christo930 said:
I am not biased. I don't hate the Pandora, the thing doesn't even exist. I just want the GP2X to has it fair run
It's not hating Pandora that makes you biased, it's your strong desire to see GP2X continue to be active. I don't think you can deny that you see Pandora as a threat to this.
I think the Pandora is too expensive to nix the GP2X, I think it will segment an already small market, take away developers and users, but not enough to support the Pandora. THis will make 2 game system markets that don't have enough devs or users to keep them going. I can't possibly see how the Pandora can remain viable without a user base at least as large as the gp2x. It might bring in new users, but I really don't think it will bring in too many. To be totally honest, I hope I am wrong. I would love to see the big boys notice and realize that there is a sizeable segment of the market that want to put their dollars in open systems., but I think it's a pipe dream.
christo930 said:
would you mind pointing out how the gp2x sucks (from a users point of view). I understand there may be things as a developer that you feel suck, but since most people are not developers, those aspects that suck don't effect most people.
Most people here have the F100, as I'm sure you know it has a terrible stick. While GP2X has moved to a d-pad for F200 I hear it's not without problems of its own. The build quality is not good, there are flaws in many units (mine for instance had a broken SD slot), the USB and power inputs are on the side which makes it almost unplayable while using either, it can't charge batteries, it doesn't come with an AC adapter and the ones users have certified to work with it are anything but cheap (thanks to the high power draw).
I personally wouldn't dare buy an F200 either; this might not apply to users so much but without USB device support you can't really remotely debug on it.
All in all, the CPU is just disappointing, and having a second one is not an attractive alternative to including a powerful main CPU. It's really not that much more powerful than GP32, and the RAM clocks are actually even worse. The 2D acceleration is also pretty pathetic, and most of the chipset features (which I'm sure draw power) are almost useless for anything aside from video playback. GP32 is a good four years older than GP2X. You might think it's too soon for GP2X to die but when you see it as more of an incremental improvement over GP32 instead of a platform that really distinguishes itself from it in terms of what it can do (and actually makes a lot of things worse) then it's no surprise that users can't stay interested in it for long. If Gamepark released a version of the GP32 with a CPU that could go a little higher, 4-8x the RAM, and SD card support instead of MMC, then I think it would have beaten the GP2X in pretty much every regard. Granted, I don't own a GP32, so this is just based on what I know about its specs, price, and what users have told me after I asked around about it. I asked if they thought this would be the case and they generally agreed.
When I first got my GP2X I was ready to send it back after 10 minutes of playing with it (and I have the latest f100 with the improved joy cap), but a half an hour later I got used to it. I can not understand why they used that stick, especially when the gp32 had the best stick on any hand held, all they had to do was improve reliability. I don't ever play it with the usb cord plugged in, so that doesn't effect me. My unit has had no problems (knock on wood). I have a battery charger and about 20 AA rechargeable batteries and to be honest the thought never even occurred to me. The high power draw is a problem for the batteries, but 1 amp power adapters can be had for $10. That is how much I paid for mine and mine and it has a bunch of different voltages and connectors so I can use it with my camera and other devices besides. As far as comparing it with the gp32, I agree. I held off for a long time to buy the gp2x for that very reason. It just wasn't a big enough jump in power to make it worth while. But the back lit screen, the SD cards, real movie playing, extra buttons, newer version of MAME, NeoGeo, Sega CD emulation and a few other things made me make the move, especially the sd cards. I needed a new card and was looking at $80 for a new sd card since by then they weren't being made anymore. I figured 3 cards will just about pay for thing (saving $60 each time) and I like to play it at night (and on the bus) and I can't play it at night because of the lack of backlight. So the deal was cinched.
christo930 said:
My expectations of the platform as far as emulation goes is that tg16 emulation should be perfect, unexpanded snes should be near perfect, and gba should have a medium rate of compatability and good performance on a medium number of games.
I'm working on a PC-Engine emulator right now, so long as "full speed" and very high compatibility for the core platform is enough for "perfect" in that's what will happen.
SNES is probably not going to happen. Lots of people have made claim that SNES emulation should be better than it is, but you don't see that motivating people to pour lots of hours into something that might not pay off a lot, or at all.
The fact that devs might not do it, is not the same as saying it can't be done. At bare minimum the sound should be put on the second cpu. Now, that may not be possible, but it worked with other emulators and I haven't heard anyone say it wouldn't work.
If you ask me GBA already does a better than that, both in compatibility and performance - I don't know what your definition of "good performance" is though.
Agreed. Most of the games I've tried have worked well enough, but I haven't tried that many
christo930 said:
It's hard to quantify that last statement and I'm not a developer. I think a big part of the performance problem of dosbox is that it's ported from PC and that on the PC it's target hardware is pretty high. The reason I think emulating a 286 pc with ega graphicsc and an 8 bit sb, is the success of the Amiga emulation and the success of the snes.
