Interview questions discussion thread

[Seehund]

mikeymike wrote:

The purpose of the "dongle" (AFAIK some code in the BIOS of the A1 that OS4 detects and allows itself to be installed/run) is for a number of reasons, take your pick:

 * The smaller the number of configurations that OS4 can be installed on, the less development and testing work that needs to be done.

Which of course is not a relevant reason, and IMO a ridiculous excuse.
You don't need market restrictions and licensing requirements placed on third party hardware to decide what hardware that your software will run on, nor to simply state to your potential customers what hardware that your software is compatible with. Hardware X in its "normally" sold form is identical to Hardware X when/if it's sold under another trademark license.

 OS4 isn't likely to set the world alight, so it makes perfect business sense to get as large a return on as small amount as possible.

It's apparently been decided that there should only be an as small as possible amount to get a return from. I.e. the return on AmigaOS can only come from the few people who are prepared to buy a Teron via Eyetech (or other licensed hardware sold by other licensees - yeah, right...), plus a bunch of Amiga owners with CS-PPCs.

 * Presumably Eyetech and Hyperion have or will have some kind of partnership, and as there's more money in selling hardware than software, both companies can have their fingers in both pies.

I don't think that anyone, not anyone who's said anything "officially" at least, would claim that there's more money to be made from hardware than from software. Larger gross income per unit, yes it's possible (at least when we're discussing consumer hw/sw), but not as a general rule to build a business plan upon.

If Eyetech/Hyperion decide to go with a completely different chipset and say the G5 in future, then Hyperion will develop the OS to cater for that as well.

That "Eyetech/Hyperion" pair is a problem. Whether there's a licensee for the hardware in question should be irrelevant for development decisions IMO. If there is a willing licensee, then great, but that should not be the stumbling block. I think the question should simply be "is a port technically and commercially feasible?" That should cover it, just as it always has for most software publishers.

Replace "[whatever hardware trademark licensee]/Hyperion" above with "Hyperion or third party driver developers", and we're set. It'd be nice if "Amiga, Inc." would fit in there too, by actually having a solid vision, set goals, some common sense and some interest in the future of AmigaOS, but everything we've seen from them seems to be shortsighted haphazardly concocted schemes to rake in some small but quick cash from very small or no efforts/investments/products (licensing fees, distribution deals, club memberships/pre-payments).

[mikeymike]

The purpose of the "dongle" (AFAIK some code in the BIOS of the A1 that OS4 detects and allows itself to be installed/run) is for a number of reasons, take your pick:
* The smaller the number of configurations that OS4 can be installed on, the less development and testing work that needs to be done.

Which of course is not a relevant reason, and IMO a ridiculous excuse.

Ok, let me put this to you simply.  Given a very small budget and that a particularly large return is unlikely, would you like to design a piece of software that works on one platform, or every desktop computer platform, mobile phone, and toaster?  Which seems like the less work to you?  Think of the amount of extra testing that has to go in, the amount of extra code to handle compatibility for each platform, even more extra bits of code to handle strange quirks of each platform.  And we're not talking in this case about just any old bit of software, but AN OPERATING SYSTEM.  The thing that has to handle loads of different types of hardware already despite the "supported hardware" standards imposed by Hyperion/Eyetech.  Operating system design is a monumental task, even for one platform and a very specific hardware set!  If you think it is so easy, where is the AmigaOS port you've been doing in your spare time that runs on all hardware platforms perfectly?
If you don't have some restrictions, customers will complain that your software doesn't work on their hardware setup, and since you didn't specifically knock on their hollow skull and tell them that it wouldn't work, it is totally your fault.  Not all operating systems makers have Microsoft's R&D budget.

It's apparently been decided that there should only be an as small as possible amount to get a return from

Or maybe they're aiming at a market that is realistic for them to aim at, at least to start off with.

I don't think that anyone, not anyone who's said anything "officially" at least, would claim that there's more money to be made from hardware than from software.

No, no-one "officially" will admit that.  Why do you think Sun Microsystems and Apple are primarily hardware makers?  Their OS comes secondary.  Microsoft are the only ones who have made it work any other way, and they played the game exceptionally well, as well as throwing in a large number of low tactics that should be illegal.

Microsoft don't like to admit that their main income is from MS Office, but it is.

Now please think about this.  An operating system can typically cost say ~$100, anywhere up to that and a little bit beyond.  There are few exceptions to this.  Now, how many man hours go into making that $100 product?  How many buyers is AmigaOS4 likely to net?  Now think about the hardware.  The variation in costs is so much bigger, all kinds of things can be done to the setup to make it cheaper or more expensive, and generally to make a much fatter return on the product.

