| [Files] See AmigaOS4 in action | ANN.lu |
|
| See AmigaOS4 in action : Comment 101 of 269 | ANN.lu |
| Posted by Phill on 22-Sep-2003 12:34 GMT | In reply to Comment 74 (Fabio Alemagna):
If you've ever worked on something in a nice safe environment and then you are faced with moving it to a more hostile one where you'll need to write loads of untested code that will be almost impossible to debug properly, then you'll know why they have waited until now to get it booting on the AmigaOne.
Personally I either start with the real hardware or wait until the end. Introducing it in the middle is a bit of a nightmare.
They probably felt they had to do it now though, to try to stop people making stupid claims. I would imagine they are frustrated enough about how long it has taken them to get this far, without people bitching all the time.
FWIW they are not the only people in the world for things to take longer than they hoped. Windows 2000 slipped by three years & more than half of the things planned were scrapped. Some things eventually showed up in XP/2003 & some may end up in longhorn ( which is something like two years away ).
I'm more shocked when software turns up on time.
Phill |
|
| See AmigaOS4 in action : Comment 102 of 269 | ANN.lu |
| Posted by Amon_Re on 22-Sep-2003 12:34 GMT | In reply to Comment 80 (Mike Bouma):
Then you can't be using windows ;)
Cheers |
|
| See AmigaOS4 in action : Comment 103 of 269 | ANN.lu |
| Posted by Fabio Alemagna on 22-Sep-2003 12:39 GMT | In reply to Comment 100 (Bernie Meyer):
> No, Fabio, I have gone one better --- I have seen for myself what a difference
> it makes whether you implement CopyRectNoMaskComplete in a P96 driver or not :)
> I don't know how AROS works its screen, and whether there is any reading from
> graphics memory involved
Well, if you want to copy stuff from one place to another, then of course you need to read memory. The VESA driver works like that.
It's indeed _slower_ than the VGA16 driver, which also needs to do c2p conversion, and that is most likely because the VGA16 driver keeps a chunky buffer internally, which then it blits to the gfx memory after doing the c2p conversion, which means that it doesn't read from gfx memory. So, yes, you've got a point there. |
|
| See AmigaOS4 in action : Comment 104 of 269 | ANN.lu |
| Posted by Rassilon on 22-Sep-2003 12:41 GMT | In reply to Comment 96 (Mike Bouma):
I think some people are underestimating the significance of this event.
Many have said why did it take so long to get AOS4 to boot on the A1, well the simple fact is it probably booted ages ago, but without the recent work done on the gfx library you wouldn't be able to see it :o)
Hyperion also said that the gfx library used was not only emulated, but was a hacked version of the original with *just* enough changes so that it will work with P96. In fact as stated they had to write a minmal chipset emulation so that the hacked gfx.library got the expected return values etc. So no wonder it was on the slow side. The whole reason for doing the quick gfx.lib hack was because seomeone needed to show AO4 booting on the A1 for a meeting.
Hyperion have been working on a proper gfx.library replacement for a while now, but it is not quite ready yet.
Rassilon |
|
| See AmigaOS4 in action : Comment 105 of 269 | ANN.lu |
| Posted by Peter Gordon on 22-Sep-2003 12:43 GMT | In reply to Comment 85 (Lando):
>Hi, well I can't comment on resizing as I wasn't using opaque resize on the
>A4k, but I mean actually the redrawing and the response time seemed very slow
Thats my point; you were using a more mature P96 driver that makes good use of the onboard blitter, rubber band sizing and moving, ReAction and intuition without the gradients and other eyecandy, and saying that OS4 is running slower than an 030.
I'd like to see you try opaque moving & sizing, and gradients on an 030+voodoo. It wouldn't be pretty.
I just wish people would look at things in perspective, is all. |
|
| See AmigaOS4 in action : Comment 106 of 269 | ANN.lu |
| Posted by Fabio Alemagna on 22-Sep-2003 12:44 GMT | In reply to Comment 101 (Phill):
> If you've ever worked on something in a nice safe environment and then you are
> faced with moving it to a more hostile one where you'll need to write loads of
> untested code that will be almost impossible to debug properly, then you'll
> know why they have waited until now to get it booting on the AmigaOne.
Hum, let's see... like writing an OS from scratch? D'oh :)
> FWIW they are not the only people in the world for things to take longer than
> they hoped. Windows 2000 slipped by three years & more than half of the things
> planned were scrapped. Some things eventually showed up in XP/2003 & some may
> end up in longhorn ( which is something like two years away ).
Ehm, so the point about comparing apples with oranges didn't even have the pleasure to be spotted by you, right?
