Have they? Would you happen to know what philosophy is behind that? I remember that direct hardware access was the way to go in the early days of the PS2, and the Vector Units seem to support that idea. (NB: I am merely curious, not in any way contesting your statement)
(off-topic) This trend of putting watermark on pictures sadden me a lot. It's growing on the internet. I don't see a reason for watermaking a single picture. There's no harm done if people re-use your pictures. The only harm I could think of would be someone ripping off an entire website (say: you made 300 scans and someone take those for his site which doesn't provide any added value). I'm ok to put a small logo somewhere if it doesn't harm the picture content. I never put watermark on my pictures and hopefully I never will. Putting a watermark on a picture you release is like putting a unremovable stamp on a videogame. It's like those pirates guys modifying a game to replace a logo by their, etc. I see no difference. Not specifically directed to you Yakumo (I appreciate what you're doing), but generally speaking, I feel this internet trend is dangerous. Think about it guys.
No, but I would guess that it makes it easier to tweak the hardware while retaining compatibility. Users are also protected from buggy games setting the WLAN antenna to full power, and then you get the benefit of simulators, as mentioned. I think the PS2 will be the last of the bigger consoles delivered with register level documentation. Both the XBox and Gamecube require developers to use a provided API and I don't think many people miss them.
Well, in fact Sony has once stated that the PS3 will require even more low-level programming, but then that was over a year ago. Frankly, I kind of don't see the point in *prohibiting* programmers from using the hardware directly... granted, there is the option of changing the hardware specs, but how often does that happen? Packaging, integration etc. regularly gets changed but in the past this has led to very few compatibility issues (GB on GBA, PSX on PS2, Nomad, etc.). I think that if programmers want to stay away from low-level programming, they should use an API, but you shouldn't stop them if they want to squeeze out just the tiny extra bit of speed :smt045 I think the PSP's reconfigurable core is a nice example. Details have been sketchy but from what I understand, it allows the programmer to reconfigure execution at a sub-instruction level. A lot of its use would be lost if there's no direct way to do that, only a few predefined templates. <end of rant>