Look, we've already been through this. 80286 CPUs are a lot more powerful and more demanding to emulate than 68k (Amiga) and 65c816 (SNES). SNES has a lot of video hardware but it can get frameskipped into submission. DOSBox for PC isn't a slouch and it's certainly not "fast enough" on high end platforms to the point where there's no interest in emulating it. Even on the best CPUs around I doubt it'll emulate at the level of more than a 75MHz Pentium or so and that's really not good enough. But it's expensive. It's true that there are a lot of things it does that are PC specific that haven't been done on the ARM version and it's true that the thing could be faster overall but doing a dynarec is a lot of work.
And I think that that target is not something that's very interesting, I'd expect most people who aren't content with 8088 level emulation will want 386+ emulation.
I think 386 emulation is unrealistic at a reasonable speed (say a 386-16) with VGA. But there are a lot of very good pc games that will work on a 286 ega - even more with vga (though I am not as confident that that is realistic). I am pretty sure I mentioned this before, but there was a pc emulator that ran on any quadra and emulated a 386 (it ran windows 3.x in 386enh mode). The gp2x is no Quadra, but were not trying to run Windows either. And saying it's a lot of work is like saying "well, we have software to port, so why write from scratch" There may be very good reasons for that. If I was a dev, I know that I wouldn't have time to write a pc emulator from scratch. But that doesn't mean it can't be done.
christo930 said:
In the end, maybe I am wrong, but an attempt has not been made (just ports of pc software).
You're WAY off on this one. A lot of GP2X emulators use ARM CPU cores, largely because they've been circulating around for a while. Some even use renderers written in ASM too. And I'll thank you not to call gpSP and psx4all PC ports, given that they both have dynarecs for ARM.
I only meant pc emulation. I am well aware of the work that has gone into some of the emulators
christo930 said:
Things that were said could not be done are a current reality like dsp support for snes (admittedly, I have not tried it as I am on a mame kick right now and havn't played snes in months).
I remember at one point people were saying sega-cd emulation was very unlikely, yet most games can be played full speed 44khz sound while underclocking the gp2x.
DSP1 support is actually not a huge deal, I think you're confusing it with the other special chips.
I already told you, it really doesn't matter what uninformed people say about what is and isn't possible. They were wrong. I never would have said that SegaCD was very unlikely, in fact I seem to remember having a conversation with quzar saying exactly that (where he thought that Dreamcast would be able to do it while GP2X wouldn't)
If I am wrong here and I've only been exposed to users point of view and not people who actually do the stuff, then forgive me for being misinformed. OTOH, nobody was denying it at the time
christo930 said:
CPS2 emulation... And as far as home brew stuff goes, I have not heard a single dev complain that their projecct was being held back by technical constraints of the machine, I even asked. I am pretty sure that high computability psx emulation is unrealistic (and anything beyond that) and the gp2x isn't very good at 3d. I may be in the minority here, but I have no interest whatsoever in 3d games (except maybe doom and a couple of others).
Then listen up because I'm complaining about it, I know Squidge and zodttd would too, I'm sure there are many others.
Beyond porting, how many home brewers have the time and artistic talent to make up large complicated 3d games beyond the capabilities of the gp2x? Since Doom runs quite well and from what I hear Quake does as well, are you telling me you need more 3d power than Quake or Doom for a 1 man operation? I may be underestimating you guys, but most modern 3d games takes thousands of man hours to produce. Where are you going to get that kind of talent for free? Sure there are some projects to port, but for new content I thought you need a lot more than a 1 or 3 man team. Now-a-days for 3d games, you need musicians, graphic artists... Do you think (and this is not a rhetorical question) that you can compete with the multi-million dollar efforts that can be bought for more conventional systems?
christo930 said:
Um, doesn't that statement prove you are biased against the gp2x and therefor to some extent biased for the Pandora?
No, it doesn't. Bias doesn't mean having an opinion, it's when your opinion of one thing influences your opinion of another thing. You're against the Pandora on the basis of what you think of GP2X. What I think of GP2X has no weight whatsoever on what I think of Pandora (which I've actually said very little of, I've mostly just been defending it against unfair comments)
I really don't see my comments as unfair. And I can't see how you wanting a system you'd rather develop on doesn't make you biased. Just as my wanting to keep the gp2x around for a while makes me biased against the Pandora (in the short run), your wanting to move on makes you biased for the Pandora. And since you are an active developer around here, I'm betting you were asked what you would like to see in a system, meaning you played a part in the design. How could you not be biased?
Chris