Atari held patents on hardware sprite collision detection. (XOR? Yes and no...) Considering that the 2600 shipped in 1977 with this technology, the patents have probably expired long ago... --
Check your history, bub -- "FUD" was invented far earlier -- in the 1970s or 80s, and was used to describe the 'preannounce' marketing tactics of IBM!
(Perhaps after IBM had it's ass dragged through antitrust court for 15 years, they shaped up and enjoyed throwing FUD label at Microsoft. But "FUD" seems so un-IBMish, so I think it mostly the Team Warp vulgarians that enjoyed that term.)
As a side note, FUD is usually defined as "Fear, Uncertaintly, Doubt", but the original definition I read was "Flaming User Demand" (by preannouncing products to defuse competittors).
--
Re:Lower-case / capital letters for filenames
on
FreeBSD VM Design
·
· Score: 1
This behavior can be defeated with the IE4 ActiveDesktop. --
Re:PLus: services don't need a gui.
on
FreeBSD VM Design
·
· Score: 1
Yes, Citrix (where WTS came from) was specifically trying to emulate the behavior of X in a NT or OS2 context. Microsoft held them off at an arms length until they figured out that is what some customers really want. --
I don't get it -- So you have enterprise support from Entex. Then you decide to add 1500 Linux servers to do what-not. Who are you going to call? LinuxCare or Entex?
My guess is that the first call will be to Entex, because they already have the contract (and perhaps right of first refusal). Only if they won't or can't support Linux will your management make another call. Score: Entex 1, LinuxCare 0.
Support/Outsourcing companies work on a "Don't worry, we'll take care of it all." basis. Some like EDS can support you from your PC XT to your latest mainframe. I just don't see the market for a company that only supports one OS, and one with a small market share at that.
With the exception of the open source celebs, LinuxCare is just another small time Unix support company, except they only support one Unix. But, now you can prove me wrong and get rich quick. --
Seagate and IBM use the same 10K RPM drive for both "server" use and "audio/video" use.
I think what you are referring to is the fact that Quantum (?) was having trouble with the fast drives and thought they could get away with labelling them "A/V" drives and unloading them. I know people who do A/V work, and bit-accuracy is just as required as any other application. (Sure, for some applications you can lose some data in the middle of a video frame, but if it's metadata, or file system data, you are just as screwed as if it were a company database. Seems like Seagates are pretty popular with these folks, although the drives do run hot.) --
When the Q950 was shipping, Apple sold it as a 33Mhz CPU (just as Moto had labeled it). Some point later, Apple retroactively doubled the clock speed on all 68040s for marketing reasons. It was a debatable decision, most people quote the old clock speeds. --
but many of the SCSI drives out there now are almost identical to their ATAPI counterparts
If this is true why do IDE drives come in strange sizes like 14GB or 22GB, while SCSI drives traditionally come as 9GB/18GB/36GB? Honest question -- if they were the same mechanisms, one would expect the same sizes.
Also, looking at IBM's lineup, IDE drives are marketed as "DeskStar" while SCSI drives are marketed as "UltraStar" -- Is this truely all marketing, or are they using better components in the "server" drives? --
AXP, sparc or even MIPS has a better architecture than the x86
Maybe, in academic theory, or if you write asm, but in the real world, a CPU architecture is only as good as the software that runs on it and how easy it is to integrate it into particular devices.
Maybe you are thinking about your kick-ass SGI workstation or the humungo Sun E10K in the server room? But, in both those cases, the CPU is the *least* interesting thing about the hardware. It would be awful difficult for the Caruso to emulate SGI's memory bandwidth or Sun's swappable CPUs, and besides, what's the market for a handheld MIPS computer anyway? People don't buy these things for the CPUs, which is exactly why Sun and SGI are covering their bets with support for Intel Merced.
It is kind of disappointing that the native ISA wasn't released, but x86 is the natural choice, given both the huge amount of software and ultra inexpensive support hardware. (Even Linux users which have a large range of platforms available to them by-in-large install binaries and run on commodity PC platforms) --
One thing to realize is that the "mainframe to PC" revolution didn't shrink the market for mainframes and minis-err-I-mean-servers, it in fact increased it quite a bit.
