Catch up on stories from the past week (and beyond) at the Slashdot story archive

 



Forgot your password?
typodupeerror
×
Handhelds Hardware

Game-development on Compaq iPaq 188

kilaasi writes "Some hard-core game-developers from Finland is making super-optimized games for the iPaq and similar devices, tweaking and tuning every bit of piece there is. These are old Commodore and Amiga-programmers that know the virtues of small-is-beautifull."
This discussion has been archived. No new comments can be posted.

Game-development on Compaq iPaq

Comments Filter:
  • Uhm... (Score:5, Funny)

    by Bob McCown ( 8411 ) on Tuesday November 06, 2001 @06:27PM (#2530092)
    These are old Commodore and Amiga-programmers that know the virtues of small-is-beautifull."

    I keep telling my sweetie that, but she doesnt believe me.

    • "In 1993, Fathammer founder Samuli Syvähuoko helped write Unreal, a 3-D demo that ran on old 386 PCs and won first prize at Assembly '92."

      It could be argued that Samuli's invention of time travel is nearly as important as having Doom on a PDA. But I don't think so.
      • As you may possilbly be a troll, i thought twice before posting this. You may just be a stupid 14 year old who has s**t for brains.

        The 3-D demo was called Unreal. Note: It did not say the 3D FPS Unreal by epic (iirc). Think Final Reality, (i am sure most people who have had windows machines have used it at least once) that was a Demo. A look at this cool shit that i can do with ASM and some Crap/Good/Odd Hardware/API.
      • "In 1993, Fathammer founder Samuli Syvähuoko helped write Unreal, a 3-D demo that ran on old 386 PCs and won first prize at Assembly '92."

        The original Unreal was released in 1992, an updated version (with GUS support) in 1993. No time travel involved, just some sloppy writing in the Wired article.

        It should also be noted that the Unreal demo has lots of parts, some in 2D, some in 3D.

    • Get one of these [penisruler.com] and you'll look bigger.

      M@
  • by Anonymous Coward on Tuesday November 06, 2001 @06:28PM (#2530094)
    ::)
  • Less is more (Score:4, Insightful)

    by Brian Kendig ( 1959 ) on Tuesday November 06, 2001 @06:29PM (#2530105)
    It's often said that the old arcade games of the early 1980's were some of the best ever created because they had so little to work with -- and therefore they were forced to focus on gameplay over glitz.

    If that same rule holds true for the iPaq, it might become one of the best gaming systems ever conceived. :-)
    • Re:Less is more (Score:2, Interesting)

      by Zspdude ( 531908 )
      The question is how much impact will this have on the glitz-greedy public? They don't all use iPaqs, and in this day and age games are sold on the basis of glitz- good gameplay is a bonus, not a requirement.
    • If only this were true, but we know that it is not. Just take a look at what we have on systems like the GameBoy. Sure, there are some good titles on it. But it's mostly crap.
    • It's often said that the old arcade games of the early 1980's were some of the best ever created because they had so little to work with -- and therefore they were forced to focus on gameplay over glitz.

      Kind of.

      I just installed Mame and tried a few of the games I remember from the past. For some of those titles that I used to think fondly of, I was shocked how boring and repetitive their gameplay actually was.
  • by cd_Csc ( 151701 ) on Tuesday November 06, 2001 @06:30PM (#2530110) Homepage
    It's great that they're doing this - it will certainly allow for some cool games in the future, but not quite yet... the iPaq has a hardware "feature" that prevents programs from detecting simultaneous usage of more than one button. Nothing sucks more than having to stop moving so you can shoot or jump. To counter this, developers have built "virtual buttons" that appear on the touch screen, but this takes up alot of the already limited screen realty. Plues, its hard to get used to not having the underappreciated tactile feedback of physical buttons.
  • people have been eeking out performance out of ARM cpus (ipaq is just a strongARM clone) for a very long time
    since the ol acorn in the corner of this roomcan atest to

    did a space invaders clone when I was 15 should dig that out (-;

    really you should look at the GBA as its the same ISA but instead of a pultry 78MHz on GBA its a whooping 200Mhz on the IPAQ

    equate that with moveing from a 486 to a P200 and you get the idea !
    (yeah yeah not the same, RISC, improved piplines.... give it a rest I know already)

    should be relitvly easy to do a GBA clone on the IPAQ as its the same ISA why havent we seen this before ?

    regards

    john jones
    • > should be relitvly easy to do a GBA clone on the IPAQ as its the same ISA why havent we seen this before ?

