Amiga.org
Amiga computer related discussion => Amiga Software Issues and Discussion => Topic started by: arnljot on February 11, 2009, 10:04:40 AM
-
I guess this is old news, but it was new to me: Announcement by bbrv about efika2 (http://www.powerdeveloper.org/forums/viewtopic.php?p=11910)
You will note we are in the process of preparation...
http://projects.powerdeveloper.org/projects.php
The "EFIKA2" has been our way of describing what is next for at least the last three to four years. The first EFIKA came to market with the 5200 and was launched at FTF 2005.
http://www.genesi-usa.com/press/2005/6/24/
At the Forum, Genesi also introduced the EFIKA 5K2, a MPC5200 Performance Evaluation Board. Genesi will now work with graphics card manufacturers to optimize the board through the use of the onboard PCI support in order to transition to a smaller two chip board solution targeted at mobile and embedded applications.
The first "EFIKA2" was the next version which featured the 5200B, an updated version of the 5200. These boards first shipped in November 2006 and continue to be sold today.
In the meanwhile, the next "EFIKA2" was planned around the 5121e/5123, but after a significant amount of development we have abandoned this effort. The chip went through too many revisions and we wasted significant time and money supporting THTF/LimePC and Cherrypal without compensation or reward. Our conclusion: 1) the 5121e/5123 is not well-suited for consumer products that require a desktop environment, and 2) the 5121e/5123 was designed as an embedded processor and will find success in the industrial and automotive segments. Nevertheless, we did our best to support Freescale and have maintained this effort until very recently:
http://en.opensuse.org/MPC5121eADS
All that to say, the next "EFIKA2" will truly be something new, so please stay tuned...
So there will be no Efika2 which is PPC, it´ll be ARM. It also seems that Genesi is abandoning the PPC platform, but maybe I´m reading too much into this :)
Now, with the (seemingly) close ties between Genesi and MorphOS team, will the CPU choice spill over to them?
-
No
-
Ok:)
Piru wrote:
No
-
@piru
But then, what is the target hardware ?
-
Bennymee wrote:
@piru
But then, what is the target hardware ?
PPC as always... Make note of the fact that a PPC Mac versionis on its way... :-)
-
well Michal has the job/bounty from genesi to port aros to the new freescale arm board.
so aros will go arm :)
-
Can we expect a broader line of supported PPC Macs, like Quicksilver G4's, Nvidia and etc...?
-
aros will go arm
I'll stick with MorphOS for now thanks...wake us all up when AROS can actually do something...not sarcasm, wake me then! :boohoo:
-
Fingers wrote:
aros will go arm
I'll stick with MorphOS for now thanks...wake us all up when AROS can actually do something...not sarcasm, wake me then! :boohoo:
Wakey Wakey!!!
http://vmwaros.blogspot.com/
http://www.youtube.com/watch?v=scY9LXEGCB0
-
Dr. Schulz just released a new test ISO (http://linuxtracker.org/download.php?id=cca86a96f8ddb79078c609f0f45c5f2cd02c8ad8&f=aros-efika-20090211.torrent) for the EFIKA I.
Dammy
-
dammy wrote:
Dr. Schulz just released a new test ISO (http://linuxtracker.org/download.php?id=cca86a96f8ddb79078c609f0f45c5f2cd02c8ad8&f=aros-efika-20090211.torrent) for the EFIKA I.
Dammy
What is the point of porting Aros to Efika? Efika already has Morphos and Aros could run on much cheaper and faster x86-hardware. Intel D945GCLF (http://www.intel.com/Products/Desktop/Motherboards/D945GCLF/D945GCLF-overview.htm) it only costs about 80 euros and it's much faster than efika.
-
What is the point of porting Aros to Efika? Efika already has Morphos and Aros could run on much cheaper and faster x86-hardware. Intel D945GCLF it only costs about 80 euros and it's much faster than efika.
Some people only want PPC. Thankfully, more and more people have realized that PPC is dead for future development and it's time to look for arch that has a real development. But for now, the old bounty is about to be finished.
Dammy
-
So, why not MorphOS and Amiga OS 4.1 are also ported to something not dead?
-
So, why not MorphOS and Amiga OS 4.1 are also ported to something not dead?
Good question.
Dammy
-
What is the point of porting Aros to Efika?
For x86 version it is a good way to find bugs. It's also a good way to be sure source code is endian agnostic and portable.