The real game is about expanding the market, not playing zero-sum ball with Wintel, DEC, IBM, or whoever the dominant player of the day might be. Of course, MS, Intel, Sun, IBM and everyone else realizes this, which is why they are all making investments into this space. --
The key is really the market the CPU is designed for. Alphas are only feasible for high end servers and workstations. MacOS's emulator was designed for Macintoshes only. The AS/400's emulator was designed for AS/400's only. Hardware Java Machines haven't shown up in realworld product yet.
The "paradigm shift" is that you can now run x86 binaries at a reasonable speed with very little power and heat. If the Crusoe required Alpha or PIII-like watts, it would hardly be a big deal. --
PowerPC was nice, but they never ran the old 680x0 code worth beans.
Actually, even a lowly Performa with a 603e/120 can run 68K code in emulation faster than the fastest real 040 Macintosh Quadra, which topped out at 40Mhz, I think.
The key to emulation is to have a CPU that's fast enough to make the emulation transparent. Early on, the PPC guys predicted their 'scalability' would be so great that they would soon be running x86 in emulation faster than any real x86 processor.
Of course, Intel wasn't about to let that happen, so Transmeta is taking the wise approach at aiming for a different market (laptops and handhelds that Intel couldn't touch without losing tons of performance.) --
Netscape was never in the browser business
on
Free Be
·
· Score: 1
Nope, Netscape never made much money from their browser. In fact their original business plan was that it would be free for all users, and would serve as the best advertising money could buy for their server products.
However, Netscape found that corporations and individuals were demanding 'support', and so they begun charging a nominal fee (~$30) for a supported version of the product. I believe this policy started with version 1.1. Corporations had to pay for the supported version of the software, and individuals did to get a manual and save the hassle of downloading over a 28.8K link.
(In fact the corporation I worked for at the time refused to deploy NS 1.0/1.1 to a few thousand desktops unless Netscape took their money and entered into a support contract. Since Netscape was so small at that time, they were a little confused at this proposition.)
They never made any attempt to restrict Internet distribution to individuals -- their "90 day evaluation period" was essentially bunk because a new patch level or beta was usually released within 90 days. Not that they ever tried enforcing this restriction, even though they easily could have.
As for where Netscape made it's money (or didn't), it was primarily their web server product. They did get a little traction with the mail/directory/calendar/certificate server line up, but by then most of the potential customers (large corporations) had standardized on Lotus Notes or MS Exchange. --
Actually, I doubt there's much useful in there for the modern Quake spoiled player. Perhaps the monster AI, but if you look at some screenshots it's clear that the graphic engine is not up to snuff with modern game programs.
The best thing about Marathon was the story -- The background plot and how it was revealed to you really kept the game interesting. (Has any Id game since Castle Wolfenstien ever had a plot?) If all of the game content has been GPLed, I would love to see Marthon's plot and terminal text ported to a modern Quake II/III engine.
For those who are interested: The Marathon Story site has a ton of background information on the plot and various references in the game. --
From the info I could find, Quake was released July 1996. Marathon 1 shipped XMas 1994, and Marathon 2 for the Mac shipped XMas 1995 -- Windows sometime later. So the games weren't out at the same time...
As a side note, I remember thinking that Marathon looked really good on a Mac Quadra machine. However, seeing some screenshots just now, it does seem kind of cheezy, all blobby and bright colored. Plus the 'viewer' seems to be tiny. Strange how Quake (I,II,III) can spoil you. --
Re:Isn't this blown all out of proportion?
on
Hole in GNU GPL?
·
· Score: 1
One thing that strikes me in this discussion is that the assumptions that went into the GPL a from an earlier by-gone era when computers were few and software was vertical market by design. When GNU was created, distribution was somewhat difficult, and the market for any particular package was tiny.
