Click Here
home features news forums classifieds faqs links search
6063 members 
Amiga Q&A /  Free for All /  Emulation /  Gaming / (Latest Posts)
Login

Nickname

Password

Lost Password?

Don't have an account yet?
Register now!

Support Amigaworld.net
Your support is needed and is appreciated as Amigaworld.net is primarily dependent upon the support of its users.
Donate

Menu
Main sections
Home
Features
News
Forums
Classifieds
Links
Downloads
Extras
OS4 Zone
IRC Network
AmigaWorld Radio
Newsfeed
Top Members
Amiga Dealers
Information
About Us
FAQs
Advertise
Polls
Terms of Service
Search

IRC Channel
Server: irc.amigaworld.net
Ports: 1024,5555, 6665-6669
SSL port: 6697
Channel: #Amigaworld
Channel Policy and Guidelines

Who's Online
12 crawler(s) on-line.
 32 guest(s) on-line.
 1 member(s) on-line.


 zipper

You are an anonymous user.
Register Now!
 zipper:  4 mins ago
 Karlos:  19 mins ago
 V8:  25 mins ago
 kolla:  1 hr 2 mins ago
 Hammer:  1 hr 8 mins ago
 wolfe:  1 hr 9 mins ago
 Gunnar:  1 hr 13 mins ago
 Aron:  1 hr 31 mins ago
 agami:  1 hr 59 mins ago
 Hypex:  2 hrs 6 mins ago

/  Forum Index
   /  Amiga OS4 Hardware
      /  X1000 CPU clue :)
Register To Post

Goto page ( Previous Page 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 Next Page )
PosterThread
vidarh 
Re: X1000 CPU clue :)
Posted on 22-Jun-2010 16:44:39
#141 ]
Cult Member
Joined: 4-Jan-2010
Posts: 580
From: London, UK (ex-pat; originally from Norway)

@koft

Quote:
You'd have to do what the A1000 did. Something nobody has seen or imagined yet. The a1000 delivered kick ass graphics and an OS that cut corners but ran fast on low cost processor while providing a better multitasking system than what was available at the time on micros that weren't running unix. It's easy to identify what was awesome in the past, what stood out above the crowd. What would be the same today? I don't know, the things I've mentioned are standard and pervasive. What would knock your socks off today? I don't know, if I did know, I'd be a wealthy guy perhaps.


Fair enough. Though I did ask for something "reasonably realistic" :)

I was thinking more in terms of what OS features and apps do you need before you'd be happy to use any of MOS, AROS or AmigaOS (assuming suitable hardware is available) full time. What would need to be fixed/added before the advantages of [any of the above] beat out the advantages of whichever you prefer of OS X, WIndows or Linux for you?

I don't consider an A1000 like step forward possible at this stage (if it was truly revolutionary, it'd get copied instantly by people with more money), but I do consider it possible to get an AmigaOS variant to the level where it is a viable alternative OS for more than just the Amiga faithful and something that's more enjoyable to use than the alternatives for enough people that it can actually grow again to a reasonably sized community. Personally I don't care about huge market share - the market today is so big that even 1%-2% would be fantastic and be viable for the long term.

The OS's need to take a few step further (this is where I harp on memory protection again), but beyond that it is largely down to 1) getting better development tools together, 2) people who want to see these OS's succeed churning out more new apps, or "salvaging" old apps (I've just started maintaining/cleaniung up/porting FrexxEd from classic AmigaOS, for example) that are worthwhile.

But this needs to happen largely *before* we can expect to get any fantastic new hardware at prices that are remotely in line with those of commodity hardware. The best we can hope for is probably a slow dance where each new project like the SAM 460 and X1000 (or Natami and FPGA Arcade or IMICA for that matter) brings in a few more users and makes it viable to aim a tiny bit higher in the next round.

The main thing is to actually get to an upwards trajectory, and I think we are pretty much there.

Until, hopefully, this community shows enough growth again to reduce the cost gap, at least some of us with the money and inclination will keep paying high premiums to play with what is - I admit - going to largely be an expensive toy. Other people buy sports cars when they start feeling old, I pick up my expensive Amiga habit again...

The way I see it in any case, don't get disappointed or annoyed with new entrants like A-Eon.

They may not turn out to be "salvation" for the Amiga community by themselves, but any new sales they generate means more hardware in play, more license money for Hyperion to update OS4, and makes it slightly more interesting for a few developers (most likely ex-Amiga users like me) to start writing some software again.