Sorry to be sarcastic, but I couldn't find any better way to express myself :) |
|
| See AmigaOS4 in action : Comment 107 of 269 | ANN.lu |
| Posted by Phill on 22-Sep-2003 12:48 GMT | In reply to Comment 87 (Alkis Tsapanidis):
> will stand with Fabio in that one, my sister had been using UAE on the PC I
>use right now, which is a 700MHz Celeron, before UAE JiT was released
Arguing about winuae speeds is pretty useless. For a start you need to include all the configuration options & all the software you are using.
To test the cpu emulation speed you need to be using p96 & no sound emulation as otherwise the overhead of all the graphics chips is too much.
With an interpreter on a 700mhz pc, I would imagine you'll get something around the speed of a 33mhz 030. Whether you think this is fast or not depends on what you do.
Comparing the speed to UAE on Linux is silly, because most of the development has only been made on WinUAE for a long time now.
I like the comment about UAE on Linux being more responsive than Linux. Which is obviously more of a damning comment about Linux than anything else. I'm just glad I run WindowsXP.
Phill |
|
| See AmigaOS4 in action : Comment 108 of 269 | ANN.lu |
| Posted by Fabio Alemagna on 22-Sep-2003 12:55 GMT | In reply to Comment 107 (Phill):
> Comparing the speed to UAE on Linux is silly, because most of the development
> has only been made on WinUAE for a long time now.
Gah, you don't want to say that, believe me, because that has not even the slightest relevance, for two main reasons:
1) CPU emulation has always been developed on the linux version first
2) If the gfx is fast under linux, it can be only faster under windows, because THAT is something which has been mainly improved for WinUAE. |
|
| See AmigaOS4 in action : Comment 109 of 269 | ANN.lu |
| Posted by Phill on 22-Sep-2003 12:56 GMT | In reply to Comment 106 (Fabio Alemagna):
>Hum, let's see... like writing an OS from scratch? D'oh :)
Oh, what Operating System have you personally brought up from scratch?
>Ehm, so the point about comparing apples with oranges didn't even have the
> pleasure to be spotted by you, right?
Apples and Oranges are both fruit, AOS4 & Windows XP are both operating systems. Apart from a classic strawman argument ( from you? never ) what is your point?
Phill |
|
| See AmigaOS4 in action : Comment 110 of 269 | ANN.lu |
| Posted by Georg Steger on 22-Sep-2003 12:57 GMT | In reply to Comment 91 (Fabio Alemagna):
>> However, if it is *not* done by the graphics card's blitter, then the CPU
>> typically needs to *read* from graphics memory. And *that* is slow as hell, no
>> matter how fast the (emulated) CPU is.
>
> Ever tried AROS with its vesa (or even VGA16!!) driver? Quite fast, I must say :)
Maybe that's because AROS vesa driver does not read from gfx memory at all ;-)
(early versions did). |
|
| See AmigaOS4 in action : Comment 111 of 269 | ANN.lu |
| Posted by Phill on 22-Sep-2003 12:58 GMT | In reply to Comment 108 (Fabio Alemagna):
>Gah, you don't want to say that, believe me, because that has not even the >slightest relevance, for two main reasons:
>1) CPU emulation has always been developed on the linux version first
>2) If the gfx is fast under linux, it can be only faster under windows,
> because THAT is something which has been mainly improved for WinUAE.
I'll ignore for the moment that you seem to know nothing about emulation & point out that most of the time improving it makes it slower.
Phill |
|
| See AmigaOS4 in action : Comment 112 of 269 | ANN.lu |
| Posted by Fabio Alemagna on 22-Sep-2003 13:00 GMT | In reply to Comment 109 (Phill):
> >Hum, let's see... like writing an OS from scratch? D'oh :)
>
> Oh, what Operating System have you personally brought up from scratch?
Do I really need to answer to that?
> >Ehm, so the point about comparing apples with oranges didn't even have the
> > pleasure to be spotted by you, right?
> Apples and Oranges are both fruit, AOS4 & Windows XP are both operating
> systems. Apart from a classic strawman argument ( from you? never ) what is
> your point?
My point? I already expressed it, but you don't care to read, which is not my fault. Do your job properly and read what you reply to.
Phill |
|
| See AmigaOS4 in action : Comment 113 of 269 | ANN.lu |
| Posted by Hagge on 22-Sep-2003 13:02 GMT | NOOOO... the search path for my statistics script was bbclonestats/ and not /bbclonestats so i haven't got any statistics for all ann.lu hits on my webpage :/
Now you all have to visit my page again! =D==D=D=D
This comment is very important, so you have to read it :) |
|
| See AmigaOS4 in action : Comment 114 of 269 | ANN.lu |
| Posted by Fabio Alemagna on 22-Sep-2003 13:04 GMT | In reply to Comment 111 (Phill):
> >1) CPU emulation has always been developed on the linux version first
> >2) If the gfx is fast under linux, it can be only faster under windows,
> > because THAT is something which has been mainly improved for WinUAE.
