Ok, check out Transmeta. The seemed to find a way to use a fraction of the power of an X86 chip, run the same apps, and perform better with a lower clock speed.
If memory serves me, a 700MHz or so Crusoe was roughly equivalent to a 500MHz PII. Should be "performs equivalently with somewhat higher clock speed..."
Uh, I believe the point of these discussions was to illustrate why the Mac has had a small open source community. Comparing them in "the apples/oranges style of idiocy," as you put it, is merely a way of showing how far apart the camps are.
As an aside, I saw mentioned in other places many different variations on NewsWatcher and other open-sourced programs -- Mac open source seems to want to fork before it wants to integrate. Why have all these separate versions, each of which has some added functionality? Why not integrate them into a "official" release? There is either a lack of organization, or a lack of cooperation amongst the authors...
I don't see what your point was comparing the distribution strategies -- clarify, if you can. I agree that both groups do have things to share and need to learn from each other. And indeed, the two groups have, for the most part, evolved separately, but I don't see that as the main reason behind the Mac's small open source community.
(And, BTW, I sense an implication that I am somehow a Linux zealot. Please don't try that -- I develop software for Windows, use Linux as a hobby, and used to write Mac software in academia. I'd like to think that I am, if not totally unbiased, then at least somewhat informed about all three.)
3. Apple didn't want you developing for the Mac unless you werre a commercial licenser.
Nonsense and lies. They did show preferential treatment to commercial houses, though, because they were more likely to invest the time to learn the system and make high-quality software than early hobbyists.
And this is exactly why the Mac has a poor tradition of open source -- commercial houses were given preferential treatment, because Apple didn't believe that hobbyists were worth supporting.
If I'm not mistaken, your first two examples (NewsWatcher and NCSA Telnet) were only open-sourced after the authors found themselves no longer able (or willing) to update the software. This is in marked contrast to what is more commonly thought of as "Open Source", where the code is typically open right from the get-go (or very shortly thereafter).
This, I think, is the fundamental difference between Open Source on the Mac and everywhere else -- source release usually only happens when the original author "gives up." Look at Apple's GameSprockets -- they gave up on them, and opened them up. Ditto on the above packages, and some others that I recall from my Mac days.
Remember, though, that they're using the Windows kernel, not the whole operating system. There should be quite a bit more memory free than would normally be expected under Windows.
"way more frames of animation" == WAAAAAAAAAAYYYYY more memory for sprites.
Each frame is a rendering of the sprite, multiplied by (bits of color / 8) bytes. Suppose you have 500 frames of animation for a 32x32x16 character. That's one megabyte of memory, roughly, for a _tiny_ character. You want 1000 frames of animation? 2 megs of memory. And so on.
All of a sudden, 3D meshes don't look so bad after all.
But, an instruction can take more than one cycle to execute. It's difficult to predict exact amounts of instructions/second because of this. Soooo...your 7 million number was purely hypothetical, right?:)
No programming language could handle the case of (literally) "structA contain[ing] a structB and structB contain[ing] a structA" -- it's an infinitely recursive definition.
Now, if you meant "structA contains a pointer to a structB" and vice versa, then the solution would be to do multi-pass compilation. (I think. hehe)
They aren't as different as you might think. The language parser remains the same, it's just that the code generation portion of the compiler is different.
If I recall, Strassen's algorithm gets you down to something like O(n^2.8something), and then some related variants get you down to O(n^2.7). Not sure about other decompositions. It's been awhile since my algorithms class, and since I don't use these things every day, I tend to forget them quickly.:)
An excellent list of suggestions! Of course, there's not room in the top 10 for all of them, but I think that more than a couple are a match for the ones in their list...
They profit off of licensing games for the system. Even though the devkits are expensive, let's look at the realities of things...would you make more money selling 1000 $20K dev kits, or licensing millions of copies of games every year?
Telemate, man! Telemate! One of the only multithreaded DOS apps I ever saw. And the only DOS comm program that let you do other thing while downloading!
Unless, of course, you were file transferring with HS/Link, in which case you could chat with the sysop...:) (aside from the obvious benefits of uploading AND downloading at the SAME TIME!)
Of course, eventually I wound up switching over from Procomm Plus in DOS to...Procomm Plus in Windows! Then, I could play Solitaire in Win 3.1 while I downloaded:)
Why assume that games would need to be installed? The hard drive could be for patches/updates (either bugfixes or game enhancements -- think of all the mods for Quake and Half-Life out there and you'll see why this is exciting for a console), for savegames (obviously), and for caching stuff during gameplay (faster than reading from the CD). Seems like a smart thing to include to me.
The article doesn't say that there's only two machines. (It says something along the lines of "both operating systems were tested on old Pentium machines", not that "there were two machines, one running NT, one running Linux".) It actually doesn't say how many machines were in the study, IIRC.