That "Eyetech/Hyperion" pair is a problem. Whether there's a licensee for the hardware in question should be irrelevant for development decisions IMO. If there is a willing licensee, then great, but that should not be the stumbling block. I think the question should simply be "is a port technically and commercially feasible?" That should cover it, just as it always has for most software publishers.

I'm not sure what you're trying to say here.

[shIva]

@mikeymike :

to my first questions : it´s clear, that os4 is  not intentioned to be a unix box, but things like ssh2, etc. are not really "only unix" anymore. these features are a must have for implementing such a system into an existing network.
for further versions a multiuser file system (afaik mufs is availible) is very important.
documentation : i´d really like to see a documentation to the system like sun (like this one) does it.

it would be very helpful for developers to have a full documented system. i know that doing a documentation like the one above is nearly impossible in normal time scales, but sth. similar would be nice.

[Seehund]

mikeymike wrote:

* The smaller the number of configurations that OS4 can be installed on, the less development and testing work that needs to be done.

Which of course is not a relevant reason, and IMO a ridiculous excuse.

Ok, let me put this to you simply.  Given a very small budget and that a particularly large return is unlikely, would you like to design a piece of software that works on one platform, or every desktop computer platform, mobile phone, and toaster?

I don't quite see why you bring those two options/scenarios up, and what it's got to do with the OS4+ distribution policies. Has someone suggested that an OS should be made to somehow automagically work on every concievable kit? Where did that come from? The distribution policies don't determine what hardware is compatible, technical issues do. Either the OS supports a particular piece of hardware, or it does not, dongles and trademarks notwithstanding.

I thought I[/i] put it simply when I wrote this:
"You don't need market restrictions and licensing requirements placed on third party hardware to decide what hardware that your software will run on, nor to simply state to your potential customers what hardware that your software is compatible with. Hardware X in its 'normally' sold form is identical to Hardware X when/if it's sold under another trademark license."

Even simpler: Imposing market restrictions has nothing to do with compatibility issues.

I just don't understand why you seem to propose that the alternative to trying to regulate a 3rd party hardware market would be to have the product in question (OS4+) run on every imaginable hardware.

If you don't have some restrictions, customers will complain that your software doesn't work on their hardware setup, and since you didn't specifically knock on their hollow skull and tell them that it wouldn't work, it is totally your fault.

The restriction should be compatibility, and it's usually presented in a hardware compatibility list and on the frigging box. Millions of people can handle things like this daily. They don't install WinXP on a C=64, because the C=64 isn't listed as compatible by MS. They don't refuel their cars with diesel oil if they're told that the car runs on petrol. If they still try these things, they'll notice soon enough that things won't work, it's their own fault, and they can't blame MS, Ford, et c.
There's at least one thing that's certain about the hardware market restrictions dragging down OS4+, they're not there to protect AInc/Hyperion from lawsuits from one or two illiterates unable to read a HCL, who might try to install the OS on incompatible hardware. That's a bizarre notion.

It's apparently been decided that there should only be an as small as possible amount to get a return from

Or maybe they're aiming at a market that is realistic for them to aim at, at least to start off with.

The worrying thing is that they have painted themselves (and AmigaOS) into a corner. It is not possible to aim for a bigger market. It has been decided that the market for AmigaOS is people who are prepared to buy a Teron from Eyetech.


Thinking about this, I recall a post by Ben Hermans on amiga.org (IIRC), that gave an impression of outright defeat - to paraphrase from memory it was something along the lines of "we don't have the infrastructure to handle a bigger market, and that's [one of] the reason for these hardware market restrictions." What kind of planning is that? They don't foresee a huge initial interest, so they'll actively work to reduce the potential market and its growth! Is this another one of those AInc "strategies"? It'd be severely disappointing if Hyperion themselves stood behind this idea. It makes "AmigaOS4 and all future versions" (as AInc says) look like a futile exercise, like artificial respiration for and a meaningless gesture toward us few already existing AmigaOS users and a hardware vendor that otherwise would've been made redundant by evolution and the death of the "Amiga hardware" concept.

If you don't dare to aim high at the moment, then at least don't weld a block to the carriage of your gun to prevent yourself from aiming higher later on. If the current infrastructure really should prove to be insufficient due to high demand, then by all means invest the greater income that this demand generates in better infrastructure and personnel!

Sometimes everything seems to be backwards and upside-down with anything even remotely related to AInc.

I don't think that anyone, not anyone who's said anything "officially" at least, would claim that there's more money to be made from hardware than from software.

No, no-one "officially" will admit that.  Why do you think Sun Microsystems and Apple are primarily hardware makers?  Their OS comes secondary.  Microsoft are the only ones who have made it work any other way, and they played the game exceptionally well, as well as throwing in a large number of low tactics that should be illegal.

