As someone working in the PC game industry, here's how the last five or more years have gone.
From 1994-1996, we were getting lots of experience with software rendering. It was slow, it was annoying, but we were getting good at it. Then the Voodoo 1 became popular, and we had more power than we knew what to do with. So we started writing games specifcially for that chipset, and we just started scratching the surface of what it could do. Then all these other chipsets started getting marketshare, those from Nvidia, those from 3Dlabs, those from Rendition. So we had to give up tuning our code for Glide and got into the whole OpenGL/Direct3D API and driver nonsense that we still haven't gotten out of. Performance was watered down as a result.
Then the Voodoo 2 hit in early 1998. We had to focus on a variety of cards and APIs by then, so we couldn't take full advantage of what it had to offer. If you consider that the coin-op San Francisco Rush was powered by a Voodoo 1, and it surpassed just about everything that was released for the Voodoo 2, you'll get a feeling for how things were. But the Voodoo 2 was cool, so we pushed back into Glide. Then the TNT was announced and gave similar performance. It might have been better on the benchmarks, but our 3D rendering code was getting pretty spread out and API-heavy at that point. We didn't push either card to its limit.
The 3DNow CPU extensions were released, but we ignored them because we had too much else to worry about.
Intel announced their so-called "Katmai" extensions for vector math, which ended up in the Pentium III. Honestly, we didn't have time to listen. Cards and drivers were changing so fast that we didn't want to blow our test suite that wide open. From what I hear, most developers ignored them too. Driver writers didn't, but driver writers were getting a "who cares, when a new card will come out in three months" attitude, so the support was poor to spotty.
Then more cards, like the GeForce, the Savage, and some from Matrox. All good cards, but only the Matrox supported hardware bump-mapping, the Voodoos still had the best multitexturing support, and the GeForce was the only one with transformation and lighting. We started ignoring most of this stuff, just shooting for decent performance on mainstream cards. Nobody really knew we weren't getting near the performance we could, because fan boys were going crazy and buying new cards as soon as they hit the streets.
Now, in 2000, we have the unfortunate situation of the Pentium III and Athlon being standard in corporate desktop machines, though the whole gig with those chips was the SIMD floating point meant for games, the support that very few game developers bothered with. AGP added more cost and complexity to motherboards, though again that was a gamer thing. And those same machines have 3D accelerators in them, soaking up more power, though once again there's no need. The gamers are happy, but of course they don't want to know that we could probably get better performance out of a Voodoo 2 than a GeForce 2, if we were able to concentrate on that chipset alone.
The end result is a complete mess that nobody's happy with. But what can we do?
There have been some bad trends in terms of power consumption for computers recently. The first two are less significant than the third, but they'll all notable when multiplied by 100 million:
1. CPUs requiring 30+ watts of power (50+ for an Athlon), so much that they need their own fans. The geeks that *think* they need 1GHz so they can get 350fps in Quake 3 are a blip on the radar, but realize that these kind of machines will shortly be common in insurance offices and secretary's desktops.
2. High-powered 3D graphics cards being standard in all machines, even though 90% of those machines are only used for word processing and web browsing. Most of the recent chipsets are hot to the touch, even with heat sinks, and many are starting to ship with their own fans.
3. The design of window managers and application software that provide incentive to purchase larger monitors. Having a big monitor is a geek goal, but it's getting more and more uncomfortable to use most window managers on reasonably sized monitors (say 14" or 15"). If you think about it, you're only focused on one application at a time, and the rest of your screen goes to other applications that you're not using at the moment. Windows managers could use a good rethinking here, as it's starting to seem silly to spend your days word processing inside of a window that only takes 50% of the screen. Sure, you could maximize it, but then you just have giant margins. I would be happy with a small monitor, if applications and desktop environments were designed for small monitors, and not huge ones. This is similar to the "gotta get me a massive SUV" trend. People think bigger is so cool, until gas prices get near $2.00 a gallon.
Interestingly, those are two of the best loved pins of all time. When Addam's Family was released, it captured the hearts of most regulars in rec.games.pinball.
Not the first time this has been forseen
on
Is Pinball Dying?
·
· Score: 2
Pinball dies every decade. In the late 1970s, when video games really took off, the end was predicted for pinball, but it hung on. In the late 80s it looked bad once again, only to be followed by a wave of some of the greatest pins ever: Black Knight 2000, Terminator 2, Bride of Pinbot, FunHouse, Fish Tales, Addam's Family, etc. Of course this time around things look worse than ever, because Williams is out of the pinball business.
The trouble with Windows is that it's generally unreliable and somewhat outdated under the hood. Rebooting every few days, if you can last that long, is still a good rule of thumb. Windows NT is much better, but too clunky for home use.
Linux is spot on technically, but crusty from most users' points of view. Even respected uber-programmers, like Jamie Zawinski and Rob Pike, don't see all being rosy with Linux.
MacOS X looks pretty close to what people have come to expect from the Mac. But it's also BSD under the hood. So the wife and grandma can get along with it just fine, and geeks can just grab a shell and keep it open all the time. This is the best of both worlds. Perhaps the most exciting part of it, from a UNIX lovers point of view, is that UNIX is going to, for the first time ever, be a mainstream consumer operating system.