The underlying assumption seems to be that any particular GPL user would be unlikely to want to get into the distribution business (because making tapes was difficult and expensive), and that computer time was so precious that any particular company would be unlikely to install and modify software that they didn't specifically need. The point was to ensure that the software licencees had the source, not that everyone had the source.
Now this scenario seems fine and dandy when there are only 45 sites using a particular PDP accounting package. But what it didn't fully anticipate was the current scenario where distribution is easy (Internet, CD-ROM), and computer usage and storage space is widespread. People have automatically come to assume that GPL == downloadable for free, when that is really a side effect of high powered personal computers and not a feature 'built-in' to the GPL.
So, think of the NSA deal as more of a "classic" use of the GPL, where one party's modifications are privately distributed to a trusted source. The point is that the NSA gets the right to modify the source code, not that you do. --
Actually, in the US, "employment" is very specifically defined in the law. I doubt very much that you could legally sell employment in your corporation for $10.
You could perhaps hire people as "consultants" or "beta program members", and require them to sign a NDA before getting your binary. It could work, but this is so far outside the normal method of selling software that most potential customers would probably choose to pass.
Of course, all of this is hypothetical crazy talk -- the largest value in open source code is the worldwide group of open source developers backing it. (Despite it's overinflated value, I don't think anyone would disagree that RedHat is a more valuable corporation than BSDi, with their closed source.) In most cases, trying to get out from under the GPL would be a fruitless struggle -- you would need a massive engineering budget to keep up with the closed source version. --
The point is that "skins" are a system-wide feature on Macintoshes, although the "appearance manager" currently only supports one Apple-blessed appearance.
Mac users are lazy bastards who want everything to look right just out of the box. Unlike Linux users, they'd rather bitch than manage skin files. (This is the right attitude, IMHO, because unlike Microsoft, Netscape is ignoring the OS features staring them in the face.)
Apple could modify Mozilla to support the appearance manager, but with the current IE deal, I wouldn't expect it. If a third party solves the problem, hopefully Netscape will include it in the standard package. --
If either IE4 or IE5 explodes, it's pretty clear that your Windows (9x?) install was hosed to begin with. Sucks, but that's Windows. Go find FORMAT.EXE.
My Win advice -- run NT from a NTFS system partition. It usually stays stable for production use. Boot W98 to play games. --
Re:Design of Mozilla must address fears of busines
on
Mozilla Status Update
·
· Score: 1
That's because the only "standard" in the real world is the behavior of your customer's browsers.
Mozilla 5 and IE 5 are nice steps towards a standards-based future, but realistically you need will need to target and test your DHTML for a long time. (For example, otherwise DOM compliant IE JavaScript will probably contain a "document.all", which is a microsoftism that Netscape made the political decision not to support.)
Even in the utopian future where NN and IE are fully standards compliant, as a web developer you will be then worrying about the broken JavaScript implementation in PalmPilots and Lincoln Navigators. As others have said, write HTML 3.2, or design around the problem, depending what the budget is. --
Yes, I know this is about the kernel. I'll leave it up to the critical slashdot reader (hah!) to determine if it's a hardware, installation, or Mozilla problem. --
"Most likely there is nothing wrong with your installation, your compiler or kernel. It very likely has something to do with your hardware. There are a variety of subsystems that can be wrong, and there is a variety of ways to fix it. Read on, and you'll find out more. There are two exceptions to this "rule". You could be running low on virtual memory, or you could be installing Red Hat 5.x or 6.x. There is more about this near the end." --
I read a "Stereo Review" article once that indicated that their selected "golden ears", in a blind test, couldn't tell the difference between a $1000 CD player, a $150 CD player, and a well mastered Dolby D cassette tape. (For obvious reasons, they don't print reviews like that anymore...)
The key here is "well mastered" and "Dolby D" -- too little, too late for the cassette as a hifi medium. I suspect it will be the same for MP3 and MiniDisc. Most people will probably wait for the next digital format, and judge the "high fidelity" quality then.
(Although, with the average set of crap speakers on the market, MP3, MiniDisc, good cassette, etc. is just fine.)