Microsoft don't like to admit that their main income is from MS Office, but it is.

Now please think about this.  An operating system can typically cost say ~$100, anywhere up to that and a little bit beyond.  There are few exceptions to this.  Now, how many man hours go into making that $100 product?  How many buyers is AmigaOS4 likely to net?  Now think about the hardware.  The variation in costs is so much bigger, all kinds of things can be done to the setup to make it cheaper or more expensive, and generally to make a much fatter return on the product.

If the general statement "there's more money in selling hardware than software" were true, then everyone who's selling software today should really be selling hardware. Anyway, Hyperion does not sell hardware. Nobody develops or produces hardware for AmigaOS. AmigaOS and its users should IMO not be seen as a means to subsidize any one particular hardware distributor just because they once used to sell a nowadays dead computer platform and are willing to rename new hardware (at an officially undisclosed but rather easy to guess extra cost to AmigaOS customers).


Hyperion makes money from the number of sold copies of AmigaOS. If they actually were to get a cut of the hardware price (you presumed that there is or will be some kind of partnership between one licensee, Eyetech, and the OS producer), there'd be yet another rotten thing in the state of Denmark; exactly who do they expect to be interested in getting an "Amiga" license for their hardware if part of the hardware price goes to the OS vendor - in addition to the AInc "Amiga" trademark license fee?
The price we pay for a copy of AmigaOS is no different (or not much different) if we buy it shrinkwrapped or bundled with a licensed Teron.

Anyway, the potential market for AmigaOS is and will be larger (by how many orders of magnitude is up for speculation) if sales of the OS is not made dependent on simultaneous sales of bundled hardware from "special" dealers. AOS simply must be ported to more hardware over time, and in order to do that it must (also) be sold independently from that hardware and be possible to buy AOS for the same hardware regardless of who sells it and regardless of whether the vendor calls it "iMac" or "iAmigaOne". Today it's "no licensee = no port, no matter how feasible it is from technical and commercial viewpoints".

There are some 10^4 times more units of one single Mac model (like the "iLamp") already in use out there than there are currently used licensed Eyetech-sold Teron boards, and more are made and sold every day, not to mention the second hand market. There are probably 3 - 10 times as many people who publicly protest against the restrictions imposed on AmigaOS and its hardware base as there are current Teron board users. There's probably twice the amount of discontinued Pegasos 1s out there, with a new model around the corner, both arguably better marketed to a wider audience and cheaper than the Terons. Just to put some perspective on things.

It all comes down to cost/profit/volume analysis.
Maybe all this is a scheme to make such analysis easier, by ignoring the volume factor...

That "Eyetech/Hyperion" pair is a problem. Whether there's a licensee for the hardware in question should be irrelevant for development decisions IMO. If there is a willing licensee, then great, but that should not be the stumbling block. I think the question should simply be "is a port technically and commercially feasible?" That should cover it, just as it always has for most software publishers.

I'm not sure what you're trying to say here.

That it's a stupid idea for an OS vendor to make the market, hardware base, development, commercial attractiveness and entire future of your OS ultimately dependent on whether or not there exists willing licensees for third party hardware. Because no such licensees exist for any hardware other than Terons, and even if they did, it would be stupid to only allow your customers to buy their hardware from such licensees.

[HyperionMP]

Seehund, if you want to use quotation marks, at least get the quote right.

All I said was that Hyperion does not have the distribution infrastructure to distribute a significant amount of OS copies.

We are a development company, not a distributor or dealer with warehousing facilities.

Which is why we are shifting the distribution to other parties via an OEM arrangement.

This OEM arrangement also guarantees us that we are exposed only to the credit risks associated with the OEM, not to countless dealers all over the world.

It also allows us to negotiate more favorable shipping conditions if we can have all copies sent over to the same location.

It also helps to cut down on piracy.

[HyperionMP]

>There are some 10^4 times more units of one single Mac >model (like the "iLamp") already in use out there than there >are currently used licensed Eyetech-sold Teron boards, and >more are made and sold every day, not to mention the >second hand market.

Your point being? Mac hardware is branded uniquely Apple through its design.

Moreover, this would require us to support this hardware with drivers for the onboard hardware which would effectively see us wade through LinuxPPC source-code as Apple certainly won't release chipset documentation to us.

All of this for a completely uncertain return.

Voodoo economics at its best.

>There are probably 3 - 10 times as many people who publicly >protest against the restrictions imposed on AmigaOS and its >hardware base as there are current Teron board users.

Pure speculation on your part. You have no numbers to back up this claim.

Given the fact that AmigaOS was always sold with the hardware when the hardware was still being produced, I doubt this very much.

>There's probably twice the amount of discontinued Pegasos >1s out there, with a new model around the corner,

Nonsense. No more than 600 Pegasos I units were ever produced and the AmigaOne sales have already overtaken that number since quite some time.

