It will undoubedly have impressive hardware - be it the MIPS+ATI graphics core, the Transmeta or PPC CPU. All these things are readily available. What makes Amiga live on is not the original hardware or software, but the vision of design.
It will be difficult for a new system to really wow people when it uses stock hardware. Or, more likely, it will have some wow factor up front, but it will be lost as soon as the next wave of improved stock hardware hits the market.
What's been keeping amazing new makes of computer from breaking into the home market is that the PC makes up for quality of overall design by getting greater horsepower every few months. So if you put something in an Amiga box that can be compared to a current offering--say ATI's RAGE 128 core--then it will start to look dated very quickly--as soon as the next generation of ATI chips shows up for the PC next quarter. Imagine if the audio/visual hardware in the original Amiga kept getting upgraded and improved--for a machine other than the Amiga.
Yes, you could upgrade this new Amiga to use the latest hardware, but then what have you gained? You might as well have bought a PC.
There are some amazing things that could be done with a new machine. How about some of the ultra low power CPUs that are being developed (like Chuck Moore's F21, which will use practically no power and cost only a buck or two to manufacture). Or how about the graphics chip developed at VM Labs? What about graph reduction hardware designed for running functional languages? For a new system to break the PC blahs, it needs to be very forward thinking. Putting PC parts in a new box won't cut it.
Now wait a minute. The entire field of computer science was spawned this century. Think of all the towering giants who made huge contributions:
Alan Turing John Von Neumann Edsger Dijkstra Don Knuth Niklaus Wirth Tony Hoare John McCarthy Frederik Brooks Ivan Sutherland Alan Kay Robin Milner Ken Thompson Jim Blinn ...etc., etc.
Somehow, running Linux is being equated with playing system administrator and mucking about with new window managers, kernel upgrades, and distributions. This is only a byproduct of Linux still being under heavy development and not an intended goal or lifestyle.
Imagine a system with the UI polish and ease of use of a Macintosh. The only difference--and 98% of users wouldn't even know this--is that there's an xterm-like application that you can fire up. From it, you can run gcc, use Perl, and do whatever you please. (Please, nobody needs to point out MacOS X.)
This is a nightmare scenario for many Linux "users," but I have no idea why.
About the article: There's this stereotype of programmers lacking any aesthetic sense; one that has been made worse by much software in the Linux world. A programmer without any real-world aesthetics is next to worthless, IMO.
About some followups: The attitude of trying to keep things difficult to use because its k00l in an "I thought The Matrix was the best movie ever made" sorta way is something that needs to go. There's a reason that clue-ridden insiders like Jamie Zawinski have been hard on Linux. Even Linus has smacked down Emacs on Usenet. What's the purpose of a "let's make it worse" attitude?
Good article. Sometimes I hate to admit that a majority of games art art-centric, but he's right.
Even for Windows, there's a severe lack of good, reasonably-priced game engines available. The next easiest thing to starting up a phony game company that never releases a product is to write a game programming library ("Those who can't, teach:)"). The old-fashioned 2D sprite packages available for Linux, are pretty horrific, for example, and not written by people who've ever worked on a game more complex than Tetris. Coming up with something general and of good quality is surprisingly difficult.
I also think the author underestimates the amount of programming that goes into a game. If it were always as easy as Myst, there wouldn't be 5 to 12 programmers working on every commercial game.
Criticizing a vapor project isn't a knock against Linux or OSS; it's just that wannabe game groups have been popping up at an amazing rate lately, mostly in the Windows world. The general pattern is:
1. Come up with a hip, alterna-skater name and a harsh, futuristic logo. 2. Give titles to your friends ("design director", "beverage coordinator"). 3. Put together an honestly impressive, graphics-laden web site. 4. Outline an ambitious game project: "A game with Unreal-style graphics set in the world of _Snow Crash_." 5. Write the background story for the game. 6. Announce your flagship title to the world, also mentioning that while you currently have a great 2D artist, musician, and team manager, you still need 3 or 4 experienced coders, plus someone to do the 3D modelling and artwork.
No offense to anyone involved--it certainly *is* an order of magnitude easier to run a website and do PR than to write a game--but this project is pretty much going by the numbers:( Sigh.
It reminds me of what the personal computer (apple, ibm pc) did to the world of mainframes. Perhaps some of you are too young to remember having to give your programs to the guy behind the window and have him run your job at his convenience. As soon as I could afford a personal computer I bought one. Then I was in control of my destiny on my computer.
Let's not be over the top here. Timesharing systems with one terminal per user definitely came between batch-only mainframes and home computers.
It's the classic mistake that all authors of game programming books make: they spend 75% of their time talking about C compilers, Windows programming, specific interfaces to joysticks and graphics and sound, and overly large utility classes. The author should have included a library to handle the system interfacing, briefly talked about it in one chapter, put the source code on the CD, and then focused on actual game design and programming issues. The end result would have been 100-150 pages of solid reading.
I suppose a 700 page tome is considered more impressive.
I find the author's anti-RedHat sentiment to be peculiar. The RedHat distribution hardly presents users with a "shell" that keeps them from ever interacting with Linux proper. Is he referring to RPM? Probably not, as other distributions have similar concepts (and many of them use RPM as well). RedHat may be popular, and some people resent that, but even RH 5.2 is still a raw presentation of basic Linux. It doesn't boot up into KDE as many people think (heck, it doesn't even install KDE by default).
Except, of course, that the Linux kenel and almost all applications and utilities--everything from gzip to bash to KDE--are compiled with gcc or egcs. Like it or not, gcc is integral to Linux and has been from the beginning.
