Am I the only person here who intentionally stopped upgrading Microsoft operating systems at Windows 2000?
Not me. XP is really only a minor bump to 2000, as near as I can tell, with some extra software like the multiuser stuff they got from Citrix bundled in with it.
It kind of feels like they came out with XP because they wanted an excuse to put the timed suicide bomb in to enforce registration.
That's creepy. I haven't tried running XP in really tight quarters but I'm running 2000 in 64M on a P266 (not even MMX) and it doesn't hurt a bit. Oh, a lot of apps are too bulky for this weeny box, but it's just fine for web and mail stuff.
Here's what he says about the requirements for Longhorn:
Longhorn will run fine on a 1 GHz computer with 256 MB of RAM, according to Microsoft corporate vice president Joe Peterson at the blogger lunch today. This is good news for today's PC users, some of whom are concerned that they won't have the PC muscle needed to run the next Windows. Of course, those people won't get the full-blown Aero Glass experience, but it will still work fine.
They're talking about a box that's somewhere between my Beige G4 (Beige G3 upgraded with a G4/466 CPU) and my Mac mini... and closer to my Mac mini. Both of these boxes run OS X with the full blown Quartz Extreme experience just fine.
It looks like Microsoft has finally leapfrogged Apple in at least one area... excessive hardware requirements. The whole point to QE is that because it's handled in the GPU it doesn't matter if the CPU is a bit wimpy... the CPU isn't doing any of the heavy lifting. Did Paul misunderstand Joe or did Microsoft miss the point?
I just pointed you at the fact that many things have been said about why exactkly the scheduler and priority management of OS/2 was better, you can ignore that and just keep pushing your own little love for Unix.
My love for UNIX is obviously why I tried to hide how the UNIX graphical user interface is poorly suited to today's desktop clients. Oh, whoops, I didn't, did I?
Unix [...] doesn't do better then OS/2
Um, I didn't say it did.
I am comparing OS/2 in CHARACTER mode, NO GUI running whatsoever, with Unix in CHARACTER mode, so again, NO GUI whatsoever.
And what exactly are you saying about OS/2 in CHARACTER MODE and UNIX in CHARACTER MODE? I want to be certain here, because way back in this thread I said that the differences between OS/2 scheduler and the UNIX scheduler weren't fundamentally important for graphical user interfaces, and both were doing about the same thing, and you just accused me of saying the UNIX scheduler did better than OS/2. Which isn't something I've said, or attempted to imply... and if I inadvertently confused you, I apologise.
But, anyway, you see how easy it is to get confused in this kind of exchange, especially when one of us isn't bothering to clearly distinguish between what they're quoting and what they're writing, so I'd really like to be sure that I'm not misunderstanding you like that.
that mouse pointer is almost completely handled in hardware, and you do still get interupts
The mouse pointer isn't controlled from an interrupt, it's crontrolled by Intuition, from an ordinary Amiga Exec task.
And with a modern accelerated video card, the mouse is almost completely handled in hardware too.
So if the fine details of the scheduler was that important to maintaining the GUI's responsive behaviour, the simplistic scheduler on the Amiga should have been a major drawback. Right?
Really, it uses LibGAIM to customise the user interface the way this new AOL IM client uses Gecko? Silly me, I thought it was using Webkit the way AOL is using Gecko. Thanks ever so much for correcting me.
Actually they treated us pretty good [good customer service story deleted]
I have no complaints with their customer service. It's their software that burned me... over and over and over again.
xenix wouldn't allow both IDE and SCSI busses on the same system
Yeh, that kind of thing. And the driver configuration. And the horrible things they did to System V system configuration. And the driver configuration again, because they kept changing it. And doing the same config in 3 places with 3 different tools because the file formats were undocumented. And Secureware. And... oh, god, I can't do this. I WILL stop now.
If they're paying attention to IIS surely they're just as likely to reduce the rankings for certain types of sites on IIS (SEO spam link sites, for example) as increase the rankings for good sites on IIS
Let's go over this again.