      The problem is that I think it has custom chips other than the StrongARM. I could be wrong. Also, the graphics speed on my GBA is a lot faster than my Palm, and seemingly than the IPAQs I've seen. They're also a touch button deficient.

      What you should be doing is turning the GBA into a PDA :)
    • Errr problem with this comparison of power between the two machines is that their architecture is completely different.

      Just taking the MHz of the ARM CPUs is not really enough. Just to clarify, the GBA has a cpu@16.7MHz, but the IPaq cannot emulate GBA games properly as it does not feature:

      X-Y/rotational/parallax scrolling.
      Up to 4096 hardware sprites with scaling/distortion/rotation.
      Hardware alpha blending/transparencies.

      Even if you were to write *really* good code on the Ipaq, you would still have a heck of a time emulating all these purely in software.

      And thats not forgetting the amazing RAM spec:
      32KB work , 256K extended work, 96KB video, 16KB sound.
      Impressive indeed :)

      Regards,

      Po
  • Anyone remember what Future Crew's C-64 demos were?

    I remember a C-64 demo called "Edge of Insanity", which displayed (amidst a funky backbeat) a hysterical tale of blood, gore, and doom that went on for page after page after page.

    Anyone remember the original authors of this thing?

    I'm damned if I can confirm it, but I vaguely remember a reference to Future Crew. But it was a hell of a long time ago, I no longer have the disk, and I could be confusing it with some other demo I enjoyed about the same time. But I do remember Future Crew from way the hell back. Far fucking out to see them still kicking ass.

    • A reference to Future Crew would probably be included in the "Greetz" section in the credits or during the demo. Future Crew was probably the most greeted demogroup out there, because of what they achived.
    • Re:C64 Demo Scene? (Score:3, Informative)

      by Anonymous Coward
      Future Crew is still kicking ass, yes! :)

      I was the group organizer in Future Crew during 1989-1994. My nick was GORE. In addition to me, three of the original FC members are working for Fathammer. The others are either working at my previous company (Remedy Entertainment, maker of Max Payne), Bitboys (www.bitboys.fi) or at some other Finnish high-tech company (e.g. F-Secure or SSH).
      • Hmm, whats going on with BitBoys? are they going to release something out soon? heard lots of promises, nothing in the end.. which is too bad...

        Hey nVidia - there are some clever guys in BitBoys - buy them ;)
  • by mj6798 ( 514047 ) on Tuesday November 06, 2001 @06:34PM (#2530136)
    "Small is beautiful"? These people are programming a machine with a 200MHz RISC chip with 32Mbytes of memory. That isn't small, that's high-end desk-top performance of a few years ago.
    • Small is beautiful"? These people are programming a machine with a 200MHz RISC chip with 32Mbytes of memory. That isn't small, that's high-end desk-top performance of a few years ago.

      Its not the iPaq they're crowing about, its the fact that they can squeeze it onto a cellphone like platform where you don't have a speedy cpu or gobs of memory. Apparently thier platform is quite svelt.
    • Keep in Mind that the iPaq only uses the 'Q' button for closing apps, so 32M/206Mhz is pretty much the peak. Also, keep in mind many 3D games use far more- not memory- but disk space: the real premium on the iPaqs.
      • Hmm. The original doom shareware zipped up to a single floppy. Unpacked to something in the area of 2 megs. The original Quake shareware was 5 megs on disk, 9 unpacked. Sure it get's expensive from there, but...

        32 M/206 Mhz *is* a lot to work with. The Atari Lynx fer instance, has 64K (8 k of which has to go for a frame buffer.) And 4 Mhz. The original Gameboy wasn't any better equipped.
        • Actually, Doom was two floppies and 5 megs installed (the registered version was about 13mb). Kind of scary I remember this stuff.

