@utri007

Quote:

utri007 wrote: @michalsc 400mhz machine can't decode video on the fly or at least it need help fron GPU. I had a 466mhz celeron wich I used to watc DVDs, it was possible because it had a ATI graphics card wich had a video codec. I've Asus EEEKeyboard, it is 1,2 mhz dual atom cpu, it has problems with video, because it video codec isn't supported generally.

Atom is the new Celeron, worse than the first time Performance of it absolutely sucks but it is intended for low power consumption.

I don't think Atoms should be compared to real CPU's...

@olegil

Quote:

olegil wrote: @Cod3r Put up, or shut up. I'm getting mighty tired of your snide remarks at our intelligence levels here. Show us an implementation of a mainstream CPU running at 400MHz that will happily decode a video using h.264 in HD (720p or better) realtime. Or alternatively, accept that you are not talking about currently used codecs. Or just bugger off. Thank you in advance. Yes, I can probably also find a codec for MY encodings that I could later use on MY decoding hardware at 400MHz but how the FOUR LETTER WORD does that help with browsing youtube?

Ok man whatever you say... i'll stop my full time commitments and make some software to prove to you that I know what I mean... not.

Even if I did you won't do or say anything different. All you will do is find the next thing to cry about and "can't be done without (fill in the blank)"

If you are such an expert, where is YOUR code? Where are YOUR videos showing your abilities? Huh huh? PUT UP OR SHUT UP.

No point at all to waste my time proving this. You prove that you can do at least HALF of what I can then we'll talk...

@Cod3r

AmigaOS has a datatype system. In the days of OS3.x when h/w could not decode even .mp3 without adding parameters to lower quality...
You could play .mp3 if you decoded first.

Is that perhaps where your thinking is? Buffers? or?

#6

Last edited by number6 on 24-Jul-2012 at 08:30 PM.

_________________
This posting, in its entirety, represents solely the perspective of the author.
*Secrecy has served us so well*

@matthey

Quote:

matthey wrote: Quote: olegil wrote: I've always used this as my reference: http://www.microapl.co.uk/Porting/ColdFire/cf_68k_diffs.html And it doesn't exactly seem to agree with you on the differences there. There are other small differences (in user mode) but the big difference is that ColdFire left out many instructions and addressing modes of the 68k. The encoding space that the CF didn't use is still free. That's why they can be trapped by Freescale's CK68KLib: http://www.microapl.co.uk/Porting/ColdFire/cf68klib.html Notice where they say missing instructions and addressing modes. Notice also where the link you posted says re-introduced or re-implemented. They are just putting back 68k instructions and addressing modes that were taken away by the CF. The only way that is possible is that the encoding space is free (no encoding conflicts). The only way that Freescale's CK68KLib can trap 68k instructions is that the encoding space is free (no encoding conflicts). Quote: Cod3r wrote: You have the right info. Matthey's reference is off. What encoding is "off" in my reference? Name one! There may be a few minor mistakes but you are going to spend a lot of time looking up encodings before you find any, if you find any. Some CF instructions I list allow Rn instead of Dn but An being open (no encoding conflicts) does not affect the accuracy of mapping encoding conflicts. I allowed Rn in the map and in the disassembler as that was how the fpga 68k core was originally going to be mapped. It's really ignorant to leave out BYTEREV An when the value being changed is an address for example. I showed a pdf map of non-conflicting encodings of 68k and CF instructions and a working disassembler with source that can decode 68k and CF instructions at the same time. Some people would call that proof while ignorant people might not understand the proof. I didn't question your lack of proof but you question my proof? You have created doubts about your credibility by your lack of knowledge of the CF and lack of ability to examine programming proof of which you claim to have ability in both.

Dude in plain simple language I will tell you this:

I work for a company that is an official Freescale partner. That means I get access to training seminars AND my brother-in-law only worked in engineering at Motorola and gave me the information on how to make my work on the ColdFire easy.

I guess he doesn't know sh*t either right? I was talking about your notes in particular, there are some inaccuracies, but either way...

To directly answer what you have presented sounds accurate in theory and theory alone. Please try to directly run some 68k binaries on a ColdFire for the same OS without emulation and you will see plain and simple...

I am not talking about simple ADD/SUB/MOVE instructions, i'm talking a full real world example of a program.

It is possible to code a program to work on both, but you have to be sure that your code produces the same results on both CPU's. You couldn't just blindly code something and expect it to run without issues without considering that first.

Heck, you even have to be careful between some variants of the 68k family like the 68060. Now if you are running code generated by a C compiler and the compiler is specifically written not to compile to instructions that may create problem binaries you are ok. But you would be restricted and the code would certainly not be optimized and you could still have problems with any inline assembly code.