What I'm suggesting is that they are going to tend to improve the ratings of sites they pay more attention to, even if they're not consciously trying to do that, because they're more likely to notice them missing from the results and be aware of other problems with them, and have those things in mind when they're working on their rules.
Which means that sites they use every day, including their intranet sites and their own sites, are going to end up at least slightly selected for. Not because they're using IIS, but because they're paying attention to those sites.
And because they're at Microsoft, a much higher percentage of those sites are going to be running on IIS, just because they're eating their own dogfood. And that means that any systematic difference between the way IIS serves pages and Apache serves pages is going to be selected for as their rules evolve. Maybe not a lot, but you don't need much of a selection effect to make a noticable result in the results.
So, it doesn't matter whether spam link sites are using IIS or not, those sites aren't the ones that are part of this natural, unconscious, and inevitable selection process.
You're notgoing to program quantum computers in anything like Java. The languages currently in use that are most likely to work well on these kinds of systems are things like Prolog.
and offers only Intel® XScale(TM) processors in their handhelds
Much as I like bashing Dell, since Intel bought the StrongARM technology from DEC the Pocket PC processor naturally followed. It's not like Dell or even Microsoft is deliberately picking Intel here, they actually picked DEC and DEC sold the business to Intel.
Remember, the guys working on the MSN search engine certainly use IIS to host their intranet sites, and whatever internal webservers they use to test against are probably IIS as well, at least in the most cases. They are likely to consider bogus results for their own sites (both internal and external) more critical... that's not malice, that's just human nature. Even if they consciously work against that, they're more likely to notice problems there first.
And search engine tweaking is more an art than a science. It's an evolutionary process, with feedback loops and strange attractors. So if there's any difference in the behaviour or design of Apache or IIS that would be visible to a search engine, it's likely to lead to a slight bias in favor of the server software that the servers they pay more attention to run.
Then use size_t instead of unsigned long : it's 32bit on Win32 and 64bit on Win64.
Well, you know, in the quarter of a century that I've been writing C code on everything from an 8080 to an Alpha I *have* in fact run across this concept before. That's not the point. The point is that this design cripples the architecture by making its most useful feature harder to use while making its least useful feature even more of an in-your-face stumbling block than it was on the Alpha.
Then you will be able to cross-compile 64-bit Windows applications.
I expect the sun will be a small hard lump of coal the size of your head long before I discover any interest in cross-compiling Windows applications. I suppose, really, I should be happy at this bit of Microsoft stupidity, except that this kind of stupidity annoys me no matter who's doing it.
When IBM and MS introduced OS/2 in the late 80s, they also filled quite a few pages about why the OS/2 scheduler does really a much better job at those things.
Well, yeh, of course they did. They have to tell you that all their design decisions are great and why they made them. Every major piece of software has lots of very smart people writing lots of copy about why they did the things they did. They're not always right.
For example, the Amiga scheduler didn't do priority promotion AT ALL, and you could tell, because when you had a background task running at 100% it hurt everything. BUT, the GUI was incredibly fast and responsive, and the pointer tracking never fumbled.
Yeah sure, that is why you see the exact thing for non GUI (ie, character based) applications.
You do see the same thing on UNIX for non-GUI applications. Interactive applications are responsive no matter what the CPU load is, so long as the memory load doesn't get so high they're deferred by swapping or excessive paging. Unless you're talking about like that, or you're seeing the behaviour of the GUI around the character mode application (you know, right, that when you run a character application inside an X-term the user interface is still handled by the X11 GUI), you're just plain wrong there. The responsiveness of UNIX character-based applications under excessive CPU loading is phenomenally good. Legendary, even.
putting some limits on how long a task can spend in the kernel while blocking the rest of the system helps as well for example.
I'm having another "But, Doctor Evil, that has ALSO already been invented" moment here, but I better stop before I start using sarcasm.
The released Alpha version was 32-bit, but they developed 64-bit Windows on Alpha and had a 64-bit version internally years ago.