          The Gameboy was even worse. It has a 1MHz processor and 8k RAM (with up to 32k extra on the cartridge)

        • While the Lynx was awesome and before its time, it's not really a good comparison. You have to remember people don't use the iPaq as a game console... not yet. So for right now, people use them to store offline web pages, e-mails, addresses, notes, media files, etc. which leaves little room for something like the 5 or 9 megs you speak of. The reason the Gameboy and Lynx aren't good examples is becuase of their 2d-ness.

          Man I miss my Lynx though... didn't it use 8 AA's?
      • Once again, I have to stress other PDAs' importance. And, once again, I have to mention that new Sharp PDA. Got a CompactFlash drive built in. And, what have we recently seen in Flash? A 1-Gigabyte card. That enough for a decent PDA/3-D game?
    • These people are programming a machine with a 200MHz RISC chip with 32Mbytes of memory. That isn't small, that's high-end desk-top performance of a few years ago.

      Not quite..
      The problem with the iPaq is that it uses the StrongArm processor which lacks a floating point unit. So, all floating point calculations are to be done using a software emulator. This is where performance really sucks. You cant use any normal mpr player for instance. Try cross compiling mpg123 for the ipaq running linu ans you will know what I mean. You need special applications that make use of Fixed point arithmetic instead. madplay which is used as an audio player on the ipaq uses fixed point arithmetic, but the decoded quality isnt quite what you would get from mpg123.
      • StrongArm processor which lacks a floating point unit

        Oh, well...that could be a problem, but it isn't. Simply because at that point they use fixed point for "scientfic calculation". Way back in the days, it was very common to use long or int with a fast library that implemented fixed point. In a stric sense this is not exactly "emulation" of floating point.
        I know there are accuracy problems and other limitations, but it has been done and it is most probably enough for this kind of game. I'm pretty sure these guys used this method.
        As an example, I would like to state "Mechwarrior" (which some of the screenshots reminded me of), that was completely written in fixed point IIRC. Gorgeous game, back in the days.

        • ...and next time I should read your comment to the end before shouting out. You knew about fixed point. :-)

          Of course fixed point won't help your mpg123 player, because the accuracy and range is not enough with fixed point. Games and MP3 players are however in a different league: games should "look good" (nowadays at least) and all tricks are allowed. MP3 decoding is more like the "sticking back together a fourrier decomposed wave", which sounds quite like what is it: exact maths.

  • by Anonymous Coward
    i'm using ibm via voice to control the screen on a bluetooth enabled pocket pc 2002 with 128 mb sd card, connecting to a ericsson t68 with bluetooth and using gprs. the 64k colors look great, the sound is stellar. most folks here trash microsoft no matter what they do, but the pocket pc 2002 os is amazing so is the compaq hardware. i tip my hat to ms on this, nicely done.
  • by igrek ( 127205 ) on Tuesday November 06, 2001 @06:37PM (#2530149)
    The PCs are bulky. Those 3D games make sense on PC.

    But PDA are small and flat. The PDA games should be 2D.

    What we need now is 1D-game. If you know what I mean.
  • Unreal? (Score:3, Funny)

    by death_denied ( 533148 ) on Tuesday November 06, 2001 @06:37PM (#2530150)
    Fathammer founder Samuli Syvähuoko helped write Unreal, a 3-D demo that ran on old 386 PCs

    I wonder if the engine could be rigged to run something else by that name.

  • Future Crew (Score:2, Insightful)

    by Anonymous Coward
    I see a good portion of these people are the same people who were in the Future Crew demo group. Those guys made the coolest demos. Hell, they are still fairly cool. Skaven and Purple Motion are actually decent composers. I still listen to the music from Unreal 2 once in a while. Back in around 93 I found their demos on a local BBS. I hadn't seen anything that good before (on a PC). I purchased Max Payne and though it was pretty good, I didn't realise until now that they were the same people.
  • by sprayNwipe ( 95435 ) on Tuesday November 06, 2001 @06:44PM (#2530191) Homepage
    I'm suprised that you didn't mention that John Romero and Co have moved to making games for the iPaq at MonkeyStone Games [monkeystone.com].

