Quote:

That's harsh. Uneducated? Did Genesi write the OS?
Genesi paid out some cash, gave away some hardware, but when they were involved the only hardware later versions of MorphOS ran on was from Genesi/bPlan.
Genesi was motivated by profit, the MorphOS team can't be doing this soley for money (there isn't enough in this for them to quit their day jobs).
Further, when Genesi was involved there was a lot more friction between MorphOS user/supporters and other Amiga camps.

And I'll stand by that original statement. Once Genesi moved away from PPCs everyone assumed that (without a hardware manufacturer) MorphOS would have to move to ARM (following Genesi).
Not only hasn't that been the case, but the development team has continued enhancing their product and selected a really good path for new hardware support.
I admire the development team and trust their judgement.

I couldn't care less about Genesi. The Pegasos was flawed when introduced. It never was that impressive and the Efika was less so. For that matter, I was never a big fan of Commodore.
You have the right to your opinion, but mine isn't the result of an uneducated conclusion. It's just different from yours.

Hey Jim

Let me ask you a simple question. What do you know about Genesi other than what you have read OTHER people write about them online? In other words do you have ANY experience working for or dealing with genesi?

When I say "uneducated" this means you have no real world knowledge about Genesi or its relationship to Morphos REALLY. Other than what RUMOURS and facts you have seen online.


As far as Pegasos being "flawed" well for first gen hardware with a FAULTY CHIP (read MAI Logic ARTICIA S). The only "flaw" as you say was this chip. Peg2 with Marvell is practically same design upgraded. So, now with this other broad statement of the Pegasos being "flawed" you have succeeded to marginalize yourself twice!

[ Edited by magnetic on 2010/11/18 2:04 ]

I'd also say that this post-Genesi era is surprinsing me every day. Until now, amigans have alwasy had this sick "mother company" dependency, and in our case, it's been Genesi, until they first diverted to Linux, and then to ARM.

Logic would say that amigans would follow Genesi as little ducks follow mother duck. But it hasn't happened: The MorphOS Team took a brilliant decision going to Mac hardware (which happens to be good and cheap). I'm very proud of what this little group has achieved, because it's been their honest hard work rewarded directly by we consumers.

Meanwhile, Genesi's PowerDeveloper.org site shows much less activity than MorphZone.org, which is quite a surprise. I agree that Genesi's support is great, but you should also agree that they have promised too many things that didn't happen. Perhaps that was becasue every one was tied to some third party agreement that was never fulfilled. Genesi's business is much harder to do, of course, but MorphOS Team delivers instead of promising, as they don't have this kind of dependencies.

> Once Genesi moved away from PPCs everyone assumed that (without a hardware
> manufacturer) MorphOS would have to move to ARM (following Genesi).

It certainly was not "everyone" :-)

> they have promised too many things that didn't happen. Perhaps that was becasue
> every one was tied to some third party agreement that was never fulfilled.

Anything specific you want to hint at with that?

Quote:

Well, those failed ventures are publicly known. The biggest one was THTF's betrayal, but I don't think pointing out failures is constructive. Well, perhaps for documenting history is, but then, there's a lot of information out there, and history shouldn't be documented by a single guy in a forum (like me).

What's interesting, though, in the THTF soap opera, is that Cherrypal did what Genesi couldn't. We could argue that it's not that Genesi couldn't, but didin't want. But I still think there's a point.

Anyway, who cares about failed projects unrelated to MorphOS (even if they were originally very related), let's move ahead.

Quote:

Oh no! Marginalized twice, but you still felt the need to counter my statements.
You are an angry fellow.

If not agreeing with you somehow diminishes me in your eyes, so be it.
You miss my point. MorphOS existed before Genesi was involved and it continues to excel after Genesi has moved on. I know who I trust and who I feel a sense of allegence to and its not Genesi, its the MorphOS development team.

And, yes, I have had some interaction with Genesi. I've exchanged some PMs with Matt about locating an MPC8640/8641 development board (for less than the $4000 Freescale currently wants for one).
I got the distinct impression that he would prefer that the development team create an ARM port, rather than run MorphOS on someone else's hardware.

I'm glad that someone (thanks jcmarcos) understood that my main point was not to denegrate Genesi.
Rather, my point was that the MorphOS development team has done quite well without Genesi and that I was very pleased with their decisions and the developments they've made to the OS.

@ jcmarcos

Quote:

If you consider the target audience, it is not surprising at all.

Compared to the MorphZone developer forums, the old MorphOS Developer Connection website or Freescale's developer-centric IMXcommunity.org website, PowerDeveloper.org holds up quite well in terms of forum activity.


