm - maybe I should become a kernel hacker instead:-)
In a real competitive market with low costs of entry, other firms will see Microsoft with such high profits and have incentive to enter the market, undercutting Microsoft.
Uhm, what low cost of entry are you talking about here? The tens of thousands of man-years required to make a product equivalent to (or hopefully better than) Windows XP's user interface, and it's huge install base? Not to mention, that were you able to pull of such a remarkable feat, the next two versions of XP would roll by in the meantime, making your shiny new product seem - well - not so shiny and new anyway.
Look at the enormous amounts of developer time that went into making Linux. Enormous. HUGE. And we're still actively debating why it is not really a force to reckon with on the desktop of the average user. How would you imagine some company coming up with even more manyears of developer time than that, and be able to pay for it?
A few years ago, a good example of a uClinux implementation was the uCsimm, a 30-pin SIMM sized machine based on the Motorola 68EZ328. 8Mbytes RAM, 2 Mbytes flash, Crystal 8900 (10 Mbit) Ethernet. The 68EZ328 powers all pre-PalmOS-5.0 units. We had a web server with complete CGI capability, as well as several additional communication front-end tools. So I know Linux runs on the 68EZ328, and I've seen references to the Palm H/W in the uClinux kernel code, though I haven't tried it on my Palm...
I'm the (mostly) happy owner of a uCSimm module, and the damn thing works just fine. The Ethernet was really slow on the kernel 2.0.38 based uCLinux that came with it (no DMA - probably a HW issue), but other than that a fine piece of equipment. The whole things runs out of about 700 kb of Flash, and the RAM footprint is roughly the same size - about half of that looks to be buffers for various Linux purposes.
The biggest problem was (and is) that the Dragonball (the nickname for the 68EZ328) has no concept of process isolation (no MMU on these babies), but that's exactly where uCLinux reigns supreme. Kind of makes the Linux experience more DOS-like, in that some stray pointer can damage OS internals. No way of avoiding that on the given HW, I know - but still a considerable headache.
Can someone please enlighten me as to the footprint and stability of the 2.4.x line of uCLinux kernels?
Have you ever used a proper Amiga, or just an unexpanded old A500 or A1200?
That's just it - an unexpanded old A500 *IS* a proper Amiga to the vast majority of the people old enough to have tried or owned one. I've seen more Amigas than most of the geeks I know, but to me a proper Amiga is an old unexpanded (OK, maybe the 1 Meg expansion) A500. They totally rocked, and that's probably because they were very useful without all the fancy schmancy stuff that caused later Amigas to become more like an incompatible PC than the cool original piece of hardware it originally was.
I'm not agreeing with the original poster or anything - I'd just like to point out that your cool A4000 and its likes are mere droplets in the ocean of old A500's.
Back in the days when 286 processors were all most people had in their PC (including mine), one guy at school got a 386SX based system running at a whopping 16 MHz. Now this was all good and fine, but for the fact that he was constantly nagging us about our systems being slow, outdated, ridiculous and whatnot. So we made a little TSR (Terminate Stay Resident) program that did nothing but count to 20000 ~18 times per second (the frequency of the DOS time clock). When it loaded, it wrote "CPU instability detected - switching to XT mode" and of course made the machine run sooooo slowly (XTs are the slowest PC-compatibles ever made, in case you're too young to know). It took him several days to figure it out, and it was REALLY difficult not to laugh our a**es off in the meantime.
sounds like something isn't proportioned just right yet. Cool gadget, but I agree with the article that this is bound to be a niche product. Not enough power for serious data usage, too high price to be a toy.
Lego is NEVER useless
on
Lego Addictions
·
· Score: 4, Insightful
Come on, michael - there is no such thing as useless Legos. They're fun, they pique the imagination, they provide an endless number of fun hours. I'm really glad my son is turning 5 soon, so I get to play with real Legos with him (IMHO, Duplos are not really funny for adults). If the cool Start Destroyer weren't so expensive, I'd get it for him (and me:-) right away.
I have so many fond memories of playing with Legos during my childhood, and no model was ever useless. They could all be combined in so many fun and surprising ways that even the hopelessly outdated models I inherited from my father were never useless. In fact, most of the Legos used by my son now is from MY childhood, and a (very small) part of those are from my fathers and uncles' childhood.