You can laugh about marketrdroids and how stupid it is that people wear Tommy H. T-shirts, but in general marketing *works*. There are thousands of really good bands out there who sell a couple of CDs every time they play at a local club or coffee house. In general, nobody seeks out these bands. People buy CDs from Moby and whoever else is on the end rack at Borders. Yeah, Moby is good, rock on and all that, but much of the success comes from going through a corporation and having a team of marketdroids come up with ad campaigns for Rolling Stone. If he were an independent who refused to go through major labels, he would be another local hero and not a worldwide phenomenon (examples of people who tried to eschew corporate record labels: Dave Alvin, Sarah Hickman).
If you want to rail against marketing, then you have to avoid mainstream movies and music, and not just Windows, because it's the same sort of thing. Someone is choosing what you have access to and you're buying it. A typical Slashdot attitude is "Oooh, look how underground I am because I like The Simpsons and X-Files and read Neal Stephenson." That's exactly the same as wearing Tommy T-shirts, but for geeks.
Those Linux zealots brought you Linux in the first place.
Not true. Linux zealots whine and complain about other operating systems and rave about how great Linux is. They don't do anything significant. The people making the big contributions tend to be more secure and much lower key.
And I think I'd want both options to use when needed. So forcing everyone to use one or the other is a monumentally stupid move. IMHO.
My goodnesss. One mention that Delphi might be the tool for some purposes and look at all the defensiveness. Why? If you re-read this thread, no one is saying that IDEs are bad across the board. In fact no one's talking about the Delphi IDE at all. They're talking about the GUI builder and compiler.
I will continue to use Emacs as my IDE with g++ and GNU make. I'm not being elitist---these are the tools which make my job the easiest and fastest to do.
You're missing the point. If you were to write a very GUI-heavy program for KDE or Gnome, would you still use straight g++ and something like GTK? Or would you want to use a very slick package that integrates everything like Delphi?*That's* the kind of thing that we're talking about here, not the Delphi IDE vs. Emacs.
I could write a version of gcc with essentially zero compile time. The problem is the simple implimentation would produce slow code. Gcc does a okay job of optimization (depends on the platform), and syntax checking. My version would compile minnimally correct code, with no concern for where warnings should be issued.
You're missing the point. Delphi does optimize code, and it does provide very nice warnings and errors (and even hints about things that could possibly be errors, though you can turn this off). You also get range checking for array accesses, if you want it.
Yes, it is also true that Object Pascal is a simpler language than C++. It is easier to parse, and it doesn't include lots of the bloat. But it does have really nice modern features, like exception handling, unlimited length reference-counted strings built into the language, a very nice module system, and so on. Similarly, you could say that Linux has a leaner and cleaner kernel than Windows. Is that a bad thing?
The bottom line here is that Delphi is very well-engineered. You can compile and link entire applications in the time that it takes gcc to compile one file. You can justify it any way you want to, but it is still (A) true, (B) wonderful, and (C) available right now for Windows ($99 for the personal edition).
I'm not saying you should never use C or gcc, not at all. Just that it would be a mistake to overlook such a nice tool simply because you're afraid it would be better than what you're used to.
Looking at games with a literary eye is on par with teaching a college course in Saturday morning cartoons. For the longest time, there has been the notion that video and computer games will advance into a form of art. But after years of this, games are a weird juvenile form of entertainment. There's been a constant notion of a certain type of games being for the kiddies, and then a "mature" type of games for people who are beyond that. "Mature" seems to be equated with top-heavy bikini babes, an obsession with blood and gore and weapons, and a fixation on dark futures. In short, fifteen year olds who want to separate themselves from the happy-go-lucky days of their youth.
At the same time, game design creativity has stagnated in a horrible way. Authors of fiction create worlds and tell stories and the results end up in bookstores. Typical authors don't start off a project by saying "Okay, that last Stephen King novel sold really well. I'm going to write the same book, only better." Yet this is what game developers always do. A game design starts out with "like Everquest, but..." or "Quake with a fantasy RPG element..." and we get the same old stuff. Yet we have no subversive element, just people writing more versions of old arcade games (but now they're Open Source).
Rather than discussing the current crop of games in an adult way, perhaps a better approach would be to try to foster a generation of game creators who can think for themselves and want to distance themselves from what's expected of the so-called game "industry." After all, it's common for writers and musicians to start out in a subversive way and grow into mainstream: Kerouac, REM, Hunter S. Thompson, Smashing Pumpkins.
This story, about Borland's upcoming Delphi for Linux, has drawn a very clear line in the sand between introverted Linux zealots and the people who are going to move Linux forward. I see endless responses like "Bah! I can just use gcc!" and "Only fools would use a RAD environment! Real Linux programmers will keep using gcc and an Emacs-based IDE."
A fantastic--maybe the best--compiler and GUI development system is coming to Linux and we should ignore it? We should keep flogging ourselves until we remember that UNIX is The Only Way? What is happening here is that an elitist crowd with a confused agenda is suddenly being confronted with the painful truth that Linux development software is far, far behind what has been available for some other systems. This realization needs to be embraced, or we can never advance. We can only fool ourselves for so long.