And while some of that may end up OS4 only, I suspect quite a few others like me would prefer to target as much of the community as possible. Even if the X1000 proves unsuccessful and only draws a small number of people, *any* additional resources will in the end benefit the entire community.

And people don't need to be developers to contribute. Bounties, constructive ideas, beta testing, graphics design, running websites, or just making the community more fun and enjoyable will all help making it more attractive for developers who are either scratching their own itches or are in it for the fun.

_________________
Wiki for new/returning Amiga users - Projects: ACE basic compiler / FrexxEd / Git

 Status: Offline
Profile     Report this post  
umisef 
Re: X1000 CPU clue :)
Posted on 22-Jun-2010 17:17:58
#142 ]
Super Member
Joined: 19-Jun-2005
Posts: 1677
From: Melbourne, Australia

@xeron
Quote:

The ShowConfig tool was completely replaced with one that just printed a message.


I am not convinced.

By now, I have seen photos of three different screens with full ShowConfig output (plus a second photo of the screen that started this thread). In addition, Rob has stated that he has a photo which shows the 12th PCI entry as "working" rather than "defective", so that would be a fourth screen. So if we were to believe that "someone found a full ShowConfig", then it appears as if at least 4 separate people "found" it --- which would suggest a rather poor effort at hiding it.

A more likely explanation would be that the ShowConfig icon has some tooltips or command line options set which caused the suppression of output to fail; Considering the people who would have done the hobbling are likely to be very comfortable with the command line, it's quite understandable that something like that could have been missed in testing[1].

So, applying Occam's razor, I will go with the most likely explanation until further information surfaces.


[1]: Good testers are highly skilled experts. Most programmers are lousy testers, whose personal preferences result in things going untested. And I include myself in that group --- Amithlon was distributed with an awful bug related to the right control key. Guess which control key I personally never use...

 Status: Offline
Profile     Report this post  
fairlanefastback 
Re: X1000 CPU clue :)
Posted on 22-Jun-2010 17:34:28
#143 ]
Team Member
Joined: 22-Jun-2005
Posts: 4880
From: New York, USA

@umisef

I guess the next question is, how hard would it be for them to entirely fake all the output?

_________________
Pegasos2 G3 running AOS 4.1 and MorphOS 2.0
Amikit user, tinkering with Icaros VM (AROS)
EFIKA owner
Amiga 1200

 Status: Offline
Profile     Report this post  
Leo 
Re: X1000 CPU clue :)
Posted on 22-Jun-2010 17:38:10
#144 ]
Super Member
Joined: 21-Aug-2003
Posts: 1596
From: Unknown

Quote:

I guess the next question is, how hard would it be for them to entirely fake all the output?

How hard ?

Quote:

If (cpuID == PASemi)
printf("you'll never know he he ;)\n");

_________________
http://www.warpdesign.fr/

 Status: Offline
Profile     Report this post  
Leo 
Re: X1000 CPU clue :)
Posted on 22-Jun-2010 17:38:30
#145 ]
Super Member
Joined: 21-Aug-2003
Posts: 1596
From: Unknown

Quote:

I guess the next question is, how hard would it be for them to entirely fake all the output?

How hard ?

Quote:

If (cpuID == 0x5000)
printf("you'll never know he he ;)\n");

_________________
http://www.warpdesign.fr/

 Status: Offline
Profile     Report this post  
Georg 
Re: X1000 CPU clue :)
Posted on 22-Jun-2010 17:51:54
#146 ]
Regular Member
Joined: 14-May-2003
Posts: 450
From: Unknown

@umisef

Or maybe the coders by mistake put the exe in the wrong place (like C:) so the original one in SYS:Tools (or wherever AOS4 has it) was still there. Now coders will typically run ShowConfig from a shell which would load it from C: So they might not have noticed that the original one is still there. But users would typically start it from WB by double clicking the ShowConfig icon in SYS:Tools which will start SYS:Tools/ShowConfig, ie. the "wrong" one.

 Status: Offline
Profile     Report this post  
MZ 
Re: X1000 CPU clue :)
Posted on 22-Jun-2010 17:53:01
#147 ]
Member
Joined: 23-Dec-2004
Posts: 70
From: Unknown

Does "showconfig debug" work with os 4.1.2?