> I'll ignore for the moment that you seem to know nothing about emulation
LOL, says who? :-)
> & point out that most of the time improving it makes it slower.
Riiight, I must be so stupid! For whatever else reason would one _improve_ picasso emulation, if NOT for making it *worse*? |
|
| See AmigaOS4 in action : Comment 115 of 269 | ANN.lu |
| Posted by Fabio Alemagna on 22-Sep-2003 13:07 GMT | In reply to Comment 113 (Hagge):
> NOOOO... the search path for my statistics script was bbclonestats/ and not
> /bbclonestats so i haven't got any statistics for all ann.lu hits on my webpage
> :/
Wrong choice for your stats script, my friend, bbclone can't bear up with high traffic at all, its dbase goes nuts. It can bear about 2k visitors a day, though. |
|
| See AmigaOS4 in action : Comment 116 of 269 | ANN.lu |
| Posted by Anonymous on 22-Sep-2003 13:07 GMT | In reply to Comment 112 (Fabio Alemagna):
> Oh, what Operating System have you personally brought up from scratch?
> Do I really need to answer to that?
Yes please. What operating system have you personally brought up on real hardware?
>My point? I already expressed it, but you don't care to read, which is not my
> fault. Do your job properly and read what you reply to.
I did read it, I just didn't agree with it. Microsoft aren't doing anything really different, just different colours and size. They have more people working on a product that does more. I would imagine that growing apples is similar to growing oranges, but I'm not a fruit expert.
Phill |
|
| See AmigaOS4 in action : Comment 117 of 269 | ANN.lu |
| Posted by Alkis Tsapanidis on 22-Sep-2003 13:08 GMT | In reply to Comment 92 (Mike Bouma):
Ah, thanx for showing that you are a worse troll than I... :-)
BTW, if you ever put me on the same line as Eva again, I'll eat you alive:-) |
|
| See AmigaOS4 in action : Comment 118 of 269 | ANN.lu |
| Posted by Phill on 22-Sep-2003 13:14 GMT | In reply to Comment 114 (Fabio Alemagna):
> > I'll ignore for the moment that you seem to know nothing about emulation
> LOL, says who? :-)
I did, are losing your short term memory?
>> & point out that most of the time improving it makes it slower.
>Riiight, I must be so stupid! For whatever else reason would one _improve_
>picasso emulation, if NOT for making it *worse*?
I was under the impression there was a bit more to UAE than picasso emulation. They seem to talk about loads of other things in the release notes.
If you've ever coded an emulator that tries to be compatible, you'll know it usually ends up slower than one that doesn't. The UAE vs Amithlon situation shows this best.
Phill |
|
| See AmigaOS4 in action : Comment 119 of 269 | ANN.lu |
| Posted by Fabio Alemagna on 22-Sep-2003 13:17 GMT | In reply to Comment 116 (Anonymous):
> > > Oh, what Operating System have you personally brought up from scratch?
> > Do I really need to answer to that?
>
> Yes please. What operating system have you personally brought up on real
> hardware?
1) You are just trying to catch me in some kind of traps, but I'm not stupid so I'll let you reply to your question yourself
2) "Real HW"? Care to elaborate?
> > My point? I already expressed it, but you don't care to read, which is not
> > my fault. Do your job properly and read what you reply to.
> I did read it, I just didn't agree with it. Microsoft aren't doing anything
> really different, just different colours and size. They have more people
> working on a product that does more.
Having more people doesn't necessarily mean going faster, it can be quite the opposite, depending on _what_ you do. Windows does a _lot_ more than AOS, and does it quite differently too, not to mention that there are often strategic reasons behind delays, not always technical ones.
> I would imagine that growing apples is
> similar to growing oranges, but I'm not a fruit expert.
You couldn't imagine anything more wrong. Perhaps not "apples" and "oranges", but not all kind of fruits grow the same way.
Phill |
|
| See AmigaOS4 in action : Comment 120 of 269 | ANN.lu |
| Posted by ikir on 22-Sep-2003 13:19 GMT | In reply to Comment 4 (Eva):
eva, it was tested with a xxxxxx pci voodoo driver, due to lack of time. It is IMPRESSIVE FAST CONSIDERING THE EMULATION OF THE DEMO; THE EMULATION OF THE VIDEO DRIVER, and that the driver was for PCI not AGP!