And, you know, this may make Linux difficult to differentiate from Windows. But hasn't that been the plan all along? I mean, we could have been devoting energies to desktop environments that aren't re-workings of Windows, right?
Actual, don't knock me off as a troll. I'm trying to say, why should I bother with Borland when I can use gcc/g++ and any ol' IDE (sorry about that obscure pun) that I want. I just don 't make sense. I already got dem tools But you for you windows boys, yes now you too can write programs for Linux!
Don't knock Delphi until you have tried it and written a app with it. It's amazing. The compiler is maybe the best compiler ever written. Compile time is effectively zero. There's no link time. Object Pascal has niceties, like having single file modules instead of source + header files, that make you wonder how you ever managed with C++. You can ignore all the GUI and OOP stuff and just use Object Pascal as a straight compiler. The built-in assembler is so nice I've even used it for all-assembly programs, as it's much faster than NASM/MASM.
And all of this is really just used as the back-end for a very slick and complete GUI design package. You can click on controls and add event handlers for just about anything. It's very complete.
The killer is that everything is extremely solid and well-implemented to a level that is foreign to Linux desktop environments. gcc may be nice, but we've all had our general oddities and bizarro error messages from it. Putting a cranky GUI layout package on top of this is not the same thing. The bottom line is that Delphi is much, much stabler and slicker than what we're used to using in Linux. Ignoring it for GUI-based programs is on par with writing an OS in assembly code instead of C. Don't wear a hair shirt because this comes from Windows. Delphi will change Linux application development, no question about it.
But there's a compiler underneath. The Object Pascal compiler--part of Delphi--is an amazing piece of work. Compile time is zero on just about any machine. That is a huge, huge reason to choose Pascal over C++ for this sort of thing.
I see much misinterpretation of the term "systems programming." It has become a euphemism for low-level bit tiwddling and device driver hacking. In fact, the term is quite literal, meaning "the programming of complete systems." If you look back at the original UNIX, it was an OS, a user-environment, and a programming language, all of which were developed to work together.
A goal of systems programming is the usability of the entire package. This is one of the stumbling blocks for what Linux has grown into. The basic command line environment is comparable to the original UNIX in this regard. Cryptic, but powerful for a certain type of user and application. But that's where the plan stops. GNU tools tend to be messy critters, with way too many command line options. Look at the man page for ls, you'll see a note about how the man page is no longer maintained (use texinfo instead). X Windows feels like a giant hack to bring UNIX into the bitmapped age. The desktop environments for Linux all feel like attempts to bring Microsoft Windows into the realm of free sofware, but they're duplicating the same mistakes: too many gadgets, too many meaningless icons, too much focus on fiddling, too many questionable interface descisions. In the end, we have an interface that's not sure what it wants to be, outside of a slap at Microsoft's market share, put on top of a byzantine graphics system that is gernerally unliked, on top of of mostly unrelated early 1970s era nuts and bolts. It is free, and this is good, but it can be viewed as a way of keeping programmed running in circles until something truly new comes along and makes everyone re-assess their goals.
It is also worth considering that the era of big operating systems is coming to a close. Who could have believed that a Palm, with it's little 68K chip, would be so bloody useful? Or that the Game Boy is closing in on 100,000,000 units (that's 100 million)? You could argue that web browsing and word processing and coding are outside the realm of such toys, but is that really true? There's a cool little Forth for the Pilot, that's probably the slickest programming system I've seen in years. And there's going to be cool stuff in the future from other non-desktop hardware. Is trying to tack Windows onto Linux really such a worthwhile endeavor?
Right now, there are three primary groups using Linux:
1. ISP's and other businesses that need rock solid networking and file serving. 2.College students, because there's some benefit to using Linux if you're a comp sci major, and also because it's free (either "free as in beer" because students are generally cash short, or "free as in freedom" because it's easier to be idealistic when you're a student). High school students are included here, too, though maybe to a lesser degree. 3. Geeks who fixate on which operating system they use. This overlaps somewhat with the previous item.
Right now, Loki is selling mostly to number 3, and I suspect this is the smallest and most volatile group. Number 2 is where the users are, but that's a tough place to make money. It's the same place crazy Napster support is coming from, and for a very obvious reason.
To an outsider whose eyes are clear of zealotry, the Linux game market looks like this:
1. There are only a handful of commercial games. 2. Those games are also available for Windows, so even if I decide to pick up Linux at Barnes & Noble I still have Windows around to play games with. Not sure why I would want to play the Linux version. 3. The freebie games are pretty horrendously unexciting. Yay! Hundreds of Tetris and Asteroids variants! Does Open Source somehow imply a lack of creativity? 4. All the nifty games like The Sims and FreeSpace 2 and Roller Coaster Tycoon aren't out for Linux. Maybe they'll get ported next year, but I don't see why I should paint myself into a corner just for the sake of supporting an operating system.
Harsh? Maybe, but let's be honest about this. There are very few reasons to even consider using Linux for games, unless you've decided to hold fast to OS-centric views. If there were something really sweet that were available only for Linux, I could see _some_ people peering over the wall to see what all the fuss was about, and maybe setting up a partition to play. But we're a long way from seeing that happen.