 Status: Offline
Profile     Report this post  
umisef 
Re: X1000 CPU clue :)
Posted on 22-Jun-2010 17:59:03
#148 ]
Super Member
Joined: 19-Jun-2005
Posts: 1677
From: Melbourne, Australia

@vidarh

Quote:
Fair enough. Though I did ask for something "reasonably realistic" :)


Well, if you want to do something revolutionary, you must be willing to ignore your past. AmigaOS is very much an 80s OS, and *anything* you could derive from it, or turn it into, if still being recognisably AmigaOS will simply have no chance of being revolutionary, the way AmigaOS was back in 1985, in its way.

So, let's say rather than "take forward AmigaOS" or something like that, you have some (a lot of) money, time, clout and dedication, and you want to create something that is to the current crop of "mainstream" stuff as AmigaOS was to the mainstream stuff of the mid 80ies....
Then I'd suggest looking into abandoning the concept of files, file I/O, file systems and anything else that differentiates between "in memory" and "on storage" data. Instead, simply treat hard drives as the last level in the (64 bit) memory hierarchy; Any data, and any programs you have will always be "in memory" (even if only some of it will be "cached" in RAM). In order to keep track of things, and be able to find them, you probably want to provide OS services to deal with sharable in-memory objects (such as describing them, indexing them, access control both for security as well as for concurrency, etc). However, chances are that with everything always in memory, you might be able to reduce the "kernel" equivalent to something that merely handles the very basics (i.e. the memory hierarchy, process scheduling, and inter process communication), and move everything else into non-kernel objects --- which would provide some AmigaOS spirit :)

Now, this is obviously neither a new idea, nor is it an easy one to actually make work, let alone make work well. There's a myriad of details to get wrong, and if it isn't done quite right, it's not going to provide the benefits. It's a huge job to get *anything* like that off the ground, unfortunately, which is almost certainly beyond the means of anyone but some very big corporations.
What's worse, this being 2010, making a revolutionary new OS is not even half the job. Users don't want an OS, they want to interact with their computer (so need a user interface), and want to do stuff (so need applications). And once you do away with one of the major concepts of all current mainstream OSs, tapping into the efforts of open source by "porting" becomes very, very hard.

So no, given the tiny amounts of money which exist in the current Amiga market, it is extremely unlikely that a paradigm-changing development like that would come from this source; Nobody is going to sink tens of millions of dollars (if not hundreds) into it. But *if* someone were to do so, they could, in a few years, have a few years head start on everyone else :) (Or they could find, in a few years, that the whole thing was a bad idea. That's the nature of speculative risk investment :)

 Status: Offline
Profile     Report this post  
umisef 
Re: X1000 CPU clue :)
Posted on 22-Jun-2010 18:19:01
#149 ]
Super Member
Joined: 19-Jun-2005
Posts: 1677
From: Melbourne, Australia

@fairlanefastback

Quote:
I guess the next question is, how hard would it be for them to entirely fake all the output?


Fairly trivial, of course.

However, there are a few reasons to strongly doubt that's what happened:

(a) It has been stated that the command line ShowConfig suppressed all PCI output. Why would anyone go to the effort of creating two distinct doctored ShowConfig executables (or two distinct doctorings of the one executable)? Spending precious developer time on hiding the CPU's identity is understandable. Wasting it on frivolously doing it twice, not so much.

(b) Outputting something made-up is easy. Outputting something realistic-looking is much harder --- and the PA6T list shown on that screenshot is quite believable. AFAICT, there is the correct number of all the devices. Why go to such lengths?

(c) Related to (b): If all you want to do is hide your CPU (and be a smug smartalec about it), why would you put in data for a CPU that *could* be in the box? It doesn't give anyone the "ha, I thought I were clever, but you are obviously cleverer" feeling. Instead, it gets people to believe in the faked CPU, and they may either (i) like it, at which point they'll get upset when they find out it was all a fake, or (ii) not like it, at which point they might just decide to give their money to someone else before finding out the truth about your product. Neither of which is good for you. So instead, you would rather put something like "0x1057 0x5806" in there... That way, people will know they were outsmarted.


And, of course, HyperionMP had the chance to say "Ha! Suckers! You all fell for it! Aren't we clever-sneaky :)", and quite pointedly avoided doing so.

 Status: Offline
Profile     Report this post  
Rudei 
Re: X1000 CPU clue :)
Posted on 22-Jun-2010 19:05:36
#150 ]
Elite Member
Joined: 20-Nov-2002
Posts: 3589
From: Dallas, Texas

@-Sam-

Quote:
This guy has seen every NDA ...ever. In between being profitable and doing his wife.


That, my friend, was absolutely hilarious!!!!!



Rude!