Infact you are a troll, because i tried the matrix trailer on the pegasos machine and the refresh was slow. Stop trolling... FOREVER. |
|
| See AmigaOS4 in action : Comment 121 of 269 | ANN.lu |
| Posted by Anonymous on 22-Sep-2003 13:22 GMT | In reply to Comment 120 (ikir):
Is it true OS4 was crashing liek hell? |
|
| See AmigaOS4 in action : Comment 122 of 269 | ANN.lu |
| Posted by Don Cox on 22-Sep-2003 13:22 GMT | In reply to Comment 50 (JoannaK):
"But. For people who have dying or Dead Amigas... Sad fact is that
OS4/aone is not even near (months) from becoming suitable replacement
for their old computer. And to them, it's obviously a big
dissappointment to see this in real life after all those big talks
about it's readiness."
OTOH those who already own AmigaOnes would probably be happy to get a beta version just to have something to play with and test. |
|
| See AmigaOS4 in action : Comment 123 of 269 | ANN.lu |
| Posted by Fabio Alemagna on 22-Sep-2003 13:22 GMT | In reply to Comment 118 (Phill):
> > LOL, says who? :-)
> I did, are losing your short term memory?
Smart reply, I congratulate with you, really, I'm astonished.
But perhaps it would be better if you told me what knowledge you have about emulators and about ME to state what you just stated. Do one thing, go to the official UAE page, scroll it a bit, and then tell me if you noticed anything "strange".
> >Riiight, I must be so stupid! For whatever else reason would one _improve_
> >picasso emulation, if NOT for making it *worse*?
> I was under the impression there was a bit more to UAE than picasso emulation.
The "more" is not what we're talking about. Besides, the "more" is just shut off when picasso is on. As said, irrelevant.
> If you've ever coded an emulator that tries to be compatible, you'll know it
> usually ends up slower than one that doesn't. The UAE vs Amithlon situation
> shows this best.
Again, that's irrelevant. |
|
| See AmigaOS4 in action : Comment 124 of 269 | ANN.lu |
| Posted by dammy on 22-Sep-2003 13:23 GMT | In reply to Comment 116 (Anonymous):
> Yes please. What operating system have you personally brought up on real
> hardware?
Phill, why are you trolling?
Dammy |
|
| See AmigaOS4 in action : Comment 125 of 269 | ANN.lu |
| Posted by Don Cox on 22-Sep-2003 13:28 GMT | In reply to Comment 79 (Anonymous):
"Sorry if I seem to be being rude, but I can't stand when people dry and dumb down this fantastic achievment - we have the OFFICIAL, ORIGINAL AmigaOS, or at least an evolution of it, running on non-legacy MODERN PPC HARDWARE! Hello people - this is an historical step!"
Correct. |
|
| See AmigaOS4 in action : Comment 126 of 269 | ANN.lu |
| Posted by Anonymous on 22-Sep-2003 13:28 GMT | In reply to Comment 119 (Fabio Alemagna):
> Yes please. What operating system have you personally brought up on real
> hardware?
>1) You are just trying to catch me in some kind of traps, but I'm not stupid
> so I'll let you reply to your question yourself
No. You didn't read my original post carefully, before you suggested that you had.
> 2) "Real HW"? Care to elaborate?
Sorry, HW is an acronym for hardware.
> Having more people doesn't necessarily mean going faster, it can be quite the
> opposite, depending on _what_ you do. Windows does a _lot_ more than AOS,
> and does it quite differently too, not to mention that there are often
> strategic reasons behind delays, not always technical ones.
Software is delayed for two reasons:
1) it's taking longer than expected to get it working.
2) the feature list changes.
I can't imagine any sane software house delaying a shrink wrapped application for any other reason.
>You couldn't imagine anything more wrong. Perhaps not "apples" and "oranges",
> but not all kind of fruits grow the same way.
But we were comparing Apples and Oranges. You can't just throw other fruit into the argument, I didn't say I thought that Apples grew the same as Bananas or even Pineapple.
Phill |
|
| See AmigaOS4 in action : Comment 127 of 269 | ANN.lu |
| Posted by Phill on 22-Sep-2003 13:31 GMT | In reply to Comment 124 (dammy):
>> Yes please. What operating system have you personally brought up on real
>> hardware?
> Phill, why are you trolling?
I wasn't. Fabio jumped on a reasonable post with one of his "I know everything" posts.