    Not only are they making games for them, but also trying to base a business on them.
  • by Blackwulf ( 34848 ) on Tuesday November 06, 2001 @06:55PM (#2530246) Homepage
    Wow, the good ol' days. I was late, I didn't start getting around to the demoscene until Second Reality came out at Assembly'94. Then I was hooked. There were many hoaxes of "Third Reality" coming out at the next big demoparty, as I recall.

    A lot of the old FC crew created a company called "Remedy" which creates the 3dmark benchmarks and recently released the game Max Payne. Purple Motion even did the music for part of 3dmark2001.

    A few people on an IRC channel I used to frequent just found a 64k intro from The Party 2000. They said "wow, when did people do this?" When I started telling them about the good ol days of MS-DOS and the demos and intros (and 4k intros!) of that time, they all turned their noses and said "EWWW DOS was NEVER good for ANYTHING! Yuck!"

    Of course, back then, the amount of polygons you could fit on a torus was the big challenge. It was what originally got me into programming. I feel so old now.

    Of course now, it's so easy to create jaw dropping images without optimized code, so it's nice to see that there is something to really test your skills on like the iPaq. I miss seeing elegant code.
    • Ahh...you beat me to it :)

      If anybody can make cool games on a handheld, it is future crew. I still remember my amazement at 'Unreal' on a 486-33 with an 8 bit soundblaster pro. Around the same time 'Ultima Underworld' came out, and also Wolf-3D. Those were the days. Fast, tight code. *sigh*

      On another note, I think I remember seeing josh jenson's name on some linux stuff somewhere. Remember him? Did all that awesome sound mixing code for the Renaissance Demo Team, also wrote his own game, Zone-66. All while in high school! Much of that sound code was used in many games of the era IIRC.

      • I got Zone-66, loved that when I was younger and couldn't believe it ran on my peice of crap system. Of coruse, it was a great system when I got it.


        I really envy some of you, a lot. A lot of the time, too. All of you, pardon the term, older geeks had it pretty good.


        I love the limited enviroment, elequent coding of days long gone. Problem is, I was never a coder in those days. I was in elementry school coloring pictures. I had an interest in computers, but I didn't have much expirience in them. I couldn't afford them until they got to where there wasn't much of a challenge.


        Maybe, that's why PDA programming appeals to me. But, I have to wonder: Where do us younger coders fit in with the code-it-small crowd that is still going on? In the days of Java and other languages that do everything, running on systems that could practically calculate the Big Bang.


        Where does one start to enjoy those times when one was never able to expirience them for one's self?

    • There still exist people today who write hardcore visual programs that are small and elegant. Unfortunatly they are usually crackers and a lot of them may end up in jail before they are old enough to get real jobs. : (

      Here's an archive of cracking intros:
      http://members.easyspace.com/erekose/f_1997.html

      I'm fond of the Paradigm intros myself. Some of them have kick ass music too!
  • It's About Time.. (Score:3, Informative)

    by Scothoser ( 523461 ) on Tuesday November 06, 2001 @06:56PM (#2530247) Homepage

    It is about time that programmers realize that embedded systems are not desktops. Hard drives are not an option with these things.

    More attention needs to be placed not only on making smaller programs perform better, but getting the program to perform closer to the hardware specs. This is what programming used to do with Assembler.

  • Lost art? (Score:2, Interesting)

    "The skills you need to optimize a software rasterizer and make it cross-platform have been forgotten by programmers relying on today's beefy (desktop) PC and console machines," said Fathammer CEO Brian Bruning. "It's something of a lost art."

    I'm curious to know why this is such a lost art. Could it be due to the fact that most engines are proprietary code? Did this lead to a state where a limited number of people have access to the code? Even fewer that would want to muck with 'legacy' code in the engine? What about publishing this in a book? I've read "The Black Art of Game Programming" which I found informative; Does this book not dive into the secrets? What are the secrets? It occurs to me that maybe these lost arts come from optimizing solutions to specific hardware platforms. Could these skills be lost because of the hardware dependencies, where as the evolution of software engineering has gravitated toward abstractions such as portability and a more OOP structure? If the knowledge of the art were important or interesting enough to distribute, where can we find it documented?