_________________
2017 Camaro 2SS

 Status: Offline
Profile     Report this post  
DiscreetFX 
Re: X1000 CPU clue :)
Posted on 22-Jun-2010 19:12:16
#151 ]
Elite Member
Joined: 12-Feb-2003
Posts: 2209
From: Chicago, IL

@umisef

You could still make Amiga OS 4.x truly revolutionary by porting it to Quantum computers.

_________________
Sent from my Quantum Computer.

 Status: Offline
Profile     Report this post  
Rob 
Re: X1000 CPU clue :)
Posted on 22-Jun-2010 19:27:57
#152 ]
Elite Member
Joined: 20-Mar-2003
Posts: 6194
From: S.Wales

@umisef

Quote:
Rob has stated that he has a photo which shows the 12th PCI entry as "working" rather than "defective"


Here it is.

http://i49.tinypic.com/2niqsk5.jpg

 Status: Offline
Profile     Report this post  
itix 
Re: X1000 CPU clue :)
Posted on 22-Jun-2010 19:59:42
#153 ]
Elite Member
Joined: 22-Dec-2004
Posts: 3398
From: Freedom world

@Georg

Some filesystems have recycled dir (i.e. SFS and PFS) and if old executable was just replaced it could be possible someone just found old binary from there... it is also possible someone just plugged in his usb memstick and ran showconfig from there. Or found an install cd somewhere...

Many many years ago when there was an Amiga event in Finland I saw how one guy wrote small app on a demo machine. If you leave machines unattended anything can happen..

_________________
Amiga Developer
Amiga 500, Efika, Mac Mini and PowerBook

 Status: Offline
Profile     Report this post  
umisef 
Re: X1000 CPU clue :)
Posted on 22-Jun-2010 20:27:18
#154 ]
Super Member
Joined: 19-Jun-2005
Posts: 1677
From: Melbourne, Australia

@DiscreetFX

Quote:
You could still make Amiga OS 4.x truly revolutionary by porting it to Quantum computers.


Again, I'll give you the benefit of the doubt, and assume that you were making a joke...

 Status: Offline
Profile     Report this post  
RodTerl 
Re: X1000 CPU clue :)
Posted on 22-Jun-2010 22:25:23
#155 ]
Cult Member
Joined: 6-Sep-2004
Posts: 589
From: Rossendale

Ive been working on th idea for years. .theoretically, my proof of concept quantum logic gate can be built using a PC with sound card, speakers, but preferably more than one mic.. and a Large cardboard box.

Unfortunately, for the wavelengths and frequencies involved, I cant afford a box that big.. at least the triple layer thing used to carry commercial photocopiers. 8(

Why do I say my quantum CPU idea for Amiga works so easilym, and others dont?.. Everyone else is working on quantum particle forms.. Im using Waveforms. ..you know.. the ones you can DVBT send 8000 simultaneous carriers through the logic?

Theres lots of fun things out there.. but teh researchers dont talk to each other. .theyre doing paair analysis to understand music, speech, text. .which in unified symbol coding. .just means theyre repeating what Collosus does. .theyre breaking the code of reality. So.. work on teh code directly.. And you find out how to understand an awful lot more of reality than you realised.


_________________
The older and more respected a scientist is, the longer it takes to prove him wrong.

 Status: Offline
Profile     Report this post  
amitv 
Re: X1000 CPU clue :)
Posted on 22-Jun-2010 22:36:11
#156 ]
Regular Member
Joined: 23-Oct-2006
Posts: 346
From: Unknown

@all


http://www.linuxfordevices.com/c/a/News/Freescale-QorIQ-P3014-P5020-P5010/?kc=rss

 Status: Offline
Profile     Report this post  
xeron 
Re: X1000 CPU clue :)
Posted on 22-Jun-2010 23:47:20
#157 ]
Elite Member
Joined: 22-Jun-2003
Posts: 2440
From: Weston-Super-Mare, Somerset, England, UK, Europe, Earth, The Milky Way, The Universe

@umisef

Quote:

umisef wrote:
@xeron
Quote:

The ShowConfig tool was completely replaced with one that just printed a message.


I am not convinced.


I'm not guessing, i'm telling you that I know for a fact that ShowConfig, at least the binary in its usual location, was replaced with one that printed a one-line message and exited.

If someone found a copy of the original binary somewhere else, i do not know.

_________________
Playstation Network ID: xeron6

 Status: Offline
Profile     Report this post  
theamigaman 
Re: X1000 CPU clue :)
Posted on 23-Jun-2010 0:45:24
#158 ]
Member
Joined: 6-May-2010
Posts: 35
From: Unknown

@amitv

VERY interesting find!

