Well I guess the title says it all... is it possible for bPlan to use a more modern southbridge in future revisions of pegasos? Not that the present one doesn't do the job but newer ones have USB2, 6 channel audio, and they are generally faster.

and according to the thread at

http://forums.extremeoverclocking.com/showthread.php?t=64473&page=1&pp=20

northbridges are nearing the end of their usefulness ?

Well until Marvell (or someone else?) designs a northbridge that encapsulates some of the functions that are missing from peg 2 then, there is still room for improvement the "easy" way (substituting the dated 686B chip with a more modern one- easy compared to designing a NB) .

ALi's M1563

Neat isn't it?!? (/me wishful thinking)

I don't think you will see a Peg IIb. With 6 Ch.
Sound or other goodies anytime soon. We even cannot use Firewire or the Digital Out (or do we?) already availabel on the PegII. So no need (for me) to implement new Hardware.

If you mean. Is it planned to make a full new
Pegasos. I can say: It has been anounced. Pegasos III (G5). But after the anouncement you didn't hear anything further.

Ways to get Multi channel out on the PegasosII
may be to develop a proper Driver for the Soundblaster live cards. OK no chipset on board but a somehow usable way.

@Van_M

The Pegasos is NOT useing the 686B, that was another board

The SB on the Peg is also from VIA, but quite a bit "younger".

On the site to your ALI-SB I can read about stuff like Hypertransport, which makes me guess that it is NOT "real" SB (a chip orginally only inteded to terminate the PCI-bus), and therefore unsuitable for "classic" NBs like Marvel's Discovery (and MAI's Articia).

It might be the the next generation of Marvs NB (for G5) do feature modern standard like VLink and Hypertransport, makeing it possible to use current "SBs", but that would offcourse need a full new mobo.

Actually the hardware part is not so important in my opinion.
Pegasos 2 use cheap and fast AGP 2.0 boards, standard PCI slots, a decent USB 1.1 connection, plus IDE etc.
Not only ... we have Spdif Audio and Firewire connection ... PLUS a 1Gbit Ehernet connection: these are all good toys that we cannot use due software problems.
We need first of all an Os capable to use FULLY the Pegasos2, coz, honestly, my Pegasos 2 G3 with an Ati 7000 is already a FAST and RESPONSIVE system!

I cannot image the performance leap that will have shortly with the 1gbit support and the Native MOS Tcp-Ip, for example.
And ... if the USB author was capable of a such IMPRESSIVE WORK on the USB stack ... I cannot image how we'll be in future the Firewire support!
And finally we need FULL drivers to support FULLy the good AGP 2.0 cards that are outhere!

Returning to the hardware section ...
AGP will remain the standard connection for the next year as all analists claim, equally is for DDR RAM: PCI-EXPPRESS and DDR2 will be a reality in the mass market only in 2005; these are the real good innovations that can give a sense to a creation of a new Pegasos, but actually it's too premature! We don't support today ATI 9000PRO ... sure we cannot create drivers for the R420
The rest ... all SouthBridge issues can be addressed with PCI cards (USB 2.0 cards, SB LIVE etc)

In the meantime ... give us Mos 1.5 (or ... 2.0, because with all the new additions it will be more correct call it 2.0 )

VT8231 isn't supported at all, i'm unhappy with this, parallel.device is only for printers... With incomplete hw support of his main hw, we will have support from 3rd party hw? bad beginnings...

Firewire? still unsopported...

programmers can have much work and no time for some things... is problem of Genesi that exploit his developers? But mos programmers are really good, but they need be more?

Patience people, patience... Remember when we first heard about a brand new PPC machine which will be capable of running Amiga programs without UAE or such emulators, and with AGP and such goodies. Only a few of us even believed that! Now the Pegasos is here, and rocks. There are a lot of developers, and things will be even better when MOS 1.5 is out. And the firmware update early this month already made things a lot better.

We need recent HW (and the Pegasos is very recent, compared to anything available in the same price-range, not counting peecees) but we don't need the most modern HW. We need to get out the most of we already have. At least this is where users of Am*ga(-like) systems were always the best.

The developers of MOS will do their best, i'm sure, because they not only doing it for money, but because for fun. We've only sit here, and wait till they're ready.

The only point where i'm agreeing with the critics, we need more programming docs about the low levels of MOS, and the internals & tricks of the firmware. With this, for eg. someone could port a better parallel.device from for eg. Linux, among others.

The only point where i'm agreeing with the critics, we need more programming docs about the low levels of MOS, and the internals & tricks of the firmware. With this, for eg. someone could port a better parallel.device from for eg. Linux, among others.
_________________________

Absolutely right!
Mos core developers are few and we have to help them; but to help in the field of drivers, low level Pegasos use ... we need docs that descibe well the MOS memory use, device implementation, relationship of the core with the OpenFirmware and so on!
On the other hand we have to remeber that under Linux the support for Firewire and the Marvell 1Gbit connection doesn't exist!

Go MOS Go!

Actually, I've seen reports of firewire working in Linux (on the Pegasos) with kernel 2.6.

Quote:

Amen to that!

@dholm

Under Linux, Firewire was already useable on the Peg at least since late 2002.

@Senex:
That was what I suspected although I wasn't able to confirm it.