I don't know what the problem is, I mean with the LLP64 IL32 compiler model this really doesn't count as a 64-bit version. It's wasting half the capacity of the register bank by not having native code use the upper half of the words directly... it's almost like this was developed for some depraved 32-bit segmented mode on Xeon or something bloody stupid like that.
Tru64 let you use 32- or 64- bit pointers on a per-application basis. Why didn't Microsoft do that, so you could at least take advantage of the improved instruction set without having to break all your Win32 code? Just recompile it in their equivalent of -taso.
Long is often the same size as int because Unix 32V (the first VAX UNIX) didn't increase the size of long going from the PDP-11 (L32 IP16) to the VAX.
Back in 1980 when I started porting code from the PDP-11 that was a surprise, long had always been longer than int up to then. I thought long should have been 64-bit back then... it was much more logical to keep a reasonable size hierarchy and use derived types and defines when size mattered. But I never imagined that people would still be using 32-bit longs on machines with 64-bit registers 25 years later. If I'd thought about it I guess I would have expected 128 bit longs, 64 bit ints, and we'd wear new short shorts or something to fill in the gap. That would give you a type that'd fit in the SSE registers.
But IL32 on a 64-bit machine? That's evil and rude.
Dude, you can get an external DVD burner that's better than a superdrive for less than the cost of adding a superdrive and removing the internal combo... even if Apple does it.
This little widget costs about $40 more than a comparable firewire external drive. If you need that storage, you need it whether you've got a mini or an iMac G5, so you're not actually able to say "you can get an iMac G5 for less". In fact, if you can get an iMac G5 for the $40 difference between a LaCie external and this, I want to know where you shop.
Slashdot Mirror
Thus I can see why they'd not give two shits about the UI.
That's why they show us a different one every time? They need to quit messing with ugly UIs until they've got a better one than Windows 2000 did.
And they're still taking after Apple's appalling metallic look.
Am I the only person here who intentionally stopped upgrading Microsoft operating systems at Windows 2000?
Not me. XP is really only a minor bump to 2000, as near as I can tell, with some extra software like the multiuser stuff they got from Citrix bundled in with it.
It kind of feels like they came out with XP because they wanted an excuse to put the timed suicide bomb in to enforce registration.
BOY do they need help.
I wonder if Max Rudberg would be willing to help them fine-tune them.
That desktop background looks quite evocative.
Here's what he says about the requirements for Longhorn:They're talking about a box that's somewhere between my Beige G4 (Beige G3 upgraded with a G4/466 CPU) and my Mac mini... and closer to my Mac mini. Both of these boxes run OS X with the full blown Quartz Extreme experience just fine.
It looks like Microsoft has finally leapfrogged Apple in at least one area... excessive hardware requirements. The whole point to QE is that because it's handled in the GPU it doesn't matter if the CPU is a bit wimpy... the CPU isn't doing any of the heavy lifting. Did Paul misunderstand Joe or did Microsoft miss the point?
I just pointed you at the fact that many things have been said about why exactkly the scheduler and priority management of OS/2 was better, you can ignore that and just keep pushing your own little love for Unix.
My love for UNIX is obviously why I tried to hide how the UNIX graphical user interface is poorly suited to today's desktop clients. Oh, whoops, I didn't, did I?
Unix [...] doesn't do better then OS/2
Um, I didn't say it did.
I am comparing OS/2 in CHARACTER mode, NO GUI running whatsoever, with Unix in CHARACTER mode, so again, NO GUI whatsoever.
And what exactly are you saying about OS/2 in CHARACTER MODE and UNIX in CHARACTER MODE? I want to be certain here, because way back in this thread I said that the differences between OS/2 scheduler and the UNIX scheduler weren't fundamentally important for graphical user interfaces, and both were doing about the same thing, and you just accused me of saying the UNIX scheduler did better than OS/2. Which isn't something I've said, or attempted to imply... and if I inadvertently confused you, I apologise.