It could very well be the freescale e5500 and Varisys is one of thier partners!

Looks pretty sweet if it is! Makes a bit more sense to me because it JUST got announced!



But its still speculation until the official word.

 Status: Offline
Profile     Report this post  
minator 
Re: X1000 CPU clue :)
Posted on 23-Jun-2010 1:26:10
#159 ]
Cult Member
Joined: 23-Mar-2004
Posts: 989
From: Cambridge

@amitv

Here's a better link.

Clearly PPC is not dead yet, that thing is rated faster per clock than a G5.


However there is something very interesting about the specs of the p5020, It looks remarkably like an enhanced PA-Semi chip.

There's clearly changes, slightly different ports some bits faster as well, less Serdes but they're twice the speed. The CPU cores appear to have changed a bit as well, they've got their own small L2s and the pipeline length has been cut in 2. Looks like an effort to boost control plane (i.e. integer) performance.

However, P5020 isn't due until 2011, but given Freescale's record that means 2012. Last time Freescale did a 64bit PPC (The PPC620) it was 4 *years* late and then only shipped in very limited quantities.


My guess is X1000 will ship with a Freescale branded CPU that is basically a rebadged PA-Semi chip. Then, when it's ready there'll be an upgrade to the p5020 (assuming A-eon can actually sell enough to keep going that long).

_________________
Whyzzat?

 Status: Offline
Profile     Report this post  
tiffers 
Re: X1000 CPU clue :)
Posted on 23-Jun-2010 4:38:15
#160 ]
Regular Member
Joined: 7-Jun-2007
Posts: 349
From: Perth, Western Australia

@koft

Quote:

You wrote:

It's not standard for the industry to do this.... NDA saying you can't tell potential customers what part you are using in your product, nah. Nobody does that. That's absurd man...


So are you saying it's normal practice or 'business as usual' for hardware designers (specifically CPU or SoC developers) not to require their partners (regardless of size) to hide details about the future use of their (SoC producer) new design in upcoming product releases?

Do I understand correctly that you're saying it's routine to have NDAs which don't include clauses which limit the information which can be given out during the development phase of any product using newly released SoC/CPUs?

Are you also saying that IF any NDA actually should happen to (senselessly, according to you) include such a limitation, it would be written in said NDA as a specific, set time and date, after which the clause no longer ceases to be valid, and can be treated as non-existant, while the rest of the NDA would still be held to be current and true and indeed still as restrictive as it was before, with the exception of being able to reveal that small bit of information, which the targeted market is so keen to hear?

Well... I guess PA Semi and A-Eon must be a right bunch of ####s then! What do you say? Who could imagine ANYONE ever using a clause such as "won't reveal details until first production run is complete", so if dates slip, the NDA will still protect. Who could imagine a company including such clauses in their NDA because they wouldn't want a bunch of 'startups' announcing upcoming products, but never actually relasing them, and the SoC developer becomes associated with 'vapourware' announcements. Why would they care about such nonsense? Surely there couldn't possibly be people in the community that would actually keep a list of vapourware and continually bring it up, smearing the names of anyone involved or attached to that list? Couldn't be.. could it?

Oh and Trevor mentions that early in 2008, they got a CPU partner. When was PA Semi sold to Apple? Would the possible sale of all IP to Apple POSSIBLY have cause PA Semi to be extra careful with the wording and restrictions used in their NDAs at the time? Was there a period of negotiation, do you think - in your professional opinion, during which PA Semi had to continue to grow their business with new partnerships, and yet try to move forward with these negotiations, privately, all the time ensuring that their existing commitments would be met, whatever the outcome of such negotiations? Or does everyone run a business in exactly the same way because everyone has exactly the same set of circumstances in which they are running said business? All of this is assuming of course that PA Semi is the mysterious partner. Without knowing the partner, you can't assume to now the NDA clauses.

Give us a break. Conspiracy theories are fun... but only for a few minutes, or until you actually use your brain and think things through.

tiffers

Last edited by tiffers on 23-Jun-2010 at 05:29 AM.

 Status: Offline
Profile     Report this post  
Goto page ( Previous Page 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 Next Page )

[ home ][ about us ][ privacy ] [ forums ][ classifieds ] [ links ][ news archive ] [ link to us ][ user account ]
Copyright (C) 2000 - 2019 Amigaworld.net.
Amigaworld.net was originally founded by David Doyle