"Both operating systems running on relatively old Pentium machines" != "Two machines, one running NT, one running Linux"
As far as I could tell from a quick re-scanning of the article, there was no indication of exactly how many machines there were. But they didn't say there were two, they said that the study used "relatively old Pentium machines."
The real damage is not in that you simply couldn't buy the book -- the damage is to their reputation and customer base. You'd probably go to Barnes & Noble if you couldn't get something from Amazon, right? I agree it's difficult to place a monetary value on the damages, but the situation isn't as cut-and-dried as "well I'll just go back once they're no longer DoS'd."
The situation with eBay and E-Trade is even worse, because a denial of service is preventing users from doing business with others, not just the company affected by the DoS.
I base my comments on developing for various versions of DirectX (from DX3 forward) under two different versions of Visual Studio. Sure, you need to convert the project files itself, but there's nothing inherent in DirectX itself that prevents you from using it with a newer/older version of a compiler (within reasonable limits -- i.e. using 16-bit compilers is a no-no), or even a competitor's product. Unlike Java, the C++ language does not change significantly every six months...
If you're going to turn this into a qualifications game, then I'd be curious to hear what your qualifications are. Your personal page on the site doesn't have much in the way of information...
As far as my "apologist" comments went, I suppose that came on a bit strong, but I really think that the situation between Nintendo and Sony is approaching that of the ages-old Apple-PC rivalry. Apple may have better hardware, but systems (and gaming systems in particular) are sold on the basis of software -- both quality and quantity. I don't need to remind anyone who is dominant in the personal computer marketplace (despite Apple's recent about-face).
The GameBoy is 10+ years old IIRC. On the one hand, it's amazing that things are still being developed for it, and that it's still a popular platform. On the other hand, the system is looking extremely long in the tooth...
Another factor to keep in mind is that margins are much more in favor of CD systems as opposed to cartridge-based systems. This is attractive not only to the developers but also the system manufacturer, as it will encourage more people to develop for the system.
P-a-a-leeeease. Both Nintendo and Sony are doing fine. Sony just has bigger sales numbers.
I hate to sound flippant, but "having bigger sales numbers" is pretty damn important to a company's success. Gaming is no longer a niche area, like how it was in the early '80s, where a variety of companies can carve out their own pieces of the market. If you're not in first place, you're endangered.
Nintendo may not be "poor" but they are certainly not in an enviable position right now -- their chief rival, who blew onto the marketplace a few years ago with all the subtlety of a tornado, is now preparing their next salvo. Developers are lining up by the truckload to put out games for the PS2, the hype is being built up (to the level where everyone knows about the PS2), and, most importantly, they're beating Nintendo to market.
I think that Nintendo is in the proverbial lull before the storm. Clear skies now, stormy weather ahead...
I fail to see any relationship between the version of a compiler being used and the version of DirectX being used, so "I'm going take if from your comment" that you don't know much about development.
And as far as DirectX being updated for WinCE, it would seem to be in Microsoft's interests to do so, since the X-Box will certainly use DirectX (and will hopefully not use a full-blown, slow version of Windows)...
I tend to agree with the earlier poster who labeled the article as "Nintendo apologist" -- Nintendo's success is currently based off of the Pokemon franchise (sickeningly so), and their market share has dwindled considerably in comparison to Sony, which came out of nowhere. Remember, the name of the economic game is not whichever system is better, but which system is the most dominant and prevalent in the home. Nintendo's not dead yet, but they better come packin' heat with the Dolphin or they will be DOA...
Let's not forget Henry Ford himself was famous for blatantly having thugs beat senseless employees who wanted to unionize. What do you think they'll want to do if you're emailing to incite sickouts?
Am I the only one who read this as "senseless employees who wanted to unionize," instead of "beating employees senseless"?:)
Not everyone is obligated to become a computer geek. Just because they aren't doesn't mean they don't have online rights that need to be looked out for.
This has to be the most insightful thing I've read in the discussion so far. We may all laugh at AOL or say "anyone who knows better would say 'no'", but the truth of the matter is that Joe Six-Pack shouldn't have his computer, or, for that matter, other competitors' products, disabled by another piece of software.
Regarding thought #1, the Win32 file system drivers are extensible -- you can plug in additional filesystems if you so desire, although I haven't really seen it done at all. (perhaps some of the esoteric "FTP filesystems"?) I haven't done this, although I've read some of the applicable literature...
Thoughts #2 and #3 I'm not really qualified to comment on, so I won't.:)
Slashdot Mirror
Ok, check out Transmeta. The seemed to find a way to use a fraction of the power of an X86 chip, run the same apps, and perform better with a lower clock speed.
If memory serves me, a 700MHz or so Crusoe was roughly equivalent to a 500MHz PII. Should be "performs equivalently with somewhat higher clock speed..."