It could open the doors to other stuff like integration of 68k emulators like Asgard (written in ppc asm for MacMame) or making AROS more compatible with MorphOS/AmigaOS4
-
bloodline wrote:
Bennymee wrote:
@piru
But then, what is the target hardware ?
PPC as always... Make note of the fact that a PPC Mac versionis on its way... :-)
But what about NEW hardware ?
-
Bennymee wrote:
bloodline wrote:
Bennymee wrote:
@piru
But then, what is the target hardware ?
PPC as always... Make note of the fact that a PPC Mac versionis on its way... :-)
But what about NEW hardware ?
Obviously you will have to ask Piru... but perhaps that is as far as they want to take it... it is their Hobby/SemiPro project...
There are plenty of Second Hand PPC Macs, that should meet the demands of any of the Amiga community interested in MOS.. yeah?
-
There are plenty of Second Hand PPC Macs, that should meet the demands of any of the Amiga community interested in MOS.. yeah?
As long we're not restricted with Mini g4's...
(http://www.applematters.com/assets/images/uploads/old/products/cache/Apple-ppc-G4-2003_300x360.jpg)
(http://lowendmac.com/imacs/art/imac20.jpg)
(http://images.csnw.com/lrg/appl_pmg4.jpg)
Those would be real sexy Amigas...
-
hbarcellos wrote:
There are plenty of Second Hand PPC Macs, that should meet the demands of any of the Amiga community interested in MOS.. yeah?
As long we're not restricted with Mini g4's...
(http://lowendmac.com/imacs/art/imac20.jpg)
Those would be real sexy Amigas...
......painted in red and white chequered
-
Dr. Schulz just released a new test ISO for the EFIKA I.
Dr.Schulz updated his blog (http://msaros.blogspot.com/2009/02/one-tiny-fix.html).
Dammy
-
bloodline wrote:
Wakey Wakey!!!
http://vmwaros.blogspot.com/
http://www.youtube.com/watch?v=scY9LXEGCB0
Well, I can't watch the video as I'm at work.
The website, although mentioning a TCP/IP stack, makes no mention of a web browser.
Also, can a single AmigaOS3.1 application run in AROS yet without porting/altering?
PZ.
-
Fingers wrote:
bloodline wrote:
Wakey Wakey!!!
http://vmwaros.blogspot.com/
http://www.youtube.com/watch?v=scY9LXEGCB0
Well, I can't watch the video as I'm at work.
The website, although mentioning a TCP/IP stack, makes no mention of a web browser.
Also, can a single AmigaOS3.1 application run in AROS yet without porting/altering?
PZ.
Actually yes, with WmwAROS's UAEBridge... but I have got to the point where I have no system friendly AmigaOS apps that I want to run anymore...
-
dammy wrote:
Dr. Schulz just released a new test ISO (http://linuxtracker.org/download.php?id=cca86a96f8ddb79078c609f0f45c5f2cd02c8ad8&f=aros-efika-20090211.torrent) for the EFIKA I.
Dammy
Why is the efika build NOT available on the official AROS website?
-
Why is the efika build NOT available on the official AROS website?
Because it hasn't been officially submitted yet as it's a WIP release?
Dammy
-
dammy wrote:
Dr. Schulz just released a new test ISO for the EFIKA I.
Dr.Schulz updated his blog (http://msaros.blogspot.com/2009/02/one-tiny-fix.html).
Dammy
Ah, sign extension. The bane of the bit banger. Bitter experience has taught me to use typedef'd unsigned ints and chars for all low level datatypes. I hope it didn't take Michal too long to find that bug, they're usually pretty annoying to debug (what do you mean it's -4? I assigned 252! is memory getting corrupted?) and one invariably feels like a tool when the bug is discovered.
-
CodeSmith wrote:
Ah, sign extension. The bane of the bit banger. Bitter experience has taught me to use typedef'd unsigned ints and chars for all low level datatypes. I hope it didn't take Michal too long to find that bug,
Well, the problem was more complex. The virtual address space for PPC architecture is slightly larger than 4GB. Since I wanted to make it clear, I was in the need (at least to some degree) of 64 bit arithmetic. Yes, my fault :) but I really didn't knew, that a cast like:
uint64_t foo(void *addr)
{
return (uint64_t)addr;
}
will sign-extend all addresses above 0x7fffffff :) Well :)
But no, it didn't took that much time. As soon as I saw that the PTE search in MMU tables fails for upper 2GB range, where it should not fail at all, I fixed it pretty quickly.