As processor power increases, the difference between good and poorly written software is becoming more dramatic. This is a greater issue that's being dwarfed by pissing games over 1GHz+ speeds. For example, twenty years ago it was common for developers to use VAX-level minicomputers to do development for emergine home computers (Atari, Apple II, early PCs). After all, who could write a decent assembler on a 1MHz system with hardly any RAM? But it turned out that native assemblers on teeny tiny systems were frequently outperforming the minicomputers by a factor of ten or more.
Fast forwarding to today, consider any C++ compiler, say gcc or Visual C++. You can never have enough processor power for such a compiler. Even on top of the line machines you're still talking minutes to rebuild a medium sized project, and seconds for an average link. Now fire up Borland's Object Pascal compiler (buried inside of a RAD tool called Delphi). On a 200MHz machine the compilation time is up near a half million lines per minute. Link time is effectively zero. In general, compilation time doesn't exist. You have to have a pretty big project before you even notice that pressing F9 is taking any time at all. If you're using a 300MHz machine or faster, this is never even an issue. It's never an issue for almost any slower machine either, but I'm playing it safe. Now, yes, Object Pascal is simpler than C++. The compiler maybe doesn't do some of the nutty stuff that gcc does. But in the end, does it matter? A compile time of zero sure does make it easy to go in there and twiddle around with the code, making it go real fast.
This kind of thing is going to be more and more common. Is a bulky application slow because the processor isn't fast enough or because it is bulky? Throwing more processor power at problems like this is a dodge.
Your X server is what controls your video card and mouse. That's all it does; X does not have menus, windows, or any of that stuff, because of the way it was designed.
And this is why it's shocking that X is as bulky as it is. It does very little, yet sure requires obscene overhead in order to accomplish it. Isn't this one of the shortcomings of Windows?
One of the very first self-published books I ran across on the web was Halcyon Days: Interviews with Classic Computer and Video Game Programmers. It was released back in 1997. I remember reading about it in a number of places, including Wired. It might be worthwhile to talk to the author and see how it turned out.
You are making some common mistakes here. The first is that cutting out the middleman is a good thing. Ask most any shareware author, or someone in a local band, and find out how much money there really is in getting people to buy something from an unknown. Getting something distributed outside of your own meager means is a huge, huge boon.
Second, there is a reason that publishers reject manuscripts. Sometimes they are wrong. Zen and the Art of Motorcycle Maintenance being a famous example of a book that was rejected dozens of times before becoming a bestseller. But often they are right. Just because you think something is great and worth reading doesn't mean that it is. You are asking web surfers to be your editor, and to read your story with the foreknowledge that it may be downright lousy. This can work, if you are very good at what you do, just as many bands act as their own producers, but most people need someone else in the loop, especially early on.
Third, you need the experience of going through a true editing and publishing process in order to gain experience. Playing music in your bedroom is completely different than trying to put a CD together. In effect, you're just going to put a tape recorder in your bedroom and give the result to the world.
Finally, the web is becoming very full of garbage and this is making people cynical. There are endless home pages of poetry from high school girls. There are sites for pseudo-companies, complete with faux press releases and corporate titles, put together by some junior high kids who decide to put together a software developer. A site with some guy hawking his Great American Unpublished Novel isn't going to garner much confidence or attention.
Attempting to get published is still the way to go.
The two major candidates in the next presidential election are viewed as cardboard cutouts thrust forward by their respective parties. Neither is a great statesman, a peacemaker, or a highly respected man of wisdom. They're not horribly offensive, and they have business-as-usual middle of the road views, if those can be considered what it takes to become president of the United States. Yes, there are other candidates involved, but they're not going to get more than an insignificant handful of the popular vote.
Two questions: 1. Aside from "rah, rah, democracy is great, rock the vote" nonsense, please justify why voting in the upcoming presidential election is worthwhile. 2. Having two bumbling, unqualified candidates makes US-style democracy laughable in the eyes of other countries. Does this hurt democracy in general?
I really, really hate projects which are "open source" but who refuse to release the source until it's "done." Too many projects these days seem to be following that path, and it's a dangerous one to take. Because what if the code is never truly "finished", as no project ever really is. Its sad.
It is not sad. You're simply parroting Linus Torvalds's "release early and often" advice. This works for an OS, because compatibility problems are the issue of the day. There are also reasons not to do so:
Most people don't take the time to upgrade, so they'll miss out on major features that are added later.
If you release something that is incomplete, people will try it, see that it is incomplete, and have a poor impression from that point on.
If promising software is released at an early stage, there is likely to be much more "cloning activity," channeling effort into doing things again--the right way--instead of tweaking something that's already most of the way there
It's like that in all the major high-tech industry areas, at least in the united states, prolly the world eh?
No. Silicon Valley is bad. Seattle is bad, but not so much as the first. Dallas is fairly inexpensive, and it's chock full of high-tech companies (lots of telecom, for example). If you compare house prices on the web you can get a good feel for the general priciness of an area.
Slashdot Mirror
As someone working in the PC game industry, here's how the last five or more years have gone.