68k binaries to CF only work if you recompile for the CF target, not if you blindly run code compiled for 68k.

Last edited by Cod3r on 25-Jul-2012 at 04:49 AM.
Last edited by Cod3r on 24-Jul-2012 at 09:20 PM.

@Yssing

Those 280 megs are for storage.

So what (and how much) memory is used for web browser?

But you can still add lots of memory to lowend CPUs my A1200/060 has 256 megs.

Of course, you can. But the talk was about web browsing on 80 MB .

_________________
RastPort

@Cod3r

Proof of burden lies on the one making bold statements. You would earn a lot more from making a h.264 codec that can run HD resolutions in realtime and software only on a 400MHz coldfire than you could ever hope to make from anything Amiga-centric.

So I don't quite see why you would make bold statements and then throw a hissy-fit when told to back up said statements with some proof.

As for me, I'm currently busy doing hardware layout on a PPC board. My second PPC board, in fact. The first has entered production. I feel that speaks for itself.

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.

@Cod3r

Quote:

When I have the time to spare and 400mhz machine i'll be back with a video

Ok, got it. When you have the time to spare and 400MHz machine (probably with ColdFire CPU) come here and show us how it decodes h264 compressed 1280x720 or 1920x1080 video stream at full frame rate (25 or 50 fps). For ease of demonstration use Big Buck Bunny movie. You may download h264 compressed stream in both 1080p and 720p sizes.

Ah, the h264 stream shall be decompressed by the CPU, not some additional video processing unit, as e.g. the one in EfikaMX or RaspberryPI.

@michalsc

Quote:

michalsc wrote: @Cod3r Quote: When I have the time to spare and 400mhz machine i'll be back with a video Ok, got it. When you have the time to spare and 400MHz machine (probably with ColdFire CPU) come here and show us how it decodes h264 compressed 1280x720 or 1920x1080 video stream at full frame rate (25 or 50 fps). For ease of demonstration use Big Buck Bunny movie. You may download h264 compressed stream in both 1080p and 720p sizes. Ah, the h264 stream shall be decompressed by the CPU, not some additional video processing unit, as e.g. the one in EfikaMX or RaspberryPI.

We said 720p at first now you say 1080p? I am optimistic about what can be done with clever programming techiniques but I am not a fool. 1080p would need hardware GPU, hands down.

I am good but I never said I can walk on water or make a blind man see!

@Cod3r

Meaning basically that there are many OSes many more times capable and use less resources.

Many? Name three hobby operating systems many times more capable than MorphOS and fully usable on 400 MHz processor with 128 MB of memory (Efika configuration). I've done some preliminary research and found none so far.

Last edited by g_kraszewski on 24-Jul-2012 at 09:38 PM.

_________________
RastPort

@g_kraszewski

Hobby: Haiku, Syllable, Menuet OS...

And there are more...

Linux variants Puppy Linux, Tiny Core Linux, Damn Small Linux...

Atari's got MiNT...

Commercial: hmm... QNX, Plan 9...

That's all I got off the top of my head. You can play videos with a 400mhz processor with 128mb on all of them. I haven't tested it on DSL but maybe it can too.

@olegil


Quote:

olegil wrote: @Cod3r Proof of burden lies on the one making bold statements. You would earn a lot more from making a h.264 codec that can run HD resolutions in realtime and software only on a 400MHz coldfire than you could ever hope to make from anything Amiga-centric. So I don't quite see why you would make bold statements and then throw a hissy-fit when told to back up said statements with some proof. As for me, I'm currently busy doing hardware layout on a PPC board. My second PPC board, in fact. The first has entered production. I feel that speaks for itself.

I am not a hardware guy so I can't speak on your abilities, but I don't think you can speak on what I can do software wise either...

I am just letting you know I had scholarships for my whole time in college all the way to a masters. My employer waited two years for me to graduate because they wanted me to work for them. Paid me full salary for two summer internships. In America it is not common for an immigrant to accomplish such things including citzenship because of my abilities within my field.

@Cod3r

Quote:

You can play videos with a 400mhz processor with 128mb on all of them.

You cannot play h264 encoded 720p videos with a 400MHz on any of them. Try it :P It could probably work great with lower resolution and much less cpu-intensive video codec, but not with h264.

If video encoding/decoding were such a simple task, that even a 400MHz CPU could do that easily, Freescale, which you cooperate so tightly with, would never have added VPU unit to their i.MX series of processors. Why would they need to bother, if all that stuff could be done on CPU with ease?