@ Jim

Quote:

Not all of it, but former employees and contractors of Genesi & Thendic France did contribute source code that is still being used in MorphOS today. Also, Genesi has licensed third-party components for MorphOS that are still being included in new OS distributions.


Quote:

Actually, the Genesi management had been a proponent of a Mac port. The developers were not "held back" by any third parties to port MorphOS to other platforms.


Quote:

I would like to point out that Genesi has contributed time, money and other resources to keep MorphZone alive and well long after MorphOS became meaningless with regard to the company's commercial ambitions.

Alright, no more negative comments on Genesi.

The fact that anyone is willing to invest in PPc hardware for Amiga derived OS' is amazing.

If I can spare kind words for Treavor and A-eon, then I should acknowledge Genesi's support for MorphOS.

I'm still quite pleasently surprised at how well things are developing without a hardware partner.

Nice videos from this event, you rock guys!

> those failed ventures are publicly known.

Yes, I could write a novel about failed Genesi projects ;-) I was more aiming at the "tied to some third party agreement that was never fulfilled" part.

> The biggest one was THTF's betrayal

Okay, so I take it THTF is that "third party" you mentioned then.

> history shouldn't be documented by a single guy in a forum (like me).

What's wrong with that? Better by a single guy than by no one :-)

> Cherrypal did what Genesi couldn't.

And so did Freescale as well as Levy and Watson from Linkbook.

> Compared to [...] Freescale's developer-centric IMXcommunity.org website,
> PowerDeveloper.org holds up quite well in terms of forum activity.

That comes as a surprise. I'd have guessed it's the other way round.

Jim

Trevor / Aeon can never compare to what genesi has done. Maybe in some future dream time :) It wasnt a personal attack I'm just not going to allow non factual statements be made to a company who has helped much to get morphos to where it is today.

@ Andre

Thanks for posting some truths :)

and now back on topic!

Don't sweat it magnetic, I don't take anything personal. I just felt that during my contacts with Matt that he wasn't too thrilled about hearing how far I'd gotten reworking Freescale's MPC8640/8641 design. And I kept having people tell me that the ATI SB600 Southbridge I was basing part of the design on wouldn't work because it wasn't a standard PCIe X4 component (it is - the X1000 uses one).

But the MorphOS development team's adoption of G4 Macs kind of rendered that project pointless. And that development made a lot of sense. It just seems like the development team's judgement has been pretty good lately and their track record of keeping their promises is reassuring and impressive.

I think too much focus has been placed on the perceived negative coments I made while the main point I was trying to present was how encouraging recent developments have been.

Anyway, great photos, good news, I'm happy to see how things are progressing (again R400 support - yes)..

Well, Matt is probably a bit special. And he was quite strongly promoting an ARM port of MorphOS. And I think generally that wouldn't be a bad thing. *IF* time and resources were huge that is. With given resources it was the best approach to focus on used ppc Macs for now. But once there will be a time when used ppc Macs won't be too sexy anymore. And what will come then is pretty unforseeable yet. Will ppc offer a future? I think it can (Freescale's pushing it again), but I wouldn't bet my last pair of trousers on that. And if ppc fails or nobody makes useable hardware?
ARM is probably an easier target for MorphOS than x86 (because of the endianess, which seems to fit according to some trustworthy ppl¹). Remains to be seen whether ARM gains some shares of the netbook, tablet, notebok or desktop market. So far ARM hasn't been too successful on this (well, the iPad is ARM; RIM's playbook, too ). No matter how much ARM was hyped during the last two years, x86 is still and by far #1 in that domain (exept recent tablets).

Anyway, MorphOS does fine with and without Genesi. Hardware shortage is no issue for the next few years. But I don't think it would be of any disadvanatge if MorphOS and Genesi will cooperate again in some yet uncertain future.

--
¹ still I'd like to see a definitive answer to the question I raised long ago "What would happen if a 68k app running on a hypthetical MorphOS ARM with transparent 68k emulation would try to access, say, 0x00000004? Would it get a pointer to exec base or would it get elsewhere?"

Umm Matt S aka Neko is not the best representation of what Genesi is, was , or does imo

ARM has some really positive points. Low cost, low power draw, wide support. I've already mentioned the PandaBoard in a thread (replying to Andreas). And there is a Chinese firm that is planning to release a 2 Ghz ARM processor (actually an Soc w/ built-in video) next year.

But if MorphOS goes ARM, do we retain PPC compatibility? And if so, how (when ARM may not be powerful enough for emulation overhead).
Would that mean a parallel development course where apps from PPC and ARM ports could not run on the other platform w/o recompilation?

And X86 has one major flaw. There too much hardware to support all X86 devices (there just no way to write that much driver code). So the developers would have to select what the platform would feature. Complicated, but not impossible.