Uh, I believe the point of these discussions was to illustrate why the Mac has had a small open source community. Comparing them in "the apples/oranges style of idiocy," as you put it, is merely a way of showing how far apart the camps are.
As an aside, I saw mentioned in other places many different variations on NewsWatcher and other open-sourced programs -- Mac open source seems to want to fork before it wants to integrate. Why have all these separate versions, each of which has some added functionality? Why not integrate them into a "official" release? There is either a lack of organization, or a lack of cooperation amongst the authors...
I don't see what your point was comparing the distribution strategies -- clarify, if you can. I agree that both groups do have things to share and need to learn from each other. And indeed, the two groups have, for the most part, evolved separately, but I don't see that as the main reason behind the Mac's small open source community.
(And, BTW, I sense an implication that I am somehow a Linux zealot. Please don't try that -- I develop software for Windows, use Linux as a hobby, and used to write Mac software in academia. I'd like to think that I am, if not totally unbiased, then at least somewhat informed about all three.)
3. Apple didn't want you developing for the Mac unless you werre a commercial licenser.
Nonsense and lies. They did show preferential treatment to commercial houses, though, because they were more likely to invest the time to learn the system and make high-quality software than early hobbyists.
And this is exactly why the Mac has a poor tradition of open source -- commercial houses were given preferential treatment, because Apple didn't believe that hobbyists were worth supporting.
If I'm not mistaken, your first two examples (NewsWatcher and NCSA Telnet) were only open-sourced after the authors found themselves no longer able (or willing) to update the software. This is in marked contrast to what is more commonly thought of as "Open Source", where the code is typically open right from the get-go (or very shortly thereafter).
This, I think, is the fundamental difference between Open Source on the Mac and everywhere else -- source release usually only happens when the original author "gives up." Look at Apple's GameSprockets -- they gave up on them, and opened them up. Ditto on the above packages, and some others that I recall from my Mac days.
Remember, though, that they're using the Windows kernel, not the whole operating system. There should be quite a bit more memory free than would normally be expected under Windows.
"way more frames of animation" == WAAAAAAAAAAYYYYY more memory for sprites.
Each frame is a rendering of the sprite, multiplied by (bits of color / 8) bytes. Suppose you have 500 frames of animation for a 32x32x16 character. That's one megabyte of memory, roughly, for a _tiny_ character. You want 1000 frames of animation? 2 megs of memory. And so on.
All of a sudden, 3D meshes don't look so bad after all.
700 MHz == 700 million cycles/second
But, an instruction can take more than one cycle to execute. It's difficult to predict exact amounts of instructions/second because of this. Soooo...your 7 million number was purely hypothetical, right?
No programming language could handle the case of (literally) "structA contain[ing] a structB and structB contain[ing] a structA" -- it's an infinitely recursive definition.
Now, if you meant "structA contains a pointer to a structB" and vice versa, then the solution would be to do multi-pass compilation. (I think. hehe)
The Age of Empires games were made by Ensemble Studios, not internally at Microsoft.
They aren't as different as you might think. The language parser remains the same, it's just that the code generation portion of the compiler is different.
A compiler takes a set of input (source code) and produces an output (machine code) based on an ordered series of steps.
I dunno, sounds a lot like an algorithm to me...
If I recall, Strassen's algorithm gets you down to something like O(n^2.8something), and then some related variants get you down to O(n^2.7). Not sure about other decompositions. It's been awhile since my algorithms class, and since I don't use these things every day, I tend to forget them quickly. :)
An excellent list of suggestions! Of course, there's not room in the top 10 for all of them, but I think that more than a couple are a match for the ones in their list...
They profit off of licensing games for the system. Even though the devkits are expensive, let's look at the realities of things...would you make more money selling 1000 $20K dev kits, or licensing millions of copies of games every year?
Telemate, man! Telemate! One of the only multithreaded DOS apps I ever saw. And the only DOS comm program that let you do other thing while downloading!
:) (aside from the obvious benefits of uploading AND downloading at the SAME TIME!)
:)
Unless, of course, you were file transferring with HS/Link, in which case you could chat with the sysop...
Of course, eventually I wound up switching over from Procomm Plus in DOS to...Procomm Plus in Windows! Then, I could play Solitaire in Win 3.1 while I downloaded
Man, those were the days!
Indeed, they were...
Why does everyone seem to think that PlayStation 2 programming is the same as programming for the PlayStation?
The PS2 may be evolutionary in some respects, but on the whole it is an entirely new beast.
Why assume that games would need to be installed? The hard drive could be for patches/updates (either bugfixes or game enhancements -- think of all the mods for Quake and Half-Life out there and you'll see why this is exciting for a console), for savegames (obviously), and for caching stuff during gameplay (faster than reading from the CD). Seems like a smart thing to include to me.