    Don't mind me. This was a stream of consciousness ramble.

    • Re:Lost art? (Score:3, Informative)

      by Junks Jerzey ( 54586 )
      I'm curious to know why this is such a lost art. Could it be due to the fact that most engines are proprietary code?

      If you want to find out how to write optimized software rasterizers, you can easily get the information from books and the web. This is old news. Everyone was into this back in 1995 and 1996, and in the end there was one generally accepted, close to optimal inner loop that was used in most texture mappers. It was published in Game Developer magazine and is available for free on the web.
      • Re:Lost art? (Score:1, Insightful)

        by Anonymous Coward
        This one "generally accepted" inner loop was written for the x86. These things vary by architecture. A lot.
    • The real lost art is the drive to dive down in and optimize. The "secrets" in those books tend to be specific to a certain chip, so how you optimized for a 386 isn't really the same as how you optimize for a Pentium, which is different from how you optimize for a Pentium 4, let alone a StrongARM.

      A large reason is that it isn't needed as much for getting higher performance. Just wait 6 months, and you have a lot more power at hand. It would have taken you 6 months to fully hand optimize the engine to begin with, assuming that you didn't need to totally rework your OO model to accomodate the optimizations.

      The biggest things people can do is to figure out how to analyze code for performance. This include making sure that you are using good algorithms at a high level (say C++), and making sure you have profile were the bottle necks are (no point optimizing in assembly something that only runs at the start up of the program). Then learn to count CPU cycles and how to use even lower level timing and profiling tools.

      It really can be a lot of fun, but it also is something that should be left to the very last step. So, if the game isn't worth playing, don't bother optimizing.
  • Ah, that's RJ Mical (Score:3, Informative)

    by Junks Jerzey ( 54586 ) on Tuesday November 06, 2001 @07:00PM (#2530262)
    He co-designed the 3DO and Atari Lynx, plus was an OS guy for the Amiga (note that he did not design the Amiga hardware; that was Jay Miner). And now he's the lead tech guy at Fathammer.

    Of course in this case it is debatable whether the best games for a system such as the iPaq should be hardcore 3D. If you take that route, then 98% of the processor time immediately goes out the window.
  • by Embedded Geek ( 532893 ) on Tuesday November 06, 2001 @07:02PM (#2530268) Homepage
    I hope we see some good game development. I'm working on my employer's next generation in flight entertainment system and we need better games for the passengers to play. It's depressing that all the serious development of late has been for higher end systems - we're running a 266MHz embedded x86 at the seat under Win CE and there's little to choose from out there.

    Although I'd like to rejoice at this news, I fear it won't help us much. With M$ pouring resources into XP and Xbox, I fear that CE (with its very reasonable liscencing terms) will become yet another orphaned child from Redmond.

    • Cool! But, er, why don't you just talk to Sony and license the PS One? I mean, it's around $100 retail, I bet that could come way down with a little bulk purchasing action. Sure, it's not top of the line anymore, but I'd guess the games available on it still might look interesting next to whatever you can get going on a 266MHz x86 system (no offense). Or?
      • Actually, on our older system several years ago we used a variation on that theme. Basically, we used a Super Nintendo as the core video engine. Worked very closely with them. They turned around and filed a patent claiming that they invented the entire concept of in seat, airline video - a claim that even we have no right to make. The thing that really sticks in our craws is that they used some of the diagrams we drew in our documents and placed them directly into the patent!!

        Also, to meet the vibration, reliability (99.9+% up time), power consumption, and EM radiation standards for running inside an aircraft we had to modify the hardware substantially. When that happened, most of the "off the shelf" savings went out the window.

        On a software front, though, the difficulty with a game engine is that these seats are expected to have a life of at least five years, but most airlines wind up pushing them as long as possible - there's a lot of equipment running out there 20 years old. Finding new games throughout the lifespan of the product is bad enough, but finding developers to maintain the system is even tougher (would you be interested in maintaining our Super NES code? Thought not - can't blame ya'!).