In some ways, this goes to show that a carefully designed interfacing system can survive many years and revisions. If only programming API's had the same level of longevity. The C standard libraries, perhaps?
and it is considered good design factor, but it is also of questionable legality, at least in some major parts of the world (the EU, for instance). I still much prefer it, but attempting to make people do stuff that has questionable legality is... not a good idea.
PS. I know that both the links in my posting are deep links - go figure:-)
I'm quite familiar with the VW Lupo 3L - it's very popular in my country (Denmark). Several of my previous co-workers bought them, and if they drive them carefully, they do get 30+ km/l out of them. Impressive.
I think a motorcycle will have difficulty getting the same mileage as the VW concept car. The reason is aerodynamics - a closed entity like the VW is MUCH easier to control in terms of resistance to wind than is the case for a motorcycle, where the rider is part of the wind resistance. But I agree that dropping the electric engine would seem prudent.
That's nice - 180 MPG is roughly equivalent to 80 km/l to the imperial-challenged crowd. But check out the 100 km/l car done by Volkswagen(yes, that's 100 km to the litre, or 235 MPG if you don't like SI units). I'm not affiliated with Volkswagen or anything - it's just cool technology.
Actually, we (the Danes) do place a lot of new windmills off-shore. At about 10km or so distance from the coast, the curvature of the Earth will have hidden them from sight. Or maybe the ever-present mist at sea? Hm, probably a combination of the two...
Avoiding visual pollution (and other pollutions, real or perceived) is a major factor in that decision. The majority of our West Coast is well-suited for windpower, but it is also a *massive* tourist attraction. To cater for both interests, we just put them out at sea to keep the coastline in its natural state, and the windmills produce electricity just the same. Good, but rather costly solution, IMHO.
As I understand it, sleep activates the brains method of "garbage collecting". In other words, while you sleep your thoughts get better organized and your mind detaches unused information. If you're willing to forego THAT (read: become a raving lunatic), then go right ahead and do without sleep. Just let the rest of the world know when you hit the one-week barrier again, so we can stay out of your reach.
But it would depend on your definition of "nap". Does it last 1 hour? 2 maybe? I seem to recall the garbage collection stuff being related to REM sleep, which starts about 1 hour into your sleep (figure pulled entirely from memory). So naps of 2 hours or so should definately help you keep your mind, but in the long run you'll probably just have to come to turns with the fact that we're DESIGNED (creationist?/Darwinist? - still true) to need sleep.
Great idea - all you need is campaign funding! Off the record, I know these two organizations that DESPERATELY needs a(nother) congressman. You could probably persuade them to give huge donations, in return for helping to make laws more... managable... to them. They're called the RIAA and the MPAA, and you can contact them by cal....
The 14 platforms would be: ZX81, BBC B (huh?), Windows 3.1, Windows for Workgroups, Windows 95, Windows 98, Windows ME, Windows NT 3.51, Windows NT 4.0 Workstation, Windows NT 4.0 Server, Windows 2000 Professional, Windows 2000 Server, Windows XP Home Edition and Windows XP Pro Edition. And that's not even counting service packs!
You make lots of good points. Really, you do. But I'm not talking about marihuana (is that an 'h' or an 'j' in there?) being addictive - I don't think it really is. It has been shown in my part of the world that the vast majority of hard-drug users started on marihuana. This is a mild "drug" with few side-effects above those of alcohol, so lots of people (wrongly) assume this goes for other drugs too. Then they try hard drugs, like the trip, and (sadly) develops physical addiction in too many cases.
Apparently, this connection from marihuana to hard drugs does NOT have a parallel "alcohol to hard drugs connection" or a "tobacco to hard drugs connection" for that matter. That's what leads me to conclude that legalizing marihuana might be more trouble than it's worth.
I don't think drug abuse should be considered a crime in itself, but most drug addict can not afford the drugs, and therefore must resort to crime to pay another criminal to get it. That's the reason why I think that it's better to give addicts drugs for free - under control, and with treatment as the endgoal. By giving it out through the authorities you eliminate a sizeable chunk of organized crime AND get regular contact with the addicts, and this contact is very valuable in terms of starting treatment. You also reduce the number of overdose deaths and deaths from impure drugs, not to mention certain transmittable diseases (AIDS, anyone?). Of course there's an inherent risk of faster escalation when hard drugs are "free", but it's well worth taking, IMHO.
Judging from your description of the US government's propaganda, I'm beginning to like not living in the US...
'No ma, government propraganda doesn't work on me!'