From 1994-1996, we were getting lots of experience with software rendering. It was slow, it was annoying, but we were getting good at it. Then the Voodoo 1 became popular, and we had more power than we knew what to do with. So we started writing games specifcially for that chipset, and we just started scratching the surface of what it could do. Then all these other chipsets started getting marketshare, those from Nvidia, those from 3Dlabs, those from Rendition. So we had to give up tuning our code for Glide and got into the whole OpenGL/Direct3D API and driver nonsense that we still haven't gotten out of. Performance was watered down as a result.
Then the Voodoo 2 hit in early 1998. We had to focus on a variety of cards and APIs by then, so we couldn't take full advantage of what it had to offer. If you consider that the coin-op San Francisco Rush was powered by a Voodoo 1, and it surpassed just about everything that was released for the Voodoo 2, you'll get a feeling for how things were. But the Voodoo 2 was cool, so we pushed back into Glide. Then the TNT was announced and gave similar performance. It might have been better on the benchmarks, but our 3D rendering code was getting pretty spread out and API-heavy at that point. We didn't push either card to its limit.
The 3DNow CPU extensions were released, but we ignored them because we had too much else to worry about.
Intel announced their so-called "Katmai" extensions for vector math, which ended up in the Pentium III. Honestly, we didn't have time to listen. Cards and drivers were changing so fast that we didn't want to blow our test suite that wide open. From what I hear, most developers ignored them too. Driver writers didn't, but driver writers were getting a "who cares, when a new card will come out in three months" attitude, so the support was poor to spotty.
Then more cards, like the GeForce, the Savage, and some from Matrox. All good cards, but only the Matrox supported hardware bump-mapping, the Voodoos still had the best multitexturing support, and the GeForce was the only one with transformation and lighting. We started ignoring most of this stuff, just shooting for decent performance on mainstream cards. Nobody really knew we weren't getting near the performance we could, because fan boys were going crazy and buying new cards as soon as they hit the streets.
Now, in 2000, we have the unfortunate situation of the Pentium III and Athlon being standard in corporate desktop machines, though the whole gig with those chips was the SIMD floating point meant for games, the support that very few game developers bothered with. AGP added more cost and complexity to motherboards, though again that was a gamer thing. And those same machines have 3D accelerators in them, soaking up more power, though once again there's no need. The gamers are happy, but of course they don't want to know that we could probably get better performance out of a Voodoo 2 than a GeForce 2, if we were able to concentrate on that chipset alone.
The end result is a complete mess that nobody's happy with. But what can we do?
There have been some bad trends in terms of power consumption for computers recently. The first two are less significant than the third, but they'll all notable when multiplied by 100 million:
1. CPUs requiring 30+ watts of power (50+ for an Athlon), so much that they need their own fans. The geeks that *think* they need 1GHz so they can get 350fps in Quake 3 are a blip on the radar, but realize that these kind of machines will shortly be common in insurance offices and secretary's desktops.
2. High-powered 3D graphics cards being standard in all machines, even though 90% of those machines are only used for word processing and web browsing. Most of the recent chipsets are hot to the touch, even with heat sinks, and many are starting to ship with their own fans.
3. The design of window managers and application software that provide incentive to purchase larger monitors. Having a big monitor is a geek goal, but it's getting more and more uncomfortable to use most window managers on reasonably sized monitors (say 14" or 15"). If you think about it, you're only focused on one application at a time, and the rest of your screen goes to other applications that you're not using at the moment. Windows managers could use a good rethinking here, as it's starting to seem silly to spend your days word processing inside of a window that only takes 50% of the screen. Sure, you could maximize it, but then you just have giant margins. I would be happy with a small monitor, if applications and desktop environments were designed for small monitors, and not huge ones. This is similar to the "gotta get me a massive SUV" trend. People think bigger is so cool, until gas prices get near $2.00 a gallon.
Think about it: Ten years ago, what would you think of an operating system created by some nobody hacker in Finland?
To be more accurate, ten years ago, UNIX--then twenty years old--was on everyone's "out" list.
In 2010 maybe we'll all be using Commodore 64's again?
Terminator
Adams Family
Interestingly, those are two of the best loved pins of all time. When Addam's Family was released, it captured the hearts of most regulars in rec.games.pinball.
Pinball dies every decade. In the late 1970s, when video games really took off, the end was predicted for pinball, but it hung on. In the late 80s it looked bad once again, only to be followed by a wave of some of the greatest pins ever: Black Knight 2000, Terminator 2, Bride of Pinbot, FunHouse, Fish Tales, Addam's Family, etc. Of course this time around things look worse than ever, because Williams is out of the pinball business.
The trouble with Windows is that it's generally unreliable and somewhat outdated under the hood. Rebooting every few days, if you can last that long, is still a good rule of thumb. Windows NT is much better, but too clunky for home use.
Linux is spot on technically, but crusty from most users' points of view. Even respected uber-programmers, like Jamie Zawinski and Rob Pike, don't see all being rosy with Linux.
MacOS X looks pretty close to what people have come to expect from the Mac. But it's also BSD under the hood. So the wife and grandma can get along with it just fine, and geeks can just grab a shell and keep it open all the time. This is the best of both worlds. Perhaps the most exciting part of it, from a UNIX lovers point of view, is that UNIX is going to, for the first time ever, be a mainstream consumer operating system.