        Another thing driving the processor/platform choice is web browsing, specifically the use of IE (which, let's face it, is the best choice to expose a non-tech savvy flying public to). That drove us to either CE or embedded NT (embedded XP wasn't on the radar yet) and we picked the cheaper choice and were able to budget in 64M of socketed RAM (again, a pitance, but the hardware guys originally wanted 8M soldered - killing any hope of upgrade).

  • Well, I just upgraded my iPaq to the unstable version of qpe, a GUI built over qt embeded for iPaqs with a basic Familiar Linux, and believe it or not it has QUAKE.

    man, this quake thing is quickly becoming omnipresent, some more time and we'll have more ports of quake than space invader or pac man...
  • Now if they could only port Rune to the ipaq... imagine the possibilities!
  • Speaking of small demos, there is an execellent, high quality 11 minute 64 kb pc demo called the product [theproduct.de]. (sorry, windows only, DX 8 required).
  • by Anonymous Coward
    We know no one wants to play a platform game with a pen, so obviously the next hack will be a joystick - but this begs the question:

    How the fsck do you hold the iPaq AND a joystick at the same time?!

    *clunk*
    "Shit... dropped the joystick again."
    (repeat 3x)

    "Okay, maybe if I hold it like *this*."
    *clunk* *crackle*
    "SHIT!!#$!@@@@@%!^^@"
    • You make a joystick that plugas into the bottom of the iPaq and makes the whole device a gameboy style device.

      Seriously until a good joystick comes along gaming on these devices suck. I've tried to play MAME games on it and can not make it past the first levels most of the time because of th shitty input interface
  • by headkase ( 533448 ) on Tuesday November 06, 2001 @07:31PM (#2530378)
    From the article: These are old Commodore and Amiga-programmers that know the virtues of small-is-beautifull.
    For their time, nothing comes close to Commodore computers, the C64 sold 22 million units between 1981 and 1987. I started out with a C128 (I rarely ran C128 programs, instead I almost always ran it in C64 mode) and migrated to the Amiga's in 1989. I started out with an Amiga 500 and moved up to the A1200. Those machines were way ahead of their time, they were multimedia machines before the phrase was coined.
    They had 4 channel digital stereo sound, could display 4096 colors out of a palette of 16 million onscreen at 1 time (this was when 16 color EGA was the rage on PC Clones). They had a fully multitasking operating system, and it was completely GUI orientated. They were also plug and play too, but they called it auto-detecting the hardware. I own a PC now, but at the time I'm glad I was an Amiga user instead of a PC user, I never had to go through all the troubles PC users were plagued with at the time (remember setting jumpers for ALL your hardware, and praying there were no conflicts?).
    • On the Amiga 500 - 16 out of 4096 unless you were in the HAM mode (Hold and Modify - very flickery) with the AGA chipset...

      I think the ECS gave more colours (256 out of 4096? not sure. anyone?)

      Ahh, those were the days (says a man who had Amiga 1000 with 1MB RAM with the Insider card)
      • ECS gave 32 out of 4096. AGA gave 256 out of 256,000.

        Josh, who had an A2000
        • Correction- OCS and ECS 16, 32, 64 (HalfBrite) and 4096 (Ham 6).

          AGA- all up to 256 colours, and 262144 (Ham 8).

          When CBM folded AAA was in development offering Ham 10 on the Mobo...
          near enough to 24bit as to not matter! THIS was amoung the Developer
          Boards that Haynie sold on E-Bay recently!

          ........