I'm pretty sure it does, but not whatever propaganda you have in mind - I'm not an American citizen, and I have never even been to the US. My local government is looking into legalizing mild drugs like marihuana (stupid idea - leads to more addicts), and we're contemplating giving hard-drug addicts free drugs under control to keep them from becoming criminals too (great idea - practically destroys organized crime). My view on drugs being different from alcohol and tobacco is - mostly - my own, and usually different from mainstream thinking.
Why would sharing my governments view in certain areas be bad, BTW? There are plenty of areas where I do NOT share their views (even though I did vote for them). It's like that with most politicians - you agree with some ideas, and disagree with others. The ones you agree with the most usually gets your vote.
Your post leaves me with the impression that the "anti-government" propaganda I hear you've got so much of in the US, have really worked on *YOU*.
I'm all for technology, but this looks like an attempt to make the "wonderpill" to cure all human ailments (sp?). I think we should devote some of that energy into preventing some of them, instead of demanding an instant cure for the problems we inflict on our own bodies. You know: stress, alcohol, drugs, tobacco etc. I'm not trying to push the view that you should abstain from all this (well, drugs you should avoid though), but realize that there are no magic cures for the problems these things cause to your body and mind.
Disregarding the fact that you're wrong, the idea is still interesting. In todays world, however, negative mass would somehow find its way into fastfood so that the dangerously obese would weigh a scant 10 kg or so. Hell, some of them would take off after a particularly large meal - new reason to sue McDonalds, perhaps?
"I swear, your Honor, I just had 6 BigMacs with fries and when I left the joint, I took off and was subjected to a traumatic view of the inner city - I demand compensation!".
I was under the impression that NTFS filesystems were readonly under Linux. This may be an outdated view of things, in which case you can just disregard the rest of this post, and *PLEASE* point me to wherever I can download full NTFS support for Linux.
You are absolutely correct that I trade the robustness of NTFS for the quite primitive FAT by doing so. I knew that too. I fully understand that this will increase the chances of me having to take action, and it does mean that in actual fact. However, the tradeoff is - in MY opinion - justified since I can not easily repair an NTFS partition. Extract stuff, perhaps (see the first paragraph) - but not repair. *I* need repairing far more than I need extraction on my boot drive, since it doesn't (by definition) contain my actual work. I wouldn't trust a FAT system to hold my own work - NTFS for that job. But I much prefer frequent easy repairs to my bootdrive, rather than difficult (or almost impossible) repairs infrequently. Whatever can get me up and running again without having to reinstall everything is a Good Thing. God, how I HATE the Windows registry.
The suggestion by platypus in reply to my post looks promising, since it will probably allow me to use NTFS (which I much prefer) *and* allow easy recovery. If I can get some more diskspace, I'll probably do it like platypus described.
It would indeed - except the problem is that I don't want the exact same configuration 6 months later. New versions of half the applications, new much-needed Service Packs from M$ and so on. That said, I usually have IT service from my company put an image on the machine when needed, but I find that I use as much time upgrading/installing my apps as I would installing it all in the first place. But your idea is interesting if I make regular images of my *own* W2K installation though - maybe I should give it a try...
Besides, having Linux on it is ALWAYS desirable:-)
First a (minor) correction - W2K allows you to use NTFS. It's not mandatory, you know.
Most of the W2K installations *I* consider to be wise have a small boot partition for W2K (~4 Gigs - W2K and Windows apps are bootdisk space hogs [sigh]) which uses FAT, just so that any disk-analyzer can find out what's wrong with it THIS time. Then put all data and programs on a secondary NTFS partition, which can be accessed when you've either
rescued the FAT boot partition
Re-installed W2K
The last option is rarely needed for an average user (they do it anyway, though), but for a developer (like myself) it's necessary with intervals of ~6 months - sigh (but that's due to DLL bloat, most of the time). If the NTFS one fails (rather unlikely barring physical disk damage) you can repair it using any of the tools already available for that job. I never ever had to repair anything running on NTFS, though. Solid as rock.
This approach has saved my a** more than a few times...
Slashdot Mirror
I liked DOS on my old machines. You could do amazing things with it, and it would just keep going. Program to snoop passwords on old Netware systems? No problem. Hook up int09, wait for someone to enter 'login' and record the next 30 keystrokes. Want to make a cooperative multitasking system out of it? Took less than two weeks of coding, and basically just involved reprogramming timer frequencies and wrapping int13 and int21 to provide primitive reentrancy. Oh, memory lane is a good place to visit :-)
Win3.1 was fun to play with, but died on me way to often for my liking. Win95 was better, but started to get in the way too much...