Great debate guys! It deserves a special mega-multi-reply!

Quote:

That's new to me, thanks for that detail (why is this the first time i read it?)

Quote:

Sure, you can say a lot of good things about Genesi. For the bad ones, they are not to guilty rally, but other parties. But then, it's strange that some ventures failed, after seemiong perfectly firm, and announced as that. But then, this is the computer business, a can of worms anyway.

I just wonder why Genesi has never had the same luck others enjoyed. Take for example that joke company Cherrypal: They are in business!

Quote:

That's a shame. I'd love a novel about success, not failure. But I guess that's what you get when your have your roots in Amiga!

Quote:

Got more info for that? What did these achieve exactly?

Quote:

And mopre details for this? How close you came to making a new computer? Full schematics?

Quote:

Both happy and sad at the same time!

Quote:

Alright, he is not here to defend himself, but I wonder why you say that. You mean technically wise?

Quote:

That's interesting, of course. But without a budget, it makes little sense.

Quote:

If it has, will it be ahead of ARM's future (which is, actually, present)?

Quote:

MorphOS could attempt then a port to a completely new platform (huuuuge work), or stop developing, leaving us happily using it on aging computers. That's not new for an amigan!

Quote:

I never, never understood the problem with endianess, no matter how many times they've explained to me. I might be completely stupid, but operating systems do exist both on little and big endian CPUs. It's a matter of programmers having endianess in mind when exchanging data with peripherals. For example, the whole TCP/IP stack is big endian, and MorphOS has one...

Quote:

And everybody writing both a network-stack or an app accesing it knows how that data is presented to him....

Thats not an issue.

All MorphOS and AmigaOS apps were written in big-endian, running them directly on a little-endian OS is impossible.


Amithlon had both an big-endian x86 gcc loosing about 50% of host performance and little-endian gcc that could only be used to compile modules similar to PowerUP.

For existing apps there are also 2 ways:

a) you run the OS big-endian giving you not just the gcc-problem explained above but also an OS that itself runs at supoptimal speed.

b) AROS were those apps only run under full emulation (UAE)


Sure that is Emumiga trying to cut the gordon knot by running big-endian code against a little-endian API, but I predict it won't get much farther than running "clock".

> there is a Chinese firm that is planning to release a 2 Ghz ARM processor
> (actually an Soc w/ built-in video) next year.

Don't forget Freescale's 4+ GHz ARM roadmap for 2011 ;-)

>> the Genesi management had been a proponent of a Mac port.

> why is this the first time i read it?

Because apparently you missed this:

"MorphOS [...] Runs on Macs by the end of 2003. This will be marketed as a special application for Mac that does not "run" on the host hard drive (but can). We will sell the application, the OS is just there to make it work or just sell the OS. [...] When things are really stable, then there is the mass-market beginning with the sale of the OS/Application to Mac Users..."
http://www.morphos-news.de/index.php?lg=en&nid=199&si=1

"Genesi came up with an idea of developing a new game. This is considered in a possible cooperation with producer Elie Chouraqui. It might then be possible to use this game as a "Trojan Horse" for the Macintosh version of MorphOS."
http://www.amiga-news.de/en/news/AN-2003-06-00140-EN.html

> I just wonder why Genesi has never had the same luck others enjoyed.
> Take for example that joke company Cherrypal: They are in business!

As far as I can see, Genesi is as well (in business, that is ;-)

>> I could write a novel about failed Genesi projects

> That's a shame.

A shame for Genesi I hope, not for me ;-)

> I'd love a novel about success, not failure.

I *could* write that one too. It's always a question of what it is that the writer of a novel wants to portray (or what he is asked for to write about). A comprehensive novel would include both failures and successes of course. My snide remark was just an answer to your mention of Genesi's "publicly known failed ventures".

>>> Cherrypal did what Genesi couldn't.

>> And so did Freescale as well as Levy and Watson from Linkbook.

> Got more info for that? What did these achieve exactly?

Doing the same thing Cherrypal does: Reselling THTF hardware.

Freescale: https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=7014&forum=11&post_id=72231#72231
Linkbook: https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=6726&forum=11&post_id=73080#73080

>> Matt wasn't too thrilled about how far I'd gotten reworking
>> Freescale's MPC8640/8641 design

> And mopre details for this? How close you came to making a new computer?

https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=6768&forum=3&post_id=69347#69347
https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=7001&forum=3&post_id=74513#74513

> he is not here to defend himself

"Last Login: 2010/11/12 7:38"
https://morph.zone/userinfo.php?uid=48

> I never, never understood the problem with endianess, no matter how many
> times they've explained to me. I might be completely stupid, but operating
> systems do exist both on little and big endian CPUs. It's a matter of programmers
> having endianess in mind when exchanging data with peripherals.

https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=6726&forum=11&post_id=77338#77338

Lots of interesting info and good news in this thread! Wow!