But, anyway, you see how easy it is to get confused in this kind of exchange, especially when one of us isn't bothering to clearly distinguish between what they're quoting and what they're writing, so I'd really like to be sure that I'm not misunderstanding you like that.
that mouse pointer is almost completely handled in hardware, and you do still get interupts
The mouse pointer isn't controlled from an interrupt, it's crontrolled by Intuition, from an ordinary Amiga Exec task.
And with a modern accelerated video card, the mouse is almost completely handled in hardware too.
So if the fine details of the scheduler was that important to maintaining the GUI's responsive behaviour, the simplistic scheduler on the Amiga should have been a major drawback. Right?
Actually, Adium uses LibGaim
Really, it uses LibGAIM to customise the user interface the way this new AOL IM client uses Gecko? Silly me, I thought it was using Webkit the way AOL is using Gecko. Thanks ever so much for correcting me.
It is actually just the output of a java decompiler (jad).
Oh. <emilylatella>nevermind</>
Sounds like they were paying attention to Adium, which is webkit rather than gecko based but otherwise sounds similar.
Well, except that Adium works with just about every IM service out there.
How long until they close the door on third party IMs?
That was my first thought, indeed.
With luck George will realise that he doesn't have to lift another finger working for the rest of his life, and someone else will do them.
Isn't this 2.1?
http://www.swapdv.net/dvarchive/source/
Actually they treated us pretty good [good customer service story deleted]
I have no complaints with their customer service. It's their software that burned me... over and over and over again.
xenix wouldn't allow both IDE and SCSI busses on the same system
Yeh, that kind of thing. And the driver configuration. And the horrible things they did to System V system configuration. And the driver configuration again, because they kept changing it. And doing the same config in 3 places with 3 different tools because the file formats were undocumented. And Secureware. And... oh, god, I can't do this. I WILL stop now.
If they're paying attention to IIS surely they're just as likely to reduce the rankings for certain types of sites on IIS (SEO spam link sites, for example) as increase the rankings for good sites on IIS
Let's go over this again.
What I'm suggesting is that they are going to tend to improve the ratings of sites they pay more attention to, even if they're not consciously trying to do that, because they're more likely to notice them missing from the results and be aware of other problems with them, and have those things in mind when they're working on their rules.
Which means that sites they use every day, including their intranet sites and their own sites, are going to end up at least slightly selected for. Not because they're using IIS, but because they're paying attention to those sites.
And because they're at Microsoft, a much higher percentage of those sites are going to be running on IIS, just because they're eating their own dogfood. And that means that any systematic difference between the way IIS serves pages and Apache serves pages is going to be selected for as their rules evolve. Maybe not a lot, but you don't need much of a selection effect to make a noticable result in the results.
So, it doesn't matter whether spam link sites are using IIS or not, those sites aren't the ones that are part of this natural, unconscious, and inevitable selection process.
You're notgoing to program quantum computers in anything like Java. The languages currently in use that are most likely to work well on these kinds of systems are things like Prolog.
and offers only Intel® XScale(TM) processors in their handhelds
Much as I like bashing Dell, since Intel bought the StrongARM technology from DEC the Pocket PC processor naturally followed. It's not like Dell or even Microsoft is deliberately picking Intel here, they actually picked DEC and DEC sold the business to Intel.
Remember, the guys working on the MSN search engine certainly use IIS to host their intranet sites, and whatever internal webservers they use to test against are probably IIS as well, at least in the most cases. They are likely to consider bogus results for their own sites (both internal and external) more critical... that's not malice, that's just human nature. Even if they consciously work against that, they're more likely to notice problems there first.
And search engine tweaking is more an art than a science. It's an evolutionary process, with feedback loops and strange attractors. So if there's any difference in the behaviour or design of Apache or IIS that would be visible to a search engine, it's likely to lead to a slight bias in favor of the server software that the servers they pay more attention to run.
Then use size_t instead of unsigned long : it's 32bit on Win32 and 64bit on Win64.