Don't get me wrong - I like my Linux box. And my new W2K box at work. I can do fun stuff with them too. I just don't get the same great feeling of control with them, since the OS will NOT move out of the way. Hmm - maybe I should become a kernel hacker instead :-)
In a real competitive market with low costs of entry, other firms will see Microsoft with such high profits and have incentive to enter the market, undercutting Microsoft.
Uhm, what low cost of entry are you talking about here? The tens of thousands of man-years required to make a product equivalent to (or hopefully better than) Windows XP's user interface, and it's huge install base? Not to mention, that were you able to pull of such a remarkable feat, the next two versions of XP would roll by in the meantime, making your shiny new product seem - well - not so shiny and new anyway.
Look at the enormous amounts of developer time that went into making Linux. Enormous. HUGE. And we're still actively debating why it is not really a force to reckon with on the desktop of the average user. How would you imagine some company coming up with even more manyears of developer time than that, and be able to pay for it?
A few years ago, a good example of a uClinux implementation was the uCsimm, a 30-pin SIMM sized machine based on the Motorola 68EZ328. 8Mbytes RAM, 2 Mbytes flash, Crystal 8900 (10 Mbit) Ethernet. The 68EZ328 powers all pre-PalmOS-5.0 units. We had a web server with complete CGI capability, as well as several additional communication front-end tools. So I know Linux runs on the 68EZ328, and I've seen references to the Palm H/W in the uClinux kernel code, though I haven't tried it on my Palm...
I'm the (mostly) happy owner of a uCSimm module, and the damn thing works just fine. The Ethernet was really slow on the kernel 2.0.38 based uCLinux that came with it (no DMA - probably a HW issue), but other than that a fine piece of equipment. The whole things runs out of about 700 kb of Flash, and the RAM footprint is roughly the same size - about half of that looks to be buffers for various Linux purposes.
The biggest problem was (and is) that the Dragonball (the nickname for the 68EZ328) has no concept of process isolation (no MMU on these babies), but that's exactly where uCLinux reigns supreme. Kind of makes the Linux experience more DOS-like, in that some stray pointer can damage OS internals. No way of avoiding that on the given HW, I know - but still a considerable headache.
Can someone please enlighten me as to the footprint and stability of the 2.4.x line of uCLinux kernels?
Have you ever used a proper Amiga, or just an unexpanded old A500 or A1200?
That's just it - an unexpanded old A500 *IS* a proper Amiga to the vast majority of the people old enough to have tried or owned one. I've seen more Amigas than most of the geeks I know, but to me a proper Amiga is an old unexpanded (OK, maybe the 1 Meg expansion) A500. They totally rocked, and that's probably because they were very useful without all the fancy schmancy stuff that caused later Amigas to become more like an incompatible PC than the cool original piece of hardware it originally was.
I'm not agreeing with the original poster or anything - I'd just like to point out that your cool A4000 and its likes are mere droplets in the ocean of old A500's.
Well, I think I speak (type?) for all of us here - did you, in fact, get some?
Back in the days when 286 processors were all most people had in their PC (including mine), one guy at school got a 386SX based system running at a whopping 16 MHz. Now this was all good and fine, but for the fact that he was constantly nagging us about our systems being slow, outdated, ridiculous and whatnot. So we made a little TSR (Terminate Stay Resident) program that did nothing but count to 20000 ~18 times per second (the frequency of the DOS time clock). When it loaded, it wrote "CPU instability detected - switching to XT mode" and of course made the machine run sooooo slowly (XTs are the slowest PC-compatibles ever made, in case you're too young to know). It took him several days to figure it out, and it was REALLY difficult not to laugh our a**es off in the meantime.
sounds like something isn't proportioned just right yet. Cool gadget, but I agree with the article that this is bound to be a niche product. Not enough power for serious data usage, too high price to be a toy.
Come on, michael - there is no such thing as useless Legos. They're fun, they pique the imagination, they provide an endless number of fun hours. I'm really glad my son is turning 5 soon, so I get to play with real Legos with him (IMHO, Duplos are not really funny for adults). If the cool Start Destroyer weren't so expensive, I'd get it for him (and me :-) right away.