Also thank you guruman for your extensive summary!

Quote:

No schematics or easily exchanged material. Mostly stuff like hand written notes listing pin equivalents between ULi Southbridge and ATI SB600 and changes necessary to rework that part of Freescale's design, specific layout info (like the BGA pad masks), and a lot of literature with highlight, personal notes, and links to other documentation or personal notes. Its all very scattered and disorganized (plus there's LOTS of it).

At the time I dropped it, it was trying to get past what I thought was a hurdle. The Freescale design uses a programmable logic component and for some reason I convinced myself that I had to now how the component was programmed to re-implement this part.
Actually, I was making this more complicated than I needed to. After a few questions from Andreas, I relized that I didn't need to know how that part of Freescale's design worked.As the component was obviously used to simply the design (mainly for glue logic related to timing) all I needed to know what it did and then I could either recreate it or design that part with descrete components.

I probably shouldn't have mentioned the entire project as I didn't follow trough with it. I only contacted a few people to bounce ideas off. Adreas was one, and I don't think he relizes how he help me examine certain things differently. Another person (that I got some advice from) was the head of the firm that designed the X1000's motherboard.
That last contact was paticularly interesting. While that firm had experience with PA Semi's PA6T, the processors they recommended were from Freescale (some of the predecessors to the 5500s that Andreas has mentioned).
Andreas ccould probably dig up thatreference easier than I could.

Two things I could have done differently. First, I bought the same layout and design package that an company I have a relationship with in PA was using. It never occured to me to consider that I could just ask Freescale for their files. When I got around to that I found out that the package I was using and the format they had their layout in wasn't compatible (and now that I think about it the $4000 cost to buy that package and access their files would have been worth it).
Second, instead of investing months in re-inventing the wheel, I could have found a partner that had already worked on a similar design (which probably would have been the same company Treavor used - although at the time I was doing this I had no idea I was consulting the same people).

Its funny, since he has always been better than me a putting things together and coming up with the right conclusions, Andreas wasn't as surprised as I was when I found out who designed the Xena.

And there are still some people in the amiga community who think Ack Systems had something to do with that or that its more vaporware (with the record that new introductions in the Amiga community have I can understand their scepticism).
My guess on the reaon the X1000 has been delayed? It probably Hyperion's fault.. That port is a pretty big project for two guys.

Great info Jim! I think your efforts, while not finalized, still have a value. Wait... you put FOUR GRAND into that? Wow. By the way, who's the man behind Xena? Some old chap that still has a Transputer t-shirt? Why did it surprise you?

I guess I wasn't clear (a problem with many of my posts), the $4000 was for a package that supported the file format Freescale was using. The package I had was a fraction of that (although still painfully expensive). Besides, the software I have produces files that the company I wanted to use to build boards could process with their own software.

@ Varisys (the company behind the X1000's motherboard) my primary contact was withPaul Gentle (although he did reference me to another employee for a few questions about the SB600 and the MPC8640/8641).

And no, I don't see following this path. I gave it up over half a year ago. A Powermac with any processor over 1.25 Ghz would outperform it in many areas (most actually except for memory bandwidth and video card support/speed).

A clean start with an Soc like the eventual Altivec enhanced succesors to the current e5500 core would be cool. Even Paul tended to favor Freescale's QorIQ communications oriented processors.
Andreas (I think) still has an open mind about Applied Micro and its Titan sucessor. But after all this time waiting, I'm not taking any bets on that company.

Finally, if something were to be designed it would either have to be led by an experienced company (like Varisys) or by a larger group of people. Before this I hadn't approached anything nearly this complicated (my last experience being a 68030 based project - and that seemed complicated compared to 40 pin DIP processors).

The only thing I walk away with after this is a better understanding of some of the design elements used in modern motherboards. The use of programmable logic components, the importance of lead lengths and paths as they retlate to timing constraints and cross-talk, and the use of BGA components (when my company last built circuit boards use of surface mounted components was much more limited), not to mention a host of other small details.

Creation of new motherboards feels like its an order of magnitude harder than it was in the '80's. If I ever sounded like I didn't admire Genesi or A-eon/Varisys' work, let me clear that up. Anyone that can design and produce a functional board has my respect.

One other thing I should have remembered. We've been buying other company's boards for a reason. Design and small scale production aren't economical.
That's why the Macs make sense (for MorphOS) as opposed to A-eon's approach.
And that's why if there isn't a new producer of PPC boards, moving to ARM might make sense.