Moreover, I doubt very much that a sufficient number of those people would be willing to buy a copy of OS 4 to merit the development costs.

>both arguably better marketed to a wider audience and >cheaper than the Terons.

You are talking about a discontinued product and a product which only exists in the form of an announcement.

If you think I'm going to bet the shop on that kind of nonsense, dream on.


>Just to put some perspective on things.
>It all comes down to cost/profit/volume analysis.
>Maybe all this is a scheme to make such analysis easier, by >ignoring the volume factor...

You lack key data such as:

1. development cost of OS 4

2. per unit license fee to Amiga

3. cost to support an additional platform

In short, your economics are way off and your conclusion particularly misguided and frankly, based on nothing but gut feeling rather than solid facts.

[Seehund]

HyperionMP wrote:

Seehund, if you want to use quotation marks, at least get the quote right.

All I said was that Hyperion does not have the distribution infrastructure to distribute a significant amount of OS copies.

As I said; "to paraphrase from memory it was something along the lines of ..."

The exact quote would be:

"We don't want to be involved in shipping end-user copies to dealers for several reasons:

1. This would put a strain on us as we would need to put into place an infrastructure for shipping, warehousing and distribution which goes beyond what we have now for selling our games."

And I've replied to that before:

(So AInc is supposed to do NOTHING but get payed for a license of their(?) IP to Hyperion? What a wonderful deal...)

If you expect to sell that many separate copies that you'd need further investments in infrastructure, I say GO FOR IT[/size] damn it! :-O
Especially considering that you won't get any relevant hardware vendors to apply for a distribution license.
Don't be afraid of success, and AInc shouldn't do everything to prevent success.

We are a development company, not a distributor or dealer with warehousing facilities.

Which is why we are shifting the distribution to other parties via an OEM arrangement.

Heh, I suppose that's one way of putting it.
Another way could be "people won't be allowed to buy the same hardware from anyone they like, and the OS must be actively prevented from being ported to better, cheaper and more widespread hardware, because the OS developers don't have enough shelves in their offices." It sounds equally silly.

If you (or AInc, or Eyetech...) honestly say no to that vast number of OS sales you're hinting at because you don't expect to be able to handle distribution, then the obvious solution IMO wouldn't really be to RESTRICT SALES. It would be to let others handle distribution of those separately sold copies (or to use some of the increased income to improve one's own distribution).

This OEM arrangement also guarantees us that we are exposed only to the credit risks associated with the OEM, not to countless dealers all over the world.

So? It's the exact same thing as choosing a distributor for additional shrinkwrapped copies of your product.

It also allows us to negotiate more favorable shipping conditions if we can have all copies sent over to the same location.

See the paragraph above. Plus, motherboards are more expensive to ship than boxes with a CD and a booklet in them, or sending ISO images and PDF files to printers/distributors. And then the irrelevant hardware distribution costs don't affect the cost of your product, AmigaOS.

But, really, shipping costs?! I haven't even heard anyone from AInc bringing up things like that to justify compulsory licensing/bundling/dongling of hardware.

It also helps to cut down on piracy.

I still don't understand how anyone can claim that.
So, under the currently presented policies, what if not even one single "unlicensed" piece of hardware ran a pirated copy of AmigaOS (which of course is a utopia)? It still means that not a single shrinkwrapped copy of AmigaOS can be bought and payed for. Not even a hypothetical gain.

Your point being? Mac hardware is branded uniquely Apple through its design.

Yes, and? Does that brand stop other OSes from running on it or being sold for it? I could swear I bought YDL for that Powerbook over there, and I could swear that the US Navy recently bought new Apple XServes with Black Lab on them.

The Amiga hardware concept is fortunately dead and buried. There's no "Amiga hardware market", for no good reason at all, based only on a brand. Oh, wait... Scratch that.

My point is that there's lots of suitable hardware already out there, and there will always be coming new hardware, but with the licensing/dongling/bundling-only policy, AInc has turned down any and all of those options for AmigaOS. AmigaOS is stuck with only licensed hardware from licensed vendors, and for any forseeable future that means Eyetech and apparently whatever hardware they can buy.

Moreover, this would require us to support this hardware with drivers for the onboard hardware which would effectively see us wade through LinuxPPC source-code as Apple certainly won't release chipset documentation to us.

Yes, that's pretty obvious. Hardware needs drivers, no matter if somebody's got a license to sell that hardware using the "Amiga" name or not. Should sales of separate OS copies be allowed, then either you or a third party could write those drivers, and they could be distributed with the OS or by the third party. Should the compulsory hardware trademark licensing stay in place, then there's no point for anyone to even consider writing