@Cod3r
There's other platforms?Lol . What have you proposed to these interested parties? remember. Only amiga makes it possible...

http://www.youtube.com/watch?v=PWeO5IkCssk

@adrianbrowne

Quote:

adrianbrowne wrote: @Cod3r There's other platforms?Lol . What have you proposed to these interested parties? remember. Only amiga makes it possible... http://www.youtube.com/watch?v=PWeO5IkCssk

There are other platforms when it comes to 68k and ColdFire I haven't proposed anything they approached me interested in what I can do.

Amiga does make it possible but a split user base makes it impossible!

@Cod3r

Quote:

Amiga does make it possible but a split user base makes it impossible!

True. But those at the top see that as a plus. There's your problem.

#6

_________________
This posting, in its entirety, represents solely the perspective of the author.
*Secrecy has served us so well*

@number6

Quote:

number6 wrote: @Cod3r Quote: Amiga does make it possible but a split user base makes it impossible! True. But those at the top see that as a plus. There's your problem. #6

You my friend have stated the obvious. And the fragmented community continues to argue about who is best despite being five steps behind the rest of the computer world.

@All

If many of you knew how hard some people in this community work hard to build a better future for Amiga users you would probably be ashamed for battling each other in this way.

It does not matter if my ColdFire solution is the answer to everyone's problems. It is the answer to some and advancement of the community. It is one piece in the puzzle.

If all of us came up with a piece to the puzzle WE would own the future of Amiga NOT a corporation who thrives on shady business practices and killing the legacy of the platform we all know and love.

FREEDOM is the answer. Old greedy misers holding on to stale patents and trademarks and holding the community hostage is not the answer. We all have been fooled for too long...

I am one man but I can't save everything alone. I need you all to help too! Instead of bickering over everything I say spend the time improving some source code or designing a helpful piece of hardware or translating some software or writing docs.

If you love the platform as much as you say you do you would. Or just keep arguing and get ready for the new Amiga cigarette lighter, Amiga microwave and Amiga toaster.... and I don't mean video toaster

@Cod3r

Quote:

Amiga does make it possible but a split user base makes it impossible!

amiga will always be split user base. i was to amiga lan parties in the 90ies , you know there were no two computers looking alike. it was like zoo or car modding meeting. what this community is all about it seems is individualism or lets say diversity, only some do not get it right. therefore it is not going to succeed, because it would contradict its romantic privilege.

@Cod3r

I just read 30 pages ,most of it pointless. For starters your heading is an untruth.
Your 'so-called' company has refused to help you. I suggest the heading be altered from a company to a coder, for now I will use the word coder loosely.

Maybe you have achieved something using CF, just maybe you want to get involved.
But so far, most of what I have read is you telling us what you have done,or can do, and how good you are,but no proof to back it up. You also keep hinting you need a 400mhz machine to advance further, well, if you are expecting someone on here, to kindly 'donate' such a system, you might be in for a long wait.
I suggest YOU soucre your own machine, then implement whatever it is, and IF you can get something ,anything, worthwhile to display,then show it. Until then stop wasting bandwidth.

So far you have gave us nothing but words, and patting yourself on the back.
Seeing is believing, IF your are serious, stop talking about it, and DO, and show it.
Once you have something worthwhile to show, then lets see it. Until that time, you /we are going around in circles.

Yes, the Amiga communtity does need to go one one direction, and it can be advanced further.
Whether that be you, or whoever, less chat more action.

PS: If you think some here are being harsh, that wont be anything once you have something on show, Then wait for the vultures to attack. Good luck.

@DarthRaider

Welcome D.R.
Been away so long your old account de-rezzed? ;)

Plaz

@Cod3r

Haiku, Syllable, Menuet OS...

Haiku and Syllable minimum requirements are set at 400 MHz and 128 MB RAM. Then we have to slash the second part of "many times more capable using less resources" statement. But are they really many times more capable? None of them has hardware 3D acceleration for example. Looking at software base of Syllable I see it is comparable to MorphOS, just a bit smaller. Software base for Haiku is bigger, but surely not many times bigger.

MenuetOS is more eficient (authors claim 90 MHz and 32 MB RAM is enough), but it does not have any useful web browser for example. So it is much more capable than MorphOS? I doubt it. Software base for this system seems to be not existing.

I can continue this comparision, but my point is not to measure how much better or worse those systems are compared to MorphOS with six digits precision. My point is to prove you that "many times more capable using less resources" is simply not true.

Last edited by g_kraszewski on 25-Jul-2012 at 07:28 AM.
Last edited by g_kraszewski on 25-Jul-2012 at 06:56 AM.

_________________
RastPort