Slashdot Mirror
Atari held patents on hardware sprite collision detection. (XOR? Yes and no...) Considering that the 2600 shipped in 1977 with this technology, the patents have probably expired long ago...
--
Check your history, bub -- "FUD" was invented far earlier -- in the 1970s or 80s, and was used to describe the 'preannounce' marketing tactics of IBM!
(Perhaps after IBM had it's ass dragged through antitrust court for 15 years, they shaped up and enjoyed throwing FUD label at Microsoft. But "FUD" seems so un-IBMish, so I think it mostly the Team Warp vulgarians that enjoyed that term.)
As a side note, FUD is usually defined as "Fear, Uncertaintly, Doubt", but the original definition I read was "Flaming User Demand" (by preannouncing products to defuse competittors).
--
This behavior can be defeated with the IE4 ActiveDesktop.
--
Yes, Citrix (where WTS came from) was specifically trying to emulate the behavior of X in a NT or OS2 context. Microsoft held them off at an arms length until they figured out that is what some customers really want.
--
I don't get it -- So you have enterprise support from Entex. Then you decide to add 1500 Linux servers to do what-not. Who are you going to call? LinuxCare or Entex?
My guess is that the first call will be to Entex, because they already have the contract (and perhaps right of first refusal). Only if they won't or can't support Linux will your management make another call. Score: Entex 1, LinuxCare 0.
Support/Outsourcing companies work on a "Don't worry, we'll take care of it all." basis. Some like EDS can support you from your PC XT to your latest mainframe. I just don't see the market for a company that only supports one OS, and one with a small market share at that.
With the exception of the open source celebs, LinuxCare is just another small time Unix support company, except they only support one Unix. But, now you can prove me wrong and get rich quick.
--
Seagate and IBM use the same 10K RPM drive for both "server" use and "audio/video" use.
I think what you are referring to is the fact that Quantum (?) was having trouble with the fast drives and thought they could get away with labelling them "A/V" drives and unloading them. I know people who do A/V work, and bit-accuracy is just as required as any other application. (Sure, for some applications you can lose some data in the middle of a video frame, but if it's metadata, or file system data, you are just as screwed as if it were a company database. Seems like Seagates are pretty popular with these folks, although the drives do run hot.)
--
No mac shipped with an 060, so yer wrong there.
When the Q950 was shipping, Apple sold it as a 33Mhz CPU (just as Moto had labeled it). Some point later, Apple retroactively doubled the clock speed on all 68040s for marketing reasons. It was a debatable decision, most people quote the old clock speeds.
--
but many of the SCSI drives out there now are almost identical to their ATAPI counterparts
If this is true why do IDE drives come in strange sizes like 14GB or 22GB, while SCSI drives traditionally come as 9GB/18GB/36GB? Honest question -- if they were the same mechanisms, one would expect the same sizes.
Also, looking at IBM's lineup, IDE drives are marketed as "DeskStar" while SCSI drives are marketed as "UltraStar" -- Is this truely all marketing, or are they using better components in the "server" drives?
--
AXP, sparc or even MIPS has a better architecture than the x86
Maybe, in academic theory, or if you write asm, but in the real world, a CPU architecture is only as good as the software that runs on it and how easy it is to integrate it into particular devices.
Maybe you are thinking about your kick-ass SGI workstation or the humungo Sun E10K in the server room? But, in both those cases, the CPU is the *least* interesting thing about the hardware. It would be awful difficult for the Caruso to emulate SGI's memory bandwidth or Sun's swappable CPUs, and besides, what's the market for a handheld MIPS computer anyway? People don't buy these things for the CPUs, which is exactly why Sun and SGI are covering their bets with support for Intel Merced.
It is kind of disappointing that the native ISA wasn't released, but x86 is the natural choice, given both the huge amount of software and ultra inexpensive support hardware. (Even Linux users which have a large range of platforms available to them by-in-large install binaries and run on commodity PC platforms)
--
One thing to realize is that the "mainframe to PC" revolution didn't shrink the market for mainframes and minis-err-I-mean-servers, it in fact increased it quite a bit.