          Amoung STANDARD Amiga capabilities still unmatched by mainstream
          Desktop Computers are .....
          * Shut-down at the Power Switch!
          * Formatting up to FOUR floppies simultaneously.
          * Continuing to work from OTHER Partitions while Scan-Disking.
          * ditto... with Re-organizing!
          * Recoverable Ram Drives,... contents intact even when re-booting
          to ANOTHER version of the OS!

          report [iprimus.com.au]

          John, on an A2000/060/32Mb with 4 Bootable Partitions on 2 HDrives.
          Standard Serial Port to Spirit v.90 modem. Original screens
          Flicker-Fixed to PicassoII display card. Amiga '88 to '01
      • The Amiga 500 had regular color modes up to 64 colors, in 64 color mode (half-brite), only 32 of them were unique the other 32 were half the intensity values of the first 32. Then there was HAM mode, 4096 colors. ECS does allow for up to 256 unique colors before its special modes like the picasso modes of 16-bit color (Of course you needed a picasso board for that). I also can't remember what other built in high color modes the A1200 had.
        I've seem some HAM pictures with more than 4096 colors too, but they switch the palette around on an interrupt as the raster draws the screen to achieve this.
    • by DGolden ( 17848 ) on Tuesday November 06, 2001 @08:13PM (#2530574) Homepage Journal
      it was completely GUI orientated.

      Well, v2.0+ also had system-wide ARexx scripting, a powerful shell, user-space filesystem drivers/translators so you could install a driver to let you cd into compressed files, the window system itself, etc. The entire GNU command-line toolset was also ported to it via a compatibility library called ixemul. The OS was built on a message-passing-by-reference system, which meant that IPC was zero-copy. There was also a very powerful networking add-on called Envoy that provided network-transparent messaging services.
      It also had fun late-binding shared libraries, that could be patched dynamically at run-time on a per-function basis, allowing third party hacks to theme the GUI and tune the OS on the fly.

      So, it had a kick-ass GUI, but it was good at lots of other stuff too. :-)

      Where the OS fell down was its complete lack of true memory protection - at the time however, this had some advantage, since it meant the computers could be made with cheaper MMU-less CPUs, and meant that task-switching was extremely quick. Amiga applications tended to be naturally multi-threaded with non-modal GUIs, so fast task-switching was a definite plus.

      Interestingly, there's a re-creation of AmigaOS for x86 available here [aros.org]. It's actually coming along very nicely, but has all of AmigaOS's weaknesses, as well as its strengths - e.g. no memory protection, but ultra-fast reboots for when you do crash :-) (a soft "reboot" actually just vectors back into the kernel entry point, skipping the BIOS and bring the back system up in seconds.)
    • For their time, nothing comes close to Commodore computers

      It's completely nerdy to bring this up, but you can't forget about the Atari 800 and friends. They were released three years before the C64 and were much better in some ways (color, interrupts, overall architecture) and poorer in others (sound, sprites). But I never met anyone with extensive experience on both machines who didn't tip his hat toward the Atari. You could either spend a week writing perfectly timed raster interrupts on the C64, or just use the supplied hardware on the 800 (sort of like the Amiga's copper).
  • Inmar Software ( http://www.inmarsoftware.com/ ) has a similarily optimized 3d engine for the Pocket PC. It has a game ( http://www.inmarsoftware.com/minigolf.htm ) that runs on StrongARM Pocket PCs and uses this impressive 3d engine. With 128 MB CF cards costing only $50 and 64 MB RAM in many new Pocket PCs, storage is not much of an issue, compared to other PDAs. The new Pocket PCs (running Pocket PC 2002, http://www.pocketpc.com/ ) all use the StrongARM cpu, so these sort of 3d games will become more common place and of higher quality in Pocket PCs with the powerful StrongARM as the cpu. The new PPC 2002 devices do not have a problem with multi-button pressing, so the quality of gaming on them will continue to advance. (The iPAQ 3635 only costs $300 after looking for a good deal and getting the $150 rebate from Compaq. I just ordered mine.)
  • I think there are a lot of gamers out there with Ipaqs who don't want extraneous 3-D graphics and action games. What is wrong with games that would be more suited to the platform. Like strategy (No, RTS is not real strategy) or RPGS?

    http://members.fortunecity.com/broadsword/Computer /FreeCiv/FreeCivScreenshots.html

    Somebody started working on a freeciv port, but I think it has been abandoned. Thats too bad. I can't think of many games more suited to the Ipaq than Civ.

