C B Felterbush said:
I can sympathize with you jdh2550, I have been a member of these forums now for a long time and there is one thing I've learned. No matter what you say, no matter how you say it there are people here who are going to twist your words. There are people who won't even read your post, they will just respond to it. This has always been a very volatile forum, people flame and discount what you have to say without regard.
This forum really isn't so bad, there are protective people but most are level headed. Much better than most other forums I've been to.
C B Felterbush said:
Exophase- I must call you a liar, I believe you did mean this sentence to be a way for you to put down jdh2550---It's the developers (are you a developer?) who will determine what it can be used for even more than the ones doing the hardware will.---- this is a bit condescending. You have done this in the past, the suggestion that if he is not a developer then his opinion is not important. It is only because he is that this topic has gone on so long, that and the question he asked got tossed aside.
Calling me a liar is really harsh. I'm sorry to you as well that that's how you interpreted it but that isn't actually what I meant, I can't believe you or anyone else would think that I would so arbitrarily assume that someone isn't a developer while knowing nothing about them. You can believe what you want but I'm telling you that I was ASKING HIM if he was a developer. Maybe I could have phrased it better but whatever, I can see you already have your own preconceived notions about what I have to say anyway.
And I think him being a developer or not has remained completely irrelevant to the responses, which is unfortunate because I really wanted an explanation
from a developer's perspective of view why the RAM is important (and not because a minimum requirement demands it).
Yes, you're right, in the past I have called people out on not being emu devs when they start telling me what is harder and easier to emulate when they don't know the first thing about the console hardware they're referring to, and I don't apologize for that
It's not at all a matter of "only a dev's opinion matters"
C B Felterbush said:
jdh2550 is not questioning craig or michael, but rather asking them a question, and not a bad one at that. If they answered " we decided not to" that would suffice.
Guess that's not the way everyone else took it.
C B Felterbush said:
UMPC, wtf? It is clearly a UMPC among many other things. Is a Powerbook G4 a computer? I owned one, it was my own personal computer. Isn't that what PC means? regardless of who coined or owns the right to say in an advertisement. I agree with jdh2550 the term UMPC should be used more loosely, it is not just a portable device made to run some variation of x86 based software.
Okay, just stop right there and go read this:
http://en.wikipedia.org/wiki/UMPC
I can't claim that Wikipedia is definition by consensus but for a pretty well known term like this it's pretty close.
It's not a
PC, meaning it's not an x86 desktop or laptop computer, that's ALL I was saying! Is this really so difficult to accept? Why are people being so offended by this?
The world has already decided what "PC" means and by extension what UMPC means, you can argue that it's a bad definition but you can't argue that the definition doesn't exist.
C B Felterbush said:
Exophase you are obviously a smart dev. that aside other than having educated guesses into the development of the Pandora, do you think your word into the "what" and "why" is more important than craigs because you are going to develop for the hardware?
No, I never thought that my word was more important than craig's, and I never said it was either. Nor have I said anything that has contradicted anything he has said. However, it's a logical fact that what the devs end up putting on the device will determine what it's used for even more than what the makers of the hardware intended (Sony never intended PSP to be used as a homebrew machine, but that's what a number of people use it for). craigix et al are not just the hardware pushers, but they are devs themselves so they will also contribute to this.
C B Felterbush said:
This is a device suited for many things, of those things many will be brand new. This is not a device suited for similar things, it is a device suited for new things. It is unimportant how much ram the PSP has or the DS- this is neither of those. The objective is to be the sheppard not another sheep.
So basically what I'm getting here is that I'm this giant condescending asshole for saying what I think the device will end up being used for, whereas you're allowed to define its purpose...
Sorry, but all of your retorts are really slanted because you're unhappy with how people's vision of Pandora doesn't meld with yours.
jdh2550 said:
On another board I saw Exophase mention that you can't create a virtual machine for an ARM. Can someone point me at a good resource to understand that?
It's not really that clear cut, it's more that in traditional ARMs you can't create a VM that fulfills all of the Popek & Goldberg requirements, whether or not this is a problem depends on what you're trying to virtualize. A big problem might be cache, if what you're virtualizing doesn't have it and you do there can be coherency issues (this is something that comes up if you're trying to virtualize say, GBA on GP32 or GP2X).
ARM has actually added "TrustZone" to their chips which does allow for true virtualization as of ARMv6 (for the ones that support it, and I believe Cortex-A8 does). This doesn't solve the coherency problem, but it's no longer really relevant at that point.
There is one more issue - when you virtualize you have to handle exceptions externally and not in software, which usually means using hardware to help emulate them. This gives you a lot less speed control, so you can't nearly as reliably unthrottle execution to let it run as fast as possible, and if the environment virtualizing isn't fast enough then it can break everything as opposed to just making it slower (this can happen even for a sudden spike in background activity, so it has pretty strict realtime requirements).