You can laugh about marketrdroids and how stupid it is that people wear Tommy H. T-shirts, but in general marketing *works*. There are thousands of really good bands out there who sell a couple of CDs every time they play at a local club or coffee house. In general, nobody seeks out these bands. People buy CDs from Moby and whoever else is on the end rack at Borders. Yeah, Moby is good, rock on and all that, but much of the success comes from going through a corporation and having a team of marketdroids come up with ad campaigns for Rolling Stone. If he were an independent who refused to go through major labels, he would be another local hero and not a worldwide phenomenon (examples of people who tried to eschew corporate record labels: Dave Alvin, Sarah Hickman).
If you want to rail against marketing, then you have to avoid mainstream movies and music, and not just Windows, because it's the same sort of thing. Someone is choosing what you have access to and you're buying it. A typical Slashdot attitude is "Oooh, look how underground I am because I like The Simpsons and X-Files and read Neal Stephenson." That's exactly the same as wearing Tommy T-shirts, but for geeks.
Those Linux zealots brought you Linux in the first place.
Not true. Linux zealots whine and complain about other operating systems and rave about how great Linux is. They don't do anything significant. The people making the big contributions tend to be more secure and much lower key.
And I think I'd want both options to use when needed. So forcing everyone to use one or the other is a monumentally stupid move. IMHO.
My goodnesss. One mention that Delphi might be the tool for some purposes and look at all the defensiveness. Why? If you re-read this thread, no one is saying that IDEs are bad across the board. In fact no one's talking about the Delphi IDE at all. They're talking about the GUI builder and compiler.
I will continue to use Emacs as my IDE with g++ and GNU make. I'm not being elitist---these are the tools which make my job the easiest and fastest to do.
You're missing the point. If you were to write a very GUI-heavy program for KDE or Gnome, would you still use straight g++ and something like GTK? Or would you want to use a very slick package that integrates everything like Delphi?*That's* the kind of thing that we're talking about here, not the Delphi IDE vs. Emacs.
I could write a version of gcc with essentially zero compile time. The problem is the simple implimentation would produce slow code. Gcc does a okay job of optimization (depends on the platform), and syntax checking. My version would compile minnimally correct code, with no concern for where warnings should be issued.
You're missing the point. Delphi does optimize code, and it does provide very nice warnings and errors (and even hints about things that could possibly be errors, though you can turn this off). You also get range checking for array accesses, if you want it.
Yes, it is also true that Object Pascal is a simpler language than C++. It is easier to parse, and it doesn't include lots of the bloat. But it does have really nice modern features, like exception handling, unlimited length reference-counted strings built into the language, a very nice module system, and so on. Similarly, you could say that Linux has a leaner and cleaner kernel than Windows. Is that a bad thing?
The bottom line here is that Delphi is very well-engineered. You can compile and link entire applications in the time that it takes gcc to compile one file. You can justify it any way you want to, but it is still (A) true, (B) wonderful, and (C) available right now for Windows ($99 for the personal edition).
I'm not saying you should never use C or gcc, not at all. Just that it would be a mistake to overlook such a nice tool simply because you're afraid it would be better than what you're used to.
Looking at games with a literary eye is on par with teaching a college course in Saturday morning cartoons. For the longest time, there has been the notion that video and computer games will advance into a form of art. But after years of this, games are a weird juvenile form of entertainment. There's been a constant notion of a certain type of games being for the kiddies, and then a "mature" type of games for people who are beyond that. "Mature" seems to be equated with top-heavy bikini babes, an obsession with blood and gore and weapons, and a fixation on dark futures. In short, fifteen year olds who want to separate themselves from the happy-go-lucky days of their youth.
At the same time, game design creativity has stagnated in a horrible way. Authors of fiction create worlds and tell stories and the results end up in bookstores. Typical authors don't start off a project by saying "Okay, that last Stephen King novel sold really well. I'm going to write the same book, only better." Yet this is what game developers always do. A game design starts out with "like Everquest, but..." or "Quake with a fantasy RPG element..." and we get the same old stuff. Yet we have no subversive element, just people writing more versions of old arcade games (but now they're Open Source).
Rather than discussing the current crop of games in an adult way, perhaps a better approach would be to try to foster a generation of game creators who can think for themselves and want to distance themselves from what's expected of the so-called game "industry." After all, it's common for writers and musicians to start out in a subversive way and grow into mainstream: Kerouac, REM, Hunter S. Thompson, Smashing Pumpkins.
This story, about Borland's upcoming Delphi for Linux, has drawn a very clear line in the sand between introverted Linux zealots and the people who are going to move Linux forward. I see endless responses like "Bah! I can just use gcc!" and "Only fools would use a RAD environment! Real Linux programmers will keep using gcc and an Emacs-based IDE."
A fantastic--maybe the best--compiler and GUI development system is coming to Linux and we should ignore it? We should keep flogging ourselves until we remember that UNIX is The Only Way? What is happening here is that an elitist crowd with a confused agenda is suddenly being confronted with the painful truth that Linux development software is far, far behind what has been available for some other systems. This realization needs to be embraced, or we can never advance. We can only fool ourselves for so long.