The real game is about expanding the market, not playing zero-sum ball with Wintel, DEC, IBM, or whoever the dominant player of the day might be. Of course, MS, Intel, Sun, IBM and everyone else realizes this, which is why they are all making investments into this space.
--
The key is really the market the CPU is designed for. Alphas are only feasible for high end servers and workstations. MacOS's emulator was designed for Macintoshes only. The AS/400's emulator was designed for AS/400's only. Hardware Java Machines haven't shown up in realworld product yet.
The "paradigm shift" is that you can now run x86 binaries at a reasonable speed with very little power and heat. If the Crusoe required Alpha or PIII-like watts, it would hardly be a big deal.
--
"Kiosks" that played music were run using vacuum tube technology for many years. They're called jukeboxes.
(If you can be a weenie, so can I!)
--
PowerPC was nice, but they never ran the old 680x0 code worth beans.
Actually, even a lowly Performa with a 603e/120 can run 68K code in emulation faster than the fastest real 040 Macintosh Quadra, which topped out at 40Mhz, I think.
The key to emulation is to have a CPU that's fast enough to make the emulation transparent. Early on, the PPC guys predicted their 'scalability' would be so great that they would soon be running x86 in emulation faster than any real x86 processor.
Of course, Intel wasn't about to let that happen, so Transmeta is taking the wise approach at aiming for a different market (laptops and handhelds that Intel couldn't touch without losing tons of performance.)
--
Nope, Netscape never made much money from their browser. In fact their original business plan was that it would be free for all users, and would serve as the best advertising money could buy for their server products.
However, Netscape found that corporations and individuals were demanding 'support', and so they begun charging a nominal fee (~$30) for a supported version of the product. I believe this policy started with version 1.1. Corporations had to pay for the supported version of the software, and individuals did to get a manual and save the hassle of downloading over a 28.8K link.
(In fact the corporation I worked for at the time refused to deploy NS 1.0/1.1 to a few thousand desktops unless Netscape took their money and entered into a support contract. Since Netscape was so small at that time, they were a little confused at this proposition.)
They never made any attempt to restrict Internet distribution to individuals -- their "90 day evaluation period" was essentially bunk because a new patch level or beta was usually released within 90 days. Not that they ever tried enforcing this restriction, even though they easily could have.
As for where Netscape made it's money (or didn't), it was primarily their web server product. They did get a little traction with the mail/directory/calendar/certificate server line up, but by then most of the potential customers (large corporations) had standardized on Lotus Notes or MS Exchange.
--
Actually, I doubt there's much useful in there for the modern Quake spoiled player. Perhaps the monster AI, but if you look at some screenshots it's clear that the graphic engine is not up to snuff with modern game programs.
The best thing about Marathon was the story -- The background plot and how it was revealed to you really kept the game interesting. (Has any Id game since Castle Wolfenstien ever had a plot?) If all of the game content has been GPLed, I would love to see Marthon's plot and terminal text ported to a modern Quake II/III engine.
For those who are interested: The Marathon Story site has a ton of background information on the plot and various references in the game.
--
From the info I could find, Quake was released July 1996. Marathon 1 shipped XMas 1994, and Marathon 2 for the Mac shipped XMas 1995 -- Windows sometime later. So the games weren't out at the same time...
As a side note, I remember thinking that Marathon looked really good on a Mac Quadra machine. However, seeing some screenshots just now, it does seem kind of cheezy, all blobby and bright colored. Plus the 'viewer' seems to be tiny. Strange how Quake (I,II,III) can spoil you.
--
One thing that strikes me in this discussion is that the assumptions that went into the GPL a from an earlier by-gone era when computers were few and software was vertical market by design. When GNU was created, distribution was somewhat difficult, and the market for any particular package was tiny.