    Anyhow.. I just think all this Ipaq gaming development is going in the wrong direction. Someone should port dos to this thing (with VGA support) then we could play all kinds of good non 3d games at 320x200.
  • by Sindre ( 534738 ) on Tuesday November 06, 2001 @08:17PM (#2530595)
    Here's a download of the demo:

    http://www.infosync.no/show.php?id=985&page=3 [infosync.no]

    It's pretty cool!
  • I thought "131mhz, should be enough for those gameboy and NES (heck even C64) emulators...

    God I got a bad surprise... not only it's unplayable, I can't beleive it's tight-assembly code, either the microsoft compiler is really really crappy for MIPS device, or WinCE sucks too much ressources, or both of these reasons... The device isn't intended for gameplay, that I can understand, but heck, at 300$ a pop, (400-500$ for ipaq?) they might as well throw in just about every features they could.

    I know that the processor in the E-115 is a crippled MIPS R3000/4000 without the FPU and some other "useless and current consuming" core components. I can overclock it but it still won't change de fact that I feel some application would greatly improve with simple lowlevel optimization.

    All that said, It's nice to see some people coding low-level and pushing the envelope... Maybe they should work on pocketquake so I can get more than 1fps :)
  • I use my Ipaq for everything. I use it's calculator function.
    I sometimes calculate how my peers on slashdot will mark my post as "redundant" when it was never said in the first place.

    I sometimes calculate how much more karma I need to boost my ego.

    I sometimes calculate how much time it will take me to meta-mod these "meanie-heads," (remember, when posting use "school language") who moderate good posts as "troll," as unfair.

    I am calculating right now how long it will take people to mark this comment as "troll" or "flamebait"

    I also use the calendar function, but I shouldn't make my post to long. (remember, 20-40 lines)

    -skoobasteve
  • A joystick is totally possible with the iPaq. It's 'ActiveSync' port just a glorified serial port. It shouldn't be hard to splice an old serial joystick to an activesync cable. Compile a new kernel and shazamm!
    • arrange it in such a way that the two are solidly connected, then you only need to hold the joystick with a little screen sticking out, or the other way around depending.


      This, of course, would be a good standard to get set very soon. Don't want each game to need a different joystick, do we?

  • but if I want handheld gaming goodness, I'll stick with a $100 Gameboy Advance. If I'm feeling frisky maybe a WonderSwan.

    If I want a PIM, I'll get a $100 Palm platform device. If I want a Super PIM capable of holding a few extras, I'll get a $200 Palm platform device.

    Either way, I still have $200-$300 to spend on my next-gen home console.

    I applaud the hacker ethic at work here, but to be pragmatic I think there are better tools to do my job.
  • The value of this article is not that a bunch of people have built 3D games for the iPAQ or other small devices.. The value of this article is that these people were able to create extremely clean and compact code. This is critical in developing good software

    It just bugs me sometimes the number of lazy programmers that are out there. In one company that I have worked at, there was always a push towards getting it done rather than writing clean code. In one instance, one piece of software made 10 SQL queries that could have been done with 1. The reasoning for this was that, "the servers can handle it."

    Remember guys as you go out there and develop code, although "the servers can handle it now.." at one point in the future, software is going to reach the capacity of the hardware. At that point, all the people that have been creating sloppy code will suffer.
  • by rcs1000 ( 462363 ) <rcs1000&gmail,com> on Tuesday November 06, 2001 @11:13PM (#2531048)
    I was at Eidos headquarters about two months ago and they showed me Tomb Raider on iPaq. I was blown away. (And trust me, I've seen/played a lot of video games.) The quality (FPS, etc.) was better than the original PSX.

    As someone already commented, the controls were... interesting... but nothing that you couldn't get used to after a little bit of practice.

    Anyway, for anyone who cares, here is a link I saw about iPAQ TR:

    http://www.pocketgamer.org/archives/00000314.sht ml
  • It's got a 200Mhz processor with 32Mb RAM, for feck's sake!

    Physically small, yes - but it's got about the same power as a good PC did 5 years ago...

1 + 1 = 3, for large values of 1.

Working...