Phill |
|
| See AmigaOS4 in action : Comment 128 of 269 | ANN.lu |
|
| See AmigaOS4 in action : Comment 129 of 269 | ANN.lu |
| Posted by Phill on 22-Sep-2003 13:36 GMT | In reply to Comment 123 (Fabio Alemagna):
>Smart reply, I congratulate with you, really, I'm astonished.
Thank you.
>But perhaps it would be better if you told me what knowledge you have about
> emulators and about ME to state what you just stated. Do one thing, go to
> the official UAE page, scroll it a bit, and then tell me if you noticed
> anything "strange".
Do you mean the couple of patches you made to UAE?
>> I was under the impression there was a bit more to UAE than picasso emulation.
>The "more" is not what we're talking about. Besides, the "more" is just shut
> off when picasso is on. As said, irrelevant.
Picasso disables sound?
>> If you've ever coded an emulator that tries to be compatible, you'll know
>> it
>> usually ends up slower than one that doesn't. The UAE vs Amithlon situation
>> shows this best.
>Again, that's irrelevant.
Why? My original post said you can't compare WinUAE speed to UAE to show how good the CPU emulation is. The fact they are based on different codebases is a good start, there were also some really bad bugs in WinUAE which caused it to run ultra slow in some cases.
Phill |
|
| See AmigaOS4 in action : Comment 130 of 269 | ANN.lu |
| Posted by Fabio Alemagna on 22-Sep-2003 13:38 GMT | In reply to Comment 125 (Don Cox):
> Correct.
I beg to differ. Hystory has much more important things to care about. |
|
| See AmigaOS4 in action : Comment 131 of 269 | ANN.lu |
| Posted by Phill on 22-Sep-2003 13:40 GMT | In reply to Comment 128 (Alkis Tsapanidis):
>If you aren't trolling, tell me what you see in www.aros.org... :-)
Ok, here is my post:
"If you've ever worked on something in a nice safe environment and then you
are faced with moving it to a more hostile one where you'll need to write
loads of untested code that will be almost impossible to debug properly,
then you'll
know why they have waited until now to get it booting on the AmigaOne. "
Here is Fabio's reply:
"Hum, let's see... like writing an OS from scratch? D'oh :) "
Please can you show me where on aros.org it says that Fabio has written an OS from scratch. He is a member of the AROS team & does write software in a nice hosted environment where you can easily debug code, much like the way that alot of the AOS4 code has been written up until now. Michal Shulz is the only person that has done the type of work that I was talking about ( first with the PC version and now PPC ).
Phill |
|
| See AmigaOS4 in action : Comment 132 of 269 | ANN.lu |
| Posted by darklite on 22-Sep-2003 13:44 GMT | In reply to Comment 16 (Ben Hermans/Hyperion):
>If you think this will take until 2004 or 2005 to convert it, you are once again
>making a fool out of yourself Eva.
Pot, kettle, black.
Remember the announced february release date? :) |
|
| See AmigaOS4 in action : Comment 133 of 269 | ANN.lu |
| Posted by Fabio Alemagna on 22-Sep-2003 13:44 GMT | In reply to Comment 126 (Anonymous):
> No. You didn't read my original post carefully, before you suggested that you
> had.
Sorry? What is that I haven't read carefully?
> > 2) "Real HW"? Care to elaborate?
> Sorry, HW is an acronym for hardware.
Oh, i know, since *I* used it, you used "hardware". So, you're not trolling, uh?
> > Having more people doesn't necessarily mean going faster, it can be quite
> > the opposite, depending on _what_ you do. Windows does a _lot_ more than
> > AOS, and does it quite differently too, not to mention that there are often
> > strategic reasons behind delays, not always technical ones.
> Software is delayed for two reasons:
> 1) it's taking longer than expected to get it working.
> 2) the feature list changes.
> I can't imagine any sane software house delaying a shrink wrapped application
> for any other reason.
Imagine you have a monopoly, imagine people are still buing your old version of the operating system, imagine that that still generates a big cash flow. Now, introducing another version of the operating system on the maket would surely lead people to NOT buy the old version anymore, but rather the NEW one, which means that you don't gain anything.
That may be another reason for delaying stuff. There are plenty of others, I let you do your homework.
> >You couldn't imagine anything more wrong. Perhaps not "apples" and "oranges",
> > but not all kind of fruits grow the same way.
> But we were comparing Apples and Oranges.
Oh dear... are you serious? Would you be happy if I said "babanas and strawberries"? You're quite amusing, you know :) |
|
| See AmigaOS4 in action : Comment 134 of 269 | ANN.lu |
| Posted by Fabio Alemagna on 22-Sep-2003 13:46 GMT | In reply to Comment 127 (Phill):
> I wasn't. Fabio jumped on a reasonable post with one of his "I know everything"
> posts.