Firewire as well as the firewire controller is standardised so if the OS has firewire support it shouldn't matter which controller you use as long as it isn't custom-made.

Anyway, why do everyone want firewire support so badly in MOS? Do you all have firewire equipment, if you don't you might as well get the USB equivalent.

Yes, I have a external firewire HD and an external firewire CD Burner. So i could use all of them in/on PC iBook Pegasos. USB (1.0) is really slow if you want to copy large files. Maybe the People coding
Video Software (Motion Studio) would appreciate it too. Most DV cameras do (only?) have a firewire connector. And the last thing: The Pegasos would look more complete. If you show it to your friends
it will look lots better if all ports are functioning.

The other thing is I don't like USB too much because:
a) I think Firewire is the better Technology (for high bandwith connenctions). For Keyboard an mice it's ok, but not for high transfer rates.
b) I did not like the way Intel tried to keep out firewire when it first appeared.

But that is only a personal preference.

Bye

Quote:

Sure, but .....

ATM all HW-drivers run in the ABox (probraly useing some secret gateway in Quark), but are planned to be moved to the QBox once that has matured (and the gateway would than be closed).

This all no problem when you have developed ALL drivers with the transition in mind, and have full access to the sources.

Writing drivers through the allready existing API (like for PCI and USB) is no prob, but stuff that is onboard should been done inhouse.

@Paochi, Chain-Q, Zylesea and others...

I agree that we need more programming docs about low levels of MOS and the internals & tricks of the firmware...

But they act with some ocultist method of work, and this isn't good, this is a minoritary OS, so they will convret MorphOS as a QNX-like but for other markets? So is probably...

No fully supported parallel, not firewire stack, not gigabit ethernet drivers... and others things still not present...

If MorphOS not supports all his main hw (pegasos), then how we will have more 3rd party hw support?

We are still without gbit ethernet drivers, this is because MorphOS development method is ocultist, more than microsoft windows... And then you can only make drivers with ?OpenPCI?

They will not release internal info of MorphOS system... So because uses his top-secret and ultra-advanced quark? LOL

For they we are only some stupid geek users that will use that they gives to us just because has amiga compatibility and an amiga-like system in some aspects (we are still using amibox for the majority of things...) and they sometimes makes things for us for still use their thing? We are only some future developers and testers for his ultra-top-secret plans?

I'm not saying who is evil and who is good, these are stupid concepts since nothing is evil or good, and a company is created for money, if not, MorphOS was been a GPL project, and you know the problems of free projects...

And I'm not saying that OS4 will be better, all contrary, if i and others aren't happy in some aspects with MorphOS, imagine OS4 with his programmers, his crappy hardware (articia with demonic possesion) and his crappy money resources (WHO #### IS KMOS?)

And I will not shut up, since I'm free of say my opinions, I don't hate Genesi and bbrv and i like MorphOS, but i like some things still not present and i have different ideas about some methods of work of Genesi...

We will see if this is going to change for better...

@Kronos

But onboard stuff isn't all supported, that's the problem...

TO ALL: What do you think about I was posted on this thread?

Perhaps the VIA 8237 could be glued onto a Peg3 in some way? It looks quite nice in my eyes:

http://www.via.com.tw/en/southbridge/vt8237_index.jsp

The main problem IMO is not these informations not available right now, but the most annoying is the silence from the core developers on this topic. I can understand that they don't want to make any promises what they can't stand later, but saying nothing is just plain worse than saying no.

I'm sure they work hard, and can't follow every discussion about their work, but this has been already told too many times to left unrecognized...

@dholm:

You could see a Peg running a FW drive at the GIGA TV presentation.
Reasons for Firewire in MOS:
*the HW has FW so why the hell not support it with its native OS?
You paid for it!
*FW is much faster than onboard USB
*there are no high-speed USB2 drivers for MOS

@takemehomegrandma:

Yeah, bplan should do a PCI-X to V-Link bridge
or ask VIA for one.
This way you can use the 4x bandwidth the Marvell
chip provides with PCI-X, and still have regular
PCI provided by the southbridge and the benefit
of its advanced features.
The current architecture/implementation stinks. :)

Quote:

Wrong, there is. But it's reliability is not proven very well yet (at least I didn't heard too much about it, maybe Palpatine can enlighten us a bit more about this issue). But I am optimistic that it will work fine, cause the current work on Poseidon was just georgeous.

@tarbos
Another reasons;

- Too useful for connect DV/miniDV cameras (if these cameras include usb, is too slow than fw...)...motionstudio?
- A newer and better replace of SCSI... fast HDDs, DVD burners...
- More fast and better than USB2 and IDE, more stable data transfer...
...Any more?

Seeing that core developers ignore us, why not make some type of massive petition (please not use petitiononline.com or similar) as send emails, protest on other sites...?

We need protest more, make that they hear to us and provide solutions to OUR problems and OUR needs, and not be as stupid windows users...

@timofonic,

Why blast the core Developers with e-mails? If they are silent
about certain parts of the core of Morphos their must be a reason.
Wouldn't it be better to e-mail someone specific and ask
them a direct question? Maybe even ask BBRV what the current
policy is regarding info/docs on underlying parts of Morphos.

Or if you are a developer begin developing, then when you have a
legitimate question on MDC, maybe someone will be more free with info!
They may only give info on a need to know basis!?

Enjoy

Katos1