And, you know, this may make Linux difficult to differentiate from Windows. But hasn't that been the plan all along? I mean, we could have been devoting energies to desktop environments that aren't re-workings of Windows, right?
The compiler is great indeed - but also Pascal is a language you can compile in one pass, versus several pass required for C or C++.
It is also an order of magnitude faster than compilers like Free Pascal. It is extremely well-done.
Actual, don't knock me off as a troll. I'm trying to say, why should I bother with Borland when I can use gcc/g++ and any ol' IDE (sorry about that obscure pun) that I want. I just don 't make sense. I already got dem tools
But you for you windows boys, yes now you too can write programs for Linux!
Don't knock Delphi until you have tried it and written a app with it. It's amazing. The compiler is maybe the best compiler ever written. Compile time is effectively zero. There's no link time. Object Pascal has niceties, like having single file modules instead of source + header files, that make you wonder how you ever managed with C++. You can ignore all the GUI and OOP stuff and just use Object Pascal as a straight compiler. The built-in assembler is so nice I've even used it for all-assembly programs, as it's much faster than NASM/MASM.
And all of this is really just used as the back-end for a very slick and complete GUI design package. You can click on controls and add event handlers for just about anything. It's very complete.
The killer is that everything is extremely solid and well-implemented to a level that is foreign to Linux desktop environments. gcc may be nice, but we've all had our general oddities and bizarro error messages from it. Putting a cranky GUI layout package on top of this is not the same thing. The bottom line is that Delphi is much, much stabler and slicker than what we're used to using in Linux. Ignoring it for GUI-based programs is on par with writing an OS in assembly code instead of C. Don't wear a hair shirt because this comes from Windows. Delphi will change Linux application development, no question about it.
But there's a compiler underneath. The Object Pascal compiler--part of Delphi--is an amazing piece of work. Compile time is zero on just about any machine. That is a huge, huge reason to choose Pascal over C++ for this sort of thing.
I see much misinterpretation of the term "systems programming." It has become a euphemism for low-level bit tiwddling and device driver hacking. In fact, the term is quite literal, meaning "the programming of complete systems." If you look back at the original UNIX, it was an OS, a user-environment, and a programming language, all of which were developed to work together.
A goal of systems programming is the usability of the entire package. This is one of the stumbling blocks for what Linux has grown into. The basic command line environment is comparable to the original UNIX in this regard. Cryptic, but powerful for a certain type of user and application. But that's where the plan stops. GNU tools tend to be messy critters, with way too many command line options. Look at the man page for ls, you'll see a note about how the man page is no longer maintained (use texinfo instead). X Windows feels like a giant hack to bring UNIX into the bitmapped age. The desktop environments for Linux all feel like attempts to bring Microsoft Windows into the realm of free sofware, but they're duplicating the same mistakes: too many gadgets, too many meaningless icons, too much focus on fiddling, too many questionable interface descisions. In the end, we have an interface that's not sure what it wants to be, outside of a slap at Microsoft's market share, put on top of a byzantine graphics system that is gernerally unliked, on top of of mostly unrelated early 1970s era nuts and bolts. It is free, and this is good, but it can be viewed as a way of keeping programmed running in circles until something truly new comes along and makes everyone re-assess their goals.
It is also worth considering that the era of big operating systems is coming to a close. Who could have believed that a Palm, with it's little 68K chip, would be so bloody useful? Or that the Game Boy is closing in on 100,000,000 units (that's 100 million)? You could argue that web browsing and word processing and coding are outside the realm of such toys, but is that really true? There's a cool little Forth for the Pilot, that's probably the slickest programming system I've seen in years. And there's going to be cool stuff in the future from other non-desktop hardware. Is trying to tack Windows onto Linux really such a worthwhile endeavor?
I expect this to be moderated down. Sigh.
.College students, because there's some benefit to using Linux if you're a comp sci major, and also because it's free (either "free as in beer" because students are generally cash short, or "free as in freedom" because it's easier to be idealistic when you're a student). High school students are included here, too, though maybe to a lesser degree.
Right now, there are three primary groups using Linux:
1. ISP's and other businesses that need rock solid networking and file serving.
2
3. Geeks who fixate on which operating system they use. This overlaps somewhat with the previous item.
Right now, Loki is selling mostly to number 3, and I suspect this is the smallest and most volatile group. Number 2 is where the users are, but that's a tough place to make money. It's the same place crazy Napster support is coming from, and for a very obvious reason.
To an outsider whose eyes are clear of zealotry, the Linux game market looks like this:
1. There are only a handful of commercial games.
2. Those games are also available for Windows, so even if I decide to pick up Linux at Barnes & Noble I still have Windows around to play games with. Not sure why I would want to play the Linux version.
3. The freebie games are pretty horrendously unexciting. Yay! Hundreds of Tetris and Asteroids variants! Does Open Source somehow imply a lack of creativity?
4. All the nifty games like The Sims and FreeSpace 2 and Roller Coaster Tycoon aren't out for Linux. Maybe they'll get ported next year, but I don't see why I should paint myself into a corner just for the sake of supporting an operating system.