I have so many fond memories of playing with Legos during my childhood, and no model was ever useless. They could all be combined in so many fun and surprising ways that even the hopelessly outdated models I inherited from my father were never useless. In fact, most of the Legos used by my son now is from MY childhood, and a (very small) part of those are from my fathers and uncles' childhood.
In some ways, this goes to show that a carefully designed interfacing system can survive many years and revisions. If only programming API's had the same level of longevity. The C standard libraries, perhaps?
and it is considered good design factor, but it is also of questionable legality, at least in some major parts of the world (the EU, for instance). I still much prefer it, but attempting to make people do stuff that has questionable legality is ... not a good idea.
PS. I know that both the links in my posting are deep links - go figure :-)
I'm quite familiar with the VW Lupo 3L - it's very popular in my country (Denmark). Several of my previous co-workers bought them, and if they drive them carefully, they do get 30+ km/l out of them. Impressive.
I think a motorcycle will have difficulty getting the same mileage as the VW concept car. The reason is aerodynamics - a closed entity like the VW is MUCH easier to control in terms of resistance to wind than is the case for a motorcycle, where the rider is part of the wind resistance. But I agree that dropping the electric engine would seem prudent.
That's nice - 180 MPG is roughly equivalent to 80 km/l to the imperial-challenged crowd. But check out the 100 km/l car done by Volkswagen(yes, that's 100 km to the litre, or 235 MPG if you don't like SI units). I'm not affiliated with Volkswagen or anything - it's just cool technology.
Actually, we (the Danes) do place a lot of new windmills off-shore. At about 10km or so distance from the coast, the curvature of the Earth will have hidden them from sight. Or maybe the ever-present mist at sea? Hm, probably a combination of the two...
Avoiding visual pollution (and other pollutions, real or perceived) is a major factor in that decision. The majority of our West Coast is well-suited for windpower, but it is also a *massive* tourist attraction. To cater for both interests, we just put them out at sea to keep the coastline in its natural state, and the windmills produce electricity just the same. Good, but rather costly solution, IMHO.
The noise would come from the protesters, protesting the ice daggers and the noise.
As I understand it, sleep activates the brains method of "garbage collecting". In other words, while you sleep your thoughts get better organized and your mind detaches unused information. If you're willing to forego THAT (read: become a raving lunatic), then go right ahead and do without sleep. Just let the rest of the world know when you hit the one-week barrier again, so we can stay out of your reach.
But it would depend on your definition of "nap". Does it last 1 hour? 2 maybe? I seem to recall the garbage collection stuff being related to REM sleep, which starts about 1 hour into your sleep (figure pulled entirely from memory). So naps of 2 hours or so should definately help you keep your mind, but in the long run you'll probably just have to come to turns with the fact that we're DESIGNED (creationist?/Darwinist? - still true) to need sleep.
I don't know - something like this?
Great idea - all you need is campaign funding! Off the record, I know these two organizations that DESPERATELY needs a(nother) congressman. You could probably persuade them to give huge donations, in return for helping to make laws more ... managable ... to them. They're called the RIAA and the MPAA, and you can contact them by cal....
Oh wait...
The 14 platforms would be: ZX81, BBC B (huh?), Windows 3.1, Windows for Workgroups, Windows 95, Windows 98, Windows ME, Windows NT 3.51, Windows NT 4.0 Workstation, Windows NT 4.0 Server, Windows 2000 Professional, Windows 2000 Server, Windows XP Home Edition and Windows XP Pro Edition. And that's not even counting service packs!
You make lots of good points. Really, you do. But I'm not talking about marihuana (is that an 'h' or an 'j' in there?) being addictive - I don't think it really is. It has been shown in my part of the world that the vast majority of hard-drug users started on marihuana. This is a mild "drug" with few side-effects above those of alcohol, so lots of people (wrongly) assume this goes for other drugs too. Then they try hard drugs, like the trip, and (sadly) develops physical addiction in too many cases.
Apparently, this connection from marihuana to hard drugs does NOT have a parallel "alcohol to hard drugs connection" or a "tobacco to hard drugs connection" for that matter. That's what leads me to conclude that legalizing marihuana might be more trouble than it's worth.
I don't think drug abuse should be considered a crime in itself, but most drug addict can not afford the drugs, and therefore must resort to crime to pay another criminal to get it. That's the reason why I think that it's better to give addicts drugs for free - under control, and with treatment as the endgoal. By giving it out through the authorities you eliminate a sizeable chunk of organized crime AND get regular contact with the addicts, and this contact is very valuable in terms of starting treatment. You also reduce the number of overdose deaths and deaths from impure drugs, not to mention certain transmittable diseases (AIDS, anyone?). Of course there's an inherent risk of faster escalation when hard drugs are "free", but it's well worth taking, IMHO.