The article doesn't say that there's only two machines. (It says something along the lines of "both operating systems were tested on old Pentium machines", not that "there were two machines, one running NT, one running Linux".) It actually doesn't say how many machines were in the study, IIRC.
"Both operating systems running on relatively old Pentium machines" != "Two machines, one running NT, one running Linux"
As far as I could tell from a quick re-scanning of the article, there was no indication of exactly how many machines there were. But they didn't say there were two, they said that the study used "relatively old Pentium machines."
The real damage is not in that you simply couldn't buy the book -- the damage is to their reputation and customer base. You'd probably go to Barnes & Noble if you couldn't get something from Amazon, right? I agree it's difficult to place a monetary value on the damages, but the situation isn't as cut-and-dried as "well I'll just go back once they're no longer DoS'd."
The situation with eBay and E-Trade is even worse, because a denial of service is preventing users from doing business with others, not just the company affected by the DoS.
I base my comments on developing for various versions of DirectX (from DX3 forward) under two different versions of Visual Studio. Sure, you need to convert the project files itself, but there's nothing inherent in DirectX itself that prevents you from using it with a newer/older version of a compiler (within reasonable limits -- i.e. using 16-bit compilers is a no-no), or even a competitor's product. Unlike Java, the C++ language does not change significantly every six months...
If you're going to turn this into a qualifications game, then I'd be curious to hear what your qualifications are. Your personal page on the site doesn't have much in the way of information...
As far as my "apologist" comments went, I suppose that came on a bit strong, but I really think that the situation between Nintendo and Sony is approaching that of the ages-old Apple-PC rivalry. Apple may have better hardware, but systems (and gaming systems in particular) are sold on the basis of software -- both quality and quantity. I don't need to remind anyone who is dominant in the personal computer marketplace (despite Apple's recent about-face).
The GameBoy is 10+ years old IIRC. On the one hand, it's amazing that things are still being developed for it, and that it's still a popular platform. On the other hand, the system is looking extremely long in the tooth...
Another factor to keep in mind is that margins are much more in favor of CD systems as opposed to cartridge-based systems. This is attractive not only to the developers but also the system manufacturer, as it will encourage more people to develop for the system.
P-a-a-leeeease. Both Nintendo and Sony are doing fine. Sony just has bigger sales numbers.
I hate to sound flippant, but "having bigger sales numbers" is pretty damn important to a company's success. Gaming is no longer a niche area, like how it was in the early '80s, where a variety of companies can carve out their own pieces of the market. If you're not in first place, you're endangered.
Nintendo may not be "poor" but they are certainly not in an enviable position right now -- their chief rival, who blew onto the marketplace a few years ago with all the subtlety of a tornado, is now preparing their next salvo. Developers are lining up by the truckload to put out games for the PS2, the hype is being built up (to the level where everyone knows about the PS2), and, most importantly, they're beating Nintendo to market.
I think that Nintendo is in the proverbial lull before the storm. Clear skies now, stormy weather ahead...
I fail to see any relationship between the version of a compiler being used and the version of DirectX being used, so "I'm going take if from your comment" that you don't know much about development.
And as far as DirectX being updated for WinCE, it would seem to be in Microsoft's interests to do so, since the X-Box will certainly use DirectX (and will hopefully not use a full-blown, slow version of Windows)...
I tend to agree with the earlier poster who labeled the article as "Nintendo apologist" -- Nintendo's success is currently based off of the Pokemon franchise (sickeningly so), and their market share has dwindled considerably in comparison to Sony, which came out of nowhere. Remember, the name of the economic game is not whichever system is better, but which system is the most dominant and prevalent in the home. Nintendo's not dead yet, but they better come packin' heat with the Dolphin or they will be DOA...
Let's not forget Henry Ford himself was famous for blatantly having thugs beat senseless employees who wanted to unionize. What do you think they'll want to do if you're emailing to incite sickouts?
:)
Am I the only one who read this as "senseless employees who wanted to unionize," instead of "beating employees senseless"?
Not everyone is obligated to become a computer geek. Just because they aren't doesn't mean they don't have online rights that need to be looked out for.
This has to be the most insightful thing I've read in the discussion so far. We may all laugh at AOL or say "anyone who knows better would say 'no'", but the truth of the matter is that Joe Six-Pack shouldn't have his computer, or, for that matter, other competitors' products, disabled by another piece of software.
Regarding thought #1, the Win32 file system drivers are extensible -- you can plug in additional filesystems if you so desire, although I haven't really seen it done at all. (perhaps some of the esoteric "FTP filesystems"?) I haven't done this, although I've read some of the applicable literature...
:)
Thoughts #2 and #3 I'm not really qualified to comment on, so I won't.