Harsh? Maybe, but let's be honest about this. There are very few reasons to even consider using Linux for games, unless you've decided to hold fast to OS-centric views. If there were something really sweet that were available only for Linux, I could see _some_ people peering over the wall to see what all the fuss was about, and maybe setting up a partition to play. But we're a long way from seeing that happen.
As processor power increases, the difference between good and poorly written software is becoming more dramatic. This is a greater issue that's being dwarfed by pissing games over 1GHz+ speeds. For example, twenty years ago it was common for developers to use VAX-level minicomputers to do development for emergine home computers (Atari, Apple II, early PCs). After all, who could write a decent assembler on a 1MHz system with hardly any RAM? But it turned out that native assemblers on teeny tiny systems were frequently outperforming the minicomputers by a factor of ten or more.
Fast forwarding to today, consider any C++ compiler, say gcc or Visual C++. You can never have enough processor power for such a compiler. Even on top of the line machines you're still talking minutes to rebuild a medium sized project, and seconds for an average link. Now fire up Borland's Object Pascal compiler (buried inside of a RAD tool called Delphi). On a 200MHz machine the compilation time is up near a half million lines per minute. Link time is effectively zero. In general, compilation time doesn't exist. You have to have a pretty big project before you even notice that pressing F9 is taking any time at all. If you're using a 300MHz machine or faster, this is never even an issue. It's never an issue for almost any slower machine either, but I'm playing it safe. Now, yes, Object Pascal is simpler than C++. The compiler maybe doesn't do some of the nutty stuff that gcc does. But in the end, does it matter? A compile time of zero sure does make it easy to go in there and twiddle around with the code, making it go real fast.
This kind of thing is going to be more and more common. Is a bulky application slow because the processor isn't fast enough or because it is bulky? Throwing more processor power at problems like this is a dodge.
Your X server is what controls your video card and mouse. That's all it does; X does not have menus, windows, or any of that stuff, because of the way it was designed.
And this is why it's shocking that X is as bulky as it is. It does very little, yet sure requires obscene overhead in order to accomplish it. Isn't this one of the shortcomings of Windows?
One of the very first self-published books I ran across on the web was Halcyon Days: Interviews with Classic Computer and Video Game Programmers. It was released back in 1997. I remember reading about it in a number of places, including Wired. It might be worthwhile to talk to the author and see how it turned out.
Maybe there are other examples of this kind?
You are making some common mistakes here. The first is that cutting out the middleman is a good thing. Ask most any shareware author, or someone in a local band, and find out how much money there really is in getting people to buy something from an unknown. Getting something distributed outside of your own meager means is a huge, huge boon.
Second, there is a reason that publishers reject manuscripts. Sometimes they are wrong. Zen and the Art of Motorcycle Maintenance being a famous example of a book that was rejected dozens of times before becoming a bestseller. But often they are right. Just because you think something is great and worth reading doesn't mean that it is. You are asking web surfers to be your editor, and to read your story with the foreknowledge that it may be downright lousy. This can work, if you are very good at what you do, just as many bands act as their own producers, but most people need someone else in the loop, especially early on.
Third, you need the experience of going through a true editing and publishing process in order to gain experience. Playing music in your bedroom is completely different than trying to put a CD together. In effect, you're just going to put a tape recorder in your bedroom and give the result to the world.
Finally, the web is becoming very full of garbage and this is making people cynical. There are endless home pages of poetry from high school girls. There are sites for pseudo-companies, complete with faux press releases and corporate titles, put together by some junior high kids who decide to put together a software developer. A site with some guy hawking his Great American Unpublished Novel isn't going to garner much confidence or attention.
Attempting to get published is still the way to go.
The two major candidates in the next presidential election are viewed as cardboard cutouts thrust forward by their respective parties. Neither is a great statesman, a peacemaker, or a highly respected man of wisdom. They're not horribly offensive, and they have business-as-usual middle of the road views, if those can be considered what it takes to become president of the United States. Yes, there are other candidates involved, but they're not going to get more than an insignificant handful of the popular vote.
Two questions:
1. Aside from "rah, rah, democracy is great, rock the vote" nonsense, please justify why voting in the upcoming presidential election is worthwhile.
2. Having two bumbling, unqualified candidates makes US-style democracy laughable in the eyes of other countries. Does this hurt democracy in general?
It is not sad. You're simply parroting Linus Torvalds's "release early and often" advice. This works for an OS, because compatibility problems are the issue of the day. There are also reasons not to do so:
Most people don't take the time to upgrade, so they'll miss out on major features that are added later.
If you release something that is incomplete, people will try it, see that it is incomplete, and have a poor impression from that point on.
If promising software is released at an early stage, there is likely to be much more "cloning activity," channeling effort into doing things again--the right way--instead of tweaking something that's already most of the way there
It's like that in all the major high-tech industry areas, at least in the united states, prolly the world eh?
No. Silicon Valley is bad. Seattle is bad, but not so much as the first. Dallas is fairly inexpensive, and it's chock full of high-tech companies (lots of telecom, for example). If you compare house prices on the web you can get a good feel for the general priciness of an area.