Slashdot Mirror
It will be difficult for a new system to really wow people when it uses stock hardware. Or, more likely, it will have some wow factor up front, but it will be lost as soon as the next wave of improved stock hardware hits the market.
What's been keeping amazing new makes of computer from breaking into the home market is that the PC makes up for quality of overall design by getting greater horsepower every few months. So if you put something in an Amiga box that can be compared to a current offering--say ATI's RAGE 128 core--then it will start to look dated very quickly--as soon as the next generation of ATI chips shows up for the PC next quarter. Imagine if the audio/visual hardware in the original Amiga kept getting upgraded and improved--for a machine other than the Amiga.
Yes, you could upgrade this new Amiga to use the latest hardware, but then what have you gained? You might as well have bought a PC.
There are some amazing things that could be done with a new machine. How about some of the ultra low power CPUs that are being developed (like Chuck Moore's F21, which will use practically no power and cost only a buck or two to manufacture). Or how about the graphics chip developed at VM Labs? What about graph reduction hardware designed for running functional languages? For a new system to break the PC blahs, it needs to be very forward thinking. Putting PC parts in a new box won't cut it.
Now wait a minute. The entire field of computer science was spawned this century. Think of all the towering giants who made huge contributions:
Alan Turing
John Von Neumann
Edsger Dijkstra
Don Knuth
Niklaus Wirth
Tony Hoare
John McCarthy
Frederik Brooks
Ivan Sutherland
Alan Kay
Robin Milner
Ken Thompson
Jim Blinn
...etc., etc.
Ugh. What a depressing post!
Somehow, running Linux is being equated with playing system administrator and mucking about with new window managers, kernel upgrades, and distributions. This is only a byproduct of Linux still being under heavy development and not an intended goal or lifestyle.
Imagine a system with the UI polish and ease of use of a Macintosh. The only difference--and 98% of users wouldn't even know this--is that there's an xterm-like application that you can fire up. From it, you can run gcc, use Perl, and do whatever you please. (Please, nobody needs to point out MacOS X.)
This is a nightmare scenario for many Linux "users," but I have no idea why.
About the article: There's this stereotype of programmers lacking any aesthetic sense; one that has been made worse by much software in the Linux world. A programmer without any real-world aesthetics is next to worthless, IMO.
About some followups: The attitude of trying to keep things difficult to use because its k00l in an "I thought The Matrix was the best movie ever made" sorta way is something that needs to go. There's a reason that clue-ridden insiders like Jamie Zawinski have been hard on Linux. Even Linus has smacked down Emacs on Usenet. What's the purpose of a "let's make it worse" attitude?
Good article. Sometimes I hate to admit that a majority of games art art-centric, but he's right.
:)"). The old-fashioned 2D sprite packages available for Linux, are pretty horrific, for example, and not written by people who've ever worked on a game more complex than Tetris. Coming up with something general and of good quality is surprisingly difficult.
Even for Windows, there's a severe lack of good, reasonably-priced game engines available. The next easiest thing to starting up a phony game company that never releases a product is to write a game programming library ("Those who can't, teach
I also think the author underestimates the amount of programming that goes into a game. If it were always as easy as Myst, there wouldn't be 5 to 12 programmers working on every commercial game.
Criticizing a vapor project isn't a knock against Linux or OSS; it's just that wannabe game groups have been popping up at an amazing rate lately, mostly in the Windows world. The general pattern is:
:( Sigh.
1. Come up with a hip, alterna-skater name and a harsh, futuristic logo.
2. Give titles to your friends ("design director", "beverage coordinator").
3. Put together an honestly impressive, graphics-laden web site.
4. Outline an ambitious game project: "A game with Unreal-style graphics set in the world of _Snow Crash_."
5. Write the background story for the game.
6. Announce your flagship title to the world, also mentioning that while you currently have a great 2D artist, musician, and team manager, you still need 3 or 4 experienced coders, plus someone to do the 3D modelling and artwork.
No offense to anyone involved--it certainly *is* an order of magnitude easier to run a website and do PR than to write a game--but this project is pretty much going by the numbers
It reminds me of what the personal computer (apple, ibm pc) did to the world of mainframes. Perhaps some of you are too young to remember having to give your programs to the guy behind the window and have him run your job at his convenience. As soon as I could afford a personal computer I bought one. Then I was in control of my destiny on my computer.
Let's not be over the top here. Timesharing systems with one terminal per user definitely came between batch-only mainframes and home computers.
It's the classic mistake that all authors of game programming books make: they spend 75% of their time talking about C compilers, Windows programming, specific interfaces to joysticks and graphics and sound, and overly large utility classes. The author should have included a library to handle the system interfacing, briefly talked about it in one chapter, put the source code on the CD, and then focused on actual game design and programming issues. The end result would have been 100-150 pages of solid reading.
I suppose a 700 page tome is considered more impressive.
I find the author's anti-RedHat sentiment to be peculiar. The RedHat distribution hardly presents users with a "shell" that keeps them from ever interacting with Linux proper. Is he referring to RPM? Probably not, as other distributions have similar concepts (and many of them use RPM as well). RedHat may be popular, and some people resent that, but even RH 5.2 is still a raw presentation of basic Linux. It doesn't boot up into KDE as many people think (heck, it doesn't even install KDE by default).
Except, of course, that the Linux kenel and almost all applications and utilities--everything from gzip to bash to KDE--are compiled with gcc or egcs. Like it or not, gcc is integral to Linux and has been from the beginning.