@Chunder

Nope, no price difference, the only chipset for the PPC970 set to be available in the neer future is the IBM CPC925 which supports up to AGP 8x and the slots are physically identical. Both Nvidia and ATI are going to continue with AGP cards, Nvidia's PCI-Express cards are simply AGP cards with a PCI-Express bridge

@Hammer

These are PCI-Express cards, usually (and correctly) abriviated to PCI-Ex, some less knowledgable peiople/companies abriviate them to PCI-X which can cause no end of confusion

@Hammer

Wow, Motorola using IBM chips never thought I would see taht, especially PowerPC chips! Shows how bad things have got at the big M!

@Hammer

Yes if 512Mb textures are needed the AGP bus would be really pushed as it needs two transfers per texture, but for most nomral applications the bus can hold several textures and teh texturing alghorithms in games are so efficient at preloading these days that they don't make full use of an 8x bus.

@blitterstorm

It would not be possible to run BeOS on this, BeOS only runs on the BeBox and 603/604 Macs, Apple would not give Be Inc the docs for the G3 systems when they laucnehed and Be had no wish nor the resources to reverse engineer a solution from the hardware, Mac OS binaries and Linux driver source code for such a small market, they stopped developing for PPC and went entirly x86. In 2001 they were brought out by Palm (now Palm Source) and all development ceased. yellowTAB make a product called Zeta which is bassed on the unfinished and unrealesed BeOS r6 aka Dano. This is x86 only and noone outside of yellowTAB even knows if they have the source code (a lot of work is simply drivers and apps, the rest can be done without the source but is not as easy as it would be with it)

I will look into the possibility of using DSP's (Motorola make a nice one, the MSC8126 StarCore http://www.freescale.com/webapp/sps/site/prod_summary.jsp?code=MSC8126&nodeId=0127958594 which has 4x400mhz cores and runs on the PPC 60x bus.

@Bobsonsirjonny


The idea of selling at cost is that everyone would be entitled to ONE board at the cost of producing+shipping etc. after that they would ahve to pay normal rates. Also bulk orders such as those by companies would not be eligable for purchasing a board at cost. This is simply an idea to help get people back into the Amiga community and I don't yet know if it is plausable, I got some half cock answer back from momentum about costs of producing the board and haven't beena ble to replky yet as I have been away (again - I am still away now but I've managed fto grab a computer) and my new monitor (20" SGI) doesn't have a cable for teh VGA poirt and I am still awiting for it to be delivered so in the mean time I ahve had to use either my SGI octane which doesn't want to work with the internet or my windows XP machine which is so god damn slow and can't post on here because of the firewall :S

@blitterstorm

I will look into the possibility of open sourcing the design, it depends on the licensing restrictions palced on the design by momentum.
Yes, I want to involve as many communities as possible but I am not sure how to go about this and I am but one man, if you know forums where you can publicse this then by all means go forth and do so, we will have to come up with ways of publicising the website once it laucnhes. Also Apple killed their clones business a long time ago and I don't think jobs is about to change his mind, although I sup[ose there is no harm in asking closer to the time.
I don't think their is a simulator available, I certainly haven't heard of one if there is and momentum and IBM don't publicise details of it, you would probably ahve to buy a board to get it anyway, that seems to be the only way to get anything out of momentum.
IBM have tried a few times to create an open PowerPC specification first with CHRP (COmmon hardware Reference Platform) then with POP (PowerPC Open Platform) but both times a lack of support from the largest PowerPC desktop manufacturer (Apple) has sunk the project before it canget off the ground. Check out www.openppc.org

@BrianK

Support from Eyetech (even if it is just manufacturing and/or sales) would be a fantastic boost to the AmigaFive projet, as Eyetech allready ahve relations ships with several maufacturers and are an established company with experience in the PowerPc market they have a good reputation which would bode well with both investors and suppliors of parts.
The last thing I want to do is kill a company that has helped to revive the AmigaOS, if I thought that this project woudl then I would never have started it but Eyetech is primeraly a reseller and so it is my beleif that this project would in actual fact strengthen their position in both the Amiga and Linux markets as well as alowing them to branch out into new markets. Don't forget taht once the ATX board is done Micro-ATX (Dual processor?) and even (possibly, depending on Power and thermal considerations - although if apple can do it with an iMac we should be able to) a mini-itx G5 (single processor).