Judging from your description of the US government's propaganda, I'm beginning to like not living in the US...
'No ma, government propraganda doesn't work on me!'
I'm pretty sure it does, but not whatever propaganda you have in mind - I'm not an American citizen, and I have never even been to the US. My local government is looking into legalizing mild drugs like marihuana (stupid idea - leads to more addicts), and we're contemplating giving hard-drug addicts free drugs under control to keep them from becoming criminals too (great idea - practically destroys organized crime). My view on drugs being different from alcohol and tobacco is - mostly - my own, and usually different from mainstream thinking.
Why would sharing my governments view in certain areas be bad, BTW? There are plenty of areas where I do NOT share their views (even though I did vote for them). It's like that with most politicians - you agree with some ideas, and disagree with others. The ones you agree with the most usually gets your vote.
Your post leaves me with the impression that the "anti-government" propaganda I hear you've got so much of in the US, have really worked on *YOU*.
I'm all for technology, but this looks like an attempt to make the "wonderpill" to cure all human ailments (sp?). I think we should devote some of that energy into preventing some of them, instead of demanding an instant cure for the problems we inflict on our own bodies. You know: stress, alcohol, drugs, tobacco etc. I'm not trying to push the view that you should abstain from all this (well, drugs you should avoid though), but realize that there are no magic cures for the problems these things cause to your body and mind.
Disregarding the fact that you're wrong, the idea is still interesting. In todays world, however, negative mass would somehow find its way into fastfood so that the dangerously obese would weigh a scant 10 kg or so. Hell, some of them would take off after a particularly large meal - new reason to sue McDonalds, perhaps?
"I swear, your Honor, I just had 6 BigMacs with fries and when I left the joint, I took off and was subjected to a traumatic view of the inner city - I demand compensation!".
I was under the impression that NTFS filesystems were readonly under Linux. This may be an outdated view of things, in which case you can just disregard the rest of this post, and *PLEASE* point me to wherever I can download full NTFS support for Linux.
You are absolutely correct that I trade the robustness of NTFS for the quite primitive FAT by doing so. I knew that too. I fully understand that this will increase the chances of me having to take action, and it does mean that in actual fact. However, the tradeoff is - in MY opinion - justified since I can not easily repair an NTFS partition. Extract stuff, perhaps (see the first paragraph) - but not repair. *I* need repairing far more than I need extraction on my boot drive, since it doesn't (by definition) contain my actual work. I wouldn't trust a FAT system to hold my own work - NTFS for that job. But I much prefer frequent easy repairs to my bootdrive, rather than difficult (or almost impossible) repairs infrequently. Whatever can get me up and running again without having to reinstall everything is a Good Thing. God, how I HATE the Windows registry.
The suggestion by platypus in reply to my post looks promising, since it will probably allow me to use NTFS (which I much prefer) *and* allow easy recovery. If I can get some more diskspace, I'll probably do it like platypus described.
It would indeed - except the problem is that I don't want the exact same configuration 6 months later. New versions of half the applications, new much-needed Service Packs from M$ and so on. That said, I usually have IT service from my company put an image on the machine when needed, but I find that I use as much time upgrading/installing my apps as I would installing it all in the first place. But your idea is interesting if I make regular images of my *own* W2K installation though - maybe I should give it a try...
Besides, having Linux on it is ALWAYS desirable :-)
First a (minor) correction - W2K allows you to use NTFS. It's not mandatory, you know.
Most of the W2K installations *I* consider to be wise have a small boot partition for W2K (~4 Gigs - W2K and Windows apps are bootdisk space hogs [sigh]) which uses FAT, just so that any disk-analyzer can find out what's wrong with it THIS time. Then put all data and programs on a secondary NTFS partition, which can be accessed when you've either
- rescued the FAT boot partition
- Re-installed W2K
The last option is rarely needed for an average user (they do it anyway, though), but for a developer (like myself) it's necessary with intervals of ~6 months - sigh (but that's due to DLL bloat, most of the time). If the NTFS one fails (rather unlikely barring physical disk damage) you can repair it using any of the tools already available for that job. I never ever had to repair anything running on NTFS, though. Solid as rock.This approach has saved my a** more than a few times...