I don't know everything, I never claimed I do, it's YOU who assumed that I DON'T know something, and I just happen to know that thing. If it disappoints you, it's your problem. |
|
| See AmigaOS4 in action : Comment 135 of 269 | ANN.lu |
| Posted by dammy on 22-Sep-2003 13:47 GMT | In reply to Comment 127 (Phill):
>> Phill, why are you trolling?
>
> I wasn't. Fabio jumped on a reasonable post with one of his "I know
> everything" posts.
Your evidently quiet content to continue to inflame the thread, that's trolling.
Dammy |
|
| See AmigaOS4 in action : Comment 136 of 269 | ANN.lu |
| Posted by Fabio Alemagna on 22-Sep-2003 13:49 GMT | In reply to Comment 129 (Phill):
> Do you mean the couple of patches you made to UAE?
No, I mean the webpage.
> > The "more" is not what we're talking about. Besides, the "more" is just shut
> > off when picasso is on. As said, irrelevant.
> Picasso disables sound?
Nitpicking, aren't you? No picasso doesn't disable sound, but I keep sound disabled both on UAE for linux and Windows, so that is not an issue for me when comparing WinUAE and UAE. Content?
>Again, that's irrelevant.
> Why? My original post said you can't compare WinUAE speed to UAE to show how
> good the CPU emulation is.
In the correct conditions you can. Period. |
|
| See AmigaOS4 in action : Comment 137 of 269 | ANN.lu |
| Posted by Fabio Alemagna on 22-Sep-2003 13:53 GMT | In reply to Comment 131 (Phill):
> Please can you show me where on aros.org it says that Fabio has written an OS
> from scratch. He is a member of the AROS team & does write software in a nice
> hosted environment where you can easily debug code, much like the way that alot
> of the AOS4 code has been written up until now. Michal Shulz is the only person
> that has done the type of work that I was talking about ( first with the PC
> version and now PPC ).
Which doesn't mean that I *don't* know what does it mean. You know, we are quite communicative in the AROS Dev team, and even if Michal is the one who more often bangs the HW, it doesn't mean that I've not done it as well. Again, you assume things that you have no knowledge about. |
|
| See AmigaOS4 in action : Comment 138 of 269 | ANN.lu |
| Posted by darklite on 22-Sep-2003 13:56 GMT | In reply to Comment 86 (Mike Bouma):
>I don't think there are only need to be two options, just like we don't only see
>black and white colors. Let's assume that we both state is the truth, then likely
>the difference must be caused by differences in OS and emulator versions. For
>instance I use WinUAE with WinXP and you use Linux/UAE, you use AmigaOS3.5 and I
>use AmigaOS3.9-2.
Sorry Mike, I used to run WinUAE (non-JIT) on Win2K alot a long time ago. It was perfectly useable (fast even) on my Duron 800, in high-resolution P96 modes. And I doubt OS3.9 is that much slower than OS3.0.
Either you are incapable of properly configuring WinUAE or the new WinUAE versions are significantly slower when JIT is disabled. |
|
| See AmigaOS4 in action : Comment 139 of 269 | ANN.lu |
| Posted by Bernie Meyer on 22-Sep-2003 13:58 GMT | In reply to Comment 103 (Fabio Alemagna):
And at least in some cases of solid window movement (most prominently when moving the OS 3.9 MP3 player partially behind other windows), the source of the "reading from somewhere" is *not* the graphics memory you see, but rather an off-screen bitmap. And that off-screen bitmap can be in main memory (when you have no blitter), or in graphics memory (when you *do* have a blitter). However, if you have a blitter, but only support it partially, the bitmap is in the card's memory, but is being read by the CPU.... |
|
| See AmigaOS4 in action : Comment 140 of 269 | ANN.lu |
| Posted by Phill on 22-Sep-2003 14:00 GMT | In reply to Comment 133 (Fabio Alemagna):
>> > 2) "Real HW"? Care to elaborate?
>> Sorry, HW is an acronym for hardware.
>Oh, i know, since *I* used it, you used "hardware". So, you're not trolling, >uh?
No I just thought I'd used it and you didn't understand. I didn't go back to my own post. Now with that misunderstanding is out of the way, real hardware means:
1) not hosted.
2) not emulated.
So you have to make sure all the chips are initialised properly etc, and of course not being able to run the code in gdb. The reason why you don't want to change environments half way through is because unless you don't want to have too many untrustworthy things. If you try to bring up an operating system where nothing is trusted then you'll never know where to look for bugs.