@AmigaMac

Yes, if you look arround in the computing industry today (non-microsoft areas) then you see this a lot, and qwuite frankly this frustrates me, for once I want someone to be bold enough to take a risk and so this is what I want to do. Just like all risk takers I can't do it alone, I need help in several areas, especailly financially.

@Bobsonsirjonny

Though I cannot speak for Hyperion, it seems from my conversations with one of the devel0opers (One of the Freiden brothers, I foerget which) that they are certailly interested in a port but can't afford a $6000 board, especially if it isn't certainly going to be a part of the Amiga market. Last time I spoke I was told that they couldn't license to anyone other than Eyetech at the moment but were talking to KMOS about it. Please don';t quote me on this though, I may have got the wrong end of the stick (I will check exactly what was said when Iget home and ask for an update+clarification)

@blitterstorm

I never actaully said that the development, as in the final design would be open source. It is an interesting possiblity but there is certainly that risk, Alhthough with the right wording in a license you could make it open source for evelauation and limited production runs but prevent either large production runs or the sale of the board without a license. Another m,anufacturer wouldn't have the esential OS4 license + dongle code though. Alhtough in the realworld that wouldn't hurt them much as they would probably be after the Linux market.

@Intuitioned

Momentum already have Linux (2.6 Kernel) running on these boards so that's not a problem, take their patches, possbily get some people that know the kernel well (I could do it but it would be ugly!) to modify them for incorporation into the main kernel tree (if they're not already) and either modify Debain, Suse, Mandrake or Yellow Dog for distribution with the board or produce your own based on LFS

@digitaldisaster
Quote:

It would not be possible to run BeOS on this....

Axel Dörfler has a Pegasos board and has been working on porting Haiku (formerly OpenBeOS) to it. Genesi has been in touch with Michael Phipps, the Haiku project leader, from time to time, and they are interested in getting Haiku back on PowerPC machines, with the Pegasos being the point of re-entry. I don't know what additional work would be involved in supporting this G5 board, but at least there is movement in that direction.

-- gary_c

_________________
zukakakina.com - themes.tikiwiki.org

@digitaldisaster

Did you ever get the moneybroker thing set up? Or is that something to be arranged when you're back properly?

@thread

Does this thread show up in anyone's "My Recent Topics" list? It isn't in mine for some reason...

_________________

@Chunder

"My Recent Topics" only shows threads you've started yourself.

@Chunder

I set it up but I'm waiting for a confirmation email

@gary_c

Ah, Haiku not BeOS, yep can't see why we can't port that, in fact I would be willing to try that myself

@gary_c

With the more recent versions of BeOS built on x86 I can't see why they'd want to do a PPC and a x86 version of Haiku. But, if someone were to write a virtual machine which would run first then the x86 Haiku were to run on top of that I'd think it'd be a better way to go about it. There there's only one foundation to worry about the core x86 Haiku and if you wanted another platform to run your OS you'd do a virual hardware on top of the core hardware. Sure it wouldn't be as fast, but it'd probably be fast enough for the average joe.

@BrianK

NewOS, the microkernel which a forkled version of is used in Haiku is designed to be as platform antagonistic as is possible. A PPC port has always been the intention and so a G5 version would require no more than a few tweaks and the nessacery drivers for the CPC925 northbridge ()and possible the AMD8111 southbridge if they don't allready exist) as long as it ahs been writen correctly

@digitaldisaster

Quote:

yellowTAB make a product called Zeta which is bassed on the unfinished and unrealesed BeOS r6 aka Dano. This is x86 only and noone outside of yellowTAB even knows if they have the source code (a lot of work is simply drivers and apps, the rest can be done without the source but is not as easy as it would be with it)

YellowTab were/are interested in a Peg port after they've done R1. As for them not having the code, they're not saying as they're under NDA - but, yes they've got it.
Also IIRC all the released versions of BeOS run on PPC (even up to R5).

@BrianK

Quote:

With the more recent versions of BeOS built on x86 I can't see why they'd want to do a PPC and a x86 version of Haiku. But, if someone were to write a virtual machine which would run first then the x86 Haiku were to run on top of that I'd think it'd be a better way to go about it.

It's probably a lot quicker to port the OS!
Also BeOS (and thus Haiku) is based on much the same philosophy as the Amiga and running on a virtual machine is pretty much 100% contary to that.

_________________
Whyzzat?