Well, you know, in the quarter of a century that I've been writing C code on everything from an 8080 to an Alpha I *have* in fact run across this concept before. That's not the point. The point is that this design cripples the architecture by making its most useful feature harder to use while making its least useful feature even more of an in-your-face stumbling block than it was on the Alpha.
Then you will be able to cross-compile 64-bit Windows applications.
I expect the sun will be a small hard lump of coal the size of your head long before I discover any interest in cross-compiling Windows applications. I suppose, really, I should be happy at this bit of Microsoft stupidity, except that this kind of stupidity annoys me no matter who's doing it.
I feel sorry for SCO's customers.
Me too, but I've always felt sorry for SCO's customers. I've been one. And, well, even when SCO was real SCO it wasn't very nice being a SCO customer.
I don't know if I want to go into details. I'd be hear all day, and I need to be watching my blood pressure.
When IBM and MS introduced OS/2 in the late 80s, they also filled quite a few pages about why the OS/2 scheduler does really a much better job at those things.
Well, yeh, of course they did. They have to tell you that all their design decisions are great and why they made them. Every major piece of software has lots of very smart people writing lots of copy about why they did the things they did. They're not always right.
For example, the Amiga scheduler didn't do priority promotion AT ALL, and you could tell, because when you had a background task running at 100% it hurt everything. BUT, the GUI was incredibly fast and responsive, and the pointer tracking never fumbled.
Yeah sure, that is why you see the exact thing for non GUI (ie, character based) applications.
You do see the same thing on UNIX for non-GUI applications. Interactive applications are responsive no matter what the CPU load is, so long as the memory load doesn't get so high they're deferred by swapping or excessive paging. Unless you're talking about like that, or you're seeing the behaviour of the GUI around the character mode application (you know, right, that when you run a character application inside an X-term the user interface is still handled by the X11 GUI), you're just plain wrong there. The responsiveness of UNIX character-based applications under excessive CPU loading is phenomenally good. Legendary, even.
putting some limits on how long a task can spend in the kernel while blocking the rest of the system helps as well for example.
I'm having another "But, Doctor Evil, that has ALSO already been invented" moment here, but I better stop before I start using sarcasm.
The released Alpha version was 32-bit, but they developed 64-bit Windows on Alpha and had a 64-bit version internally years ago.
I don't know what the problem is, I mean with the LLP64 IL32 compiler model this really doesn't count as a 64-bit version. It's wasting half the capacity of the register bank by not having native code use the upper half of the words directly... it's almost like this was developed for some depraved 32-bit segmented mode on Xeon or something bloody stupid like that.
Tru64 let you use 32- or 64- bit pointers on a per-application basis. Why didn't Microsoft do that, so you could at least take advantage of the improved instruction set without having to break all your Win32 code? Just recompile it in their equivalent of -taso.
Long is often the same size as int because Unix 32V (the first VAX UNIX) didn't increase the size of long going from the PDP-11 (L32 IP16) to the VAX.
Back in 1980 when I started porting code from the PDP-11 that was a surprise, long had always been longer than int up to then. I thought long should have been 64-bit back then... it was much more logical to keep a reasonable size hierarchy and use derived types and defines when size mattered. But I never imagined that people would still be using 32-bit longs on machines with 64-bit registers 25 years later. If I'd thought about it I guess I would have expected 128 bit longs, 64 bit ints, and we'd wear new short shorts or something to fill in the gap. That would give you a type that'd fit in the SSE registers.
But IL32 on a 64-bit machine? That's evil and rude.
And then you can make a Beowulf cluster of them!
Whoops, wrong thread.
Properly outfitted?
Dude, you can get an external DVD burner that's better than a superdrive for less than the cost of adding a superdrive and removing the internal combo... even if Apple does it.
This little widget costs about $40 more than a comparable firewire external drive. If you need that storage, you need it whether you've got a mini or an iMac G5, so you're not actually able to say "you can get an iMac G5 for less". In fact, if you can get an iMac G5 for the $40 difference between a LaCie external and this, I want to know where you shop.
Yeh, Orion was the hottest thing this side of project Pluto.