The underlying assumption seems to be that any particular GPL user would be unlikely to want to get into the distribution business (because making tapes was difficult and expensive), and that computer time was so precious that any particular company would be unlikely to install and modify software that they didn't specifically need. The point was to ensure that the software licencees had the source, not that everyone had the source.
Now this scenario seems fine and dandy when there are only 45 sites using a particular PDP accounting package. But what it didn't fully anticipate was the current scenario where distribution is easy (Internet, CD-ROM), and computer usage and storage space is widespread. People have automatically come to assume that GPL == downloadable for free, when that is really a side effect of high powered personal computers and not a feature 'built-in' to the GPL.
So, think of the NSA deal as more of a "classic" use of the GPL, where one party's modifications are privately distributed to a trusted source. The point is that the NSA gets the right to modify the source code, not that you do.
--
Here is his original
I don't think he's a troll, just confused.
--
Actually, in the US, "employment" is very specifically defined in the law. I doubt very much that you could legally sell employment in your corporation for $10.
You could perhaps hire people as "consultants" or "beta program members", and require them to sign a NDA before getting your binary. It could work, but this is so far outside the normal method of selling software that most potential customers would probably choose to pass.
Of course, all of this is hypothetical crazy talk -- the largest value in open source code is the worldwide group of open source developers backing it. (Despite it's overinflated value, I don't think anyone would disagree that RedHat is a more valuable corporation than BSDi, with their closed source.) In most cases, trying to get out from under the GPL would be a fruitless struggle -- you would need a massive engineering budget to keep up with the closed source version.
--
The point is that "skins" are a system-wide feature on Macintoshes, although the "appearance manager" currently only supports one Apple-blessed appearance.
Mac users are lazy bastards who want everything to look right just out of the box. Unlike Linux users, they'd rather bitch than manage skin files. (This is the right attitude, IMHO, because unlike Microsoft, Netscape is ignoring the OS features staring them in the face.)
Apple could modify Mozilla to support the appearance manager, but with the current IE deal, I wouldn't expect it. If a third party solves the problem, hopefully Netscape will include it in the standard package.
--
If either IE4 or IE5 explodes, it's pretty clear that your Windows (9x?) install was hosed to begin with. Sucks, but that's Windows. Go find FORMAT.EXE.
My Win advice -- run NT from a NTFS system partition. It usually stays stable for production use. Boot W98 to play games.
--
That's because the only "standard" in the real world is the behavior of your customer's browsers.
Mozilla 5 and IE 5 are nice steps towards a standards-based future, but realistically you need will need to target and test your DHTML for a long time. (For example, otherwise DOM compliant IE JavaScript will probably contain a "document.all", which is a microsoftism that Netscape made the political decision not to support.)
Even in the utopian future where NN and IE are fully standards compliant, as a web developer you will be then worrying about the broken JavaScript implementation in PalmPilots and Lincoln Navigators. As others have said, write HTML 3.2, or design around the problem, depending what the budget is.
--
Yes, I know this is about the kernel. I'll leave it up to the critical slashdot reader (hah!) to determine if it's a hardware, installation, or Mozilla problem.
--
Check out http://www.bitwizard.nl/sig11/
"Most likely there is nothing wrong with your installation, your compiler or kernel. It very likely has something to do with your hardware. There are a variety of subsystems that can be wrong, and there is a variety of ways to fix it. Read on, and you'll find out more. There are two exceptions to this "rule". You could be running low on virtual memory, or you could be installing Red Hat 5.x or 6.x. There is more about this near the end."
--
I read a "Stereo Review" article once that indicated that their selected "golden ears", in a blind test, couldn't tell the difference between a $1000 CD player, a $150 CD player, and a well mastered Dolby D cassette tape. (For obvious reasons, they don't print reviews like that anymore...)
The key here is "well mastered" and "Dolby D" -- too little, too late for the cassette as a hifi medium. I suspect it will be the same for MP3 and MiniDisc. Most people will probably wait for the next digital format, and judge the "high fidelity" quality then.
(Although, with the average set of crap speakers on the market, MP3, MiniDisc, good cassette, etc. is just fine.)
--