>Imagine you have a monopoly, imagine people are still buing your old version
> of the operating system, imagine that that still generates a big cash flow.
> Now, introducing another version of the operating system on the maket would
> surely lead people to NOT buy the old version anymore, but rather the NEW
> one, which means that you don't gain anything.
So you are saying there is a software house that would announce a piece of software is going to be released on a certain date & then delay it, to get more sales on the old version?
>That may be another reason for delaying stuff. There are plenty of others, I >let you do your homework.
No it's because you can't think of anything else, but admitting that wouldn't help your argument, now much would it.
>Oh dear... are you serious? Would you be happy if I said "babanas and
> strawberries"? You're quite amusing, you know :)
No I wasn't serious, but I'm not going to fall for your straw man argument. Of course I'm not going to say that Apples and Oranges are exactly the same, but that was your main point for suggesting that AOS4 and Longhorn had nothing in common either. So if I admit that Apples and Oranges are in fact different then you would probably say something like "LOL see I was right".
Phill |
|
| See AmigaOS4 in action : Comment 141 of 269 | ANN.lu |
| Posted by Raffaele on 22-Sep-2003 14:06 GMT | In reply to Comment 126 (Anonymous):
An Anonymous wrote:
>But we were comparing Apples and Oranges.
>You can't just throw other fruit into the argument,
>I didn't say I thought that Apples grew the same
>as Bananas or even Pineapple.
Well Apples and Oranges are not the same, or a "computer firm" will be the same as an "english telephone provider"
:->
<LOL>
But yes, you can mix apples, bananas and oranges into a big FRUIT SALAD!
AND
When Speaking of OSes...
Think about the fact that Windows 95 was intended to be released as:
Windows 85
(Not to mention the fact it needed over 40 coders hired by M$ at full working-time daily & "full pay packet" AND 400 registered betatesters)
What a loooong development duration time...
And how slooooow the programmers were...
;-))))))))))))))))))))))))))))
<STRALOL> |
|
| See AmigaOS4 in action : Comment 142 of 269 | ANN.lu |
| Posted by Phill on 22-Sep-2003 14:10 GMT | In reply to Comment 137 (Fabio Alemagna):
>Which doesn't mean that I *don't* know what does it mean. You know, we are
> quite communicative in the AROS Dev team, and even if Michal is the one who
> more often bangs the HW, it doesn't mean that I've not done it as well.
> Again, you assume things that you have no knowledge about.
No it doesn't mean you haven't done it as well. Which is why when you replied, I actually asked you what OS you had written from scratch. To be honest, talking to someone that has done it really doesn't count as doing it. It's like the difference between parachuting out of a plane and hearing about it from a friend. So really unless you've done it yourself, then you don't know.
Phill |
|
| See AmigaOS4 in action : Comment 143 of 269 | ANN.lu |
| Posted by hooligan/dcs on 22-Sep-2003 14:17 GMT | In reply to Comment 141 (Raffaele):
What the HELL does STRALOL mean? :) |
|
| See AmigaOS4 in action : Comment 144 of 269 | ANN.lu |
| Posted by samface on 22-Sep-2003 14:22 GMT | In reply to Comment 138 (darklite):
Argh! For christ sake, just drop it and compare once the product is released, ok? I mean, I'm quite sure there is alot more to things than just wether it's JIT or not. For example, doesn't Picasso96 for WinUAE make use of Windows' 2D drivers while this AmigaOS4.0 beta hasn't any proper 2D drivers yet and runs with a simple framebuffer device or something? I don't know, but I'm quite sure it's not as simple as comparing with WinUAE speeds in order to determine wether AmigaOS4 68k emulation is slow or not (or whatever you might want to conclude by this comparison).
Then of course there is also the debug output. I remember when people complained at MorphOS speeds when displayed in it's early stages, you know... |
|
| See AmigaOS4 in action : Comment 145 of 269 | ANN.lu |
| Posted by Raffaele on 22-Sep-2003 14:22 GMT | In reply to Comment 143 (hooligan/dcs):
Whoops... Excuse me Hooligan...
STRALOL is intended to be "half italianized" in italian computer jergon.
I must had used:
XTRALOL
(More understandable now, yup?) |
|
| See AmigaOS4 in action : Comment 146 of 269 | ANN.lu |
| Posted by Jürgen Schober on 22-Sep-2003 14:25 GMT | In reply to Comment 90 (Jacek Piszczek):
>One thing one could do very quickly is patching workbench (or intuition) to do
>LOL. Could you please learn how refreshing works before trying to brainwash
>people here? Smartrefresh can only speed things up when you have very few
Hm, guess I don't do any brainwashing here, but...
>windows on screen and none of them are covering each other. Once you get more
>windows (>10 big windows) smart refresh becomes really slow because you need to
>do a lot of syncing with offscreen smart bitmaps and screen display. This
>involves plenty of reads from graphics card memory which is always slow (you
>can check this easily by comparing gfx mem read and write speed tests).
..might be that I am wrong. But I doubt it. With AGP/PCI and fast blitter support a memtransfer is hardly slower then a bunch of refresh messages and redraw commands calling a (68k implementation?) of reaction gadgets.
Maybe that's true for original Amiga systems with real slow bus bandwidth, but I doubt it is for the AmigaOne. Though as has been mentioned, if the blitter isn't used to "SwapPixels", then this slows down the whole thing a lot.
It will, too, if you have a lot of open window and Offscreen video memory is running out, and you have to buffer it in "FastRAM". Rearranging bitmaps can help...guess why it slows down on big windows...
>Anyway... smart won't help you at all when opaque resizing windows, it will
>only slow down things A LOT because there's no clipping when refreshing window
>contents and the covered parts of a window will have to be redrawn to smart >bitmap.
?? You mean the window which gets resized ? True, you'd have to redraw to the new size every time. But this does not apply to the background defective rectangle. But maybe someone could enlight me how the refresh is done in the current implementation...might be that I am totaly wrong. |
|
| See AmigaOS4 in action : Comment 147 of 269 | ANN.lu |
| Posted by Mike Bouma on 22-Sep-2003 14:25 GMT | In reply to Comment 138 (darklite):
@ darklite
> And I doubt OS3.9 is that much slower than OS3.0.
As I stated above, AmigaOS3.1 is running here faster with JIT disabled than AmigaOS3.9-2, but I nearly always use AmigaOS3.9-2 jmyself on my system.
But of course the bottom line is that JIT emulation is ALOT faster than interpreted emulation. And that PPC native code is ALOT faster than JIT emulated code.
My mention of "nearly unusable" may mean different things to different people. |
|
| See AmigaOS4 in action : Comment 148 of 269 | ANN.lu |
| Posted by Phill on 22-Sep-2003 14:25 GMT | In reply to Comment 136 (Fabio Alemagna):
>No, I mean the webpage.
Well you'll have to point out the strange thing I'm looking for then.
>> Why? My original post said you can't compare WinUAE speed to UAE to show how
>> good the CPU emulation is.
> In the correct conditions you can. Period.
But I was replying to comparisons being made between WinUAE & UAE that were on different machines. There wasn't any mention of which version was being used & what options were used. As the codebase has split & the options you use have such an impact on performance that it would become very difficult ( or even impossible ) to get the correct conditions to do some comparisons.
Phill |
|
| See AmigaOS4 in action : Comment 149 of 269 | ANN.lu |
| Posted by Fabio Alemagna on 22-Sep-2003 14:45 GMT | In reply to Comment 148 (Phill):
> >No, I mean the webpage.
> Well you'll have to point out the strange thing I'm looking for then.
Come on! Of course I was referring to those patches!
> In the correct conditions you can. Period.
> But I was replying to comparisons being made between WinUAE & UAE that were on
> different machines. There wasn't any mention of which version was being used &
> what options were used. As the codebase has split & the options you use have
> such an impact on performance that it would become very difficult ( or even
> impossible ) to get the correct conditions to do some comparisons.
It doesn't matter, because in those conditions even comparing WinUAE vs WinUAE is meaningless. Your point is just moot, please realize it and move on. |
|
| See AmigaOS4 in action : Comment 150 of 269 | ANN.lu |
| Posted by Phill on 22-Sep-2003 14:49 GMT | In reply to Comment 134 (Fabio Alemagna):
>I don't know everything, I never claimed I do, it's YOU who assumed that I
> DON'T know something, and I just happen to know that thing. If it
> disappoints you, it's your problem.
I was led to believe you didn't know about emulation, by the comments you were making. I still stand by my decision & if you don't like it then thats your problem.
Phill |
|
|
- User Menu
-
- About ANN archives
- The ANN archives is powered by #AmigaZeux. It was updated daily (news last: 22-Oct-2004; comments last: 18-May-2005).
ANN.lu was created, previously owned and maintained by Christian Kemp, www.ckemp.com.
- Contribute
- Not possible at this time!
- Search ANN archives
- Advanced search
- Hosting
- ANN.lu was hosted by Dreamhost. Sign up through this link, mention "ckemp" as referrer and he will get a 10% commission on any account you purchase.
Please show your appreciation for any past, present and future work on ANN.lu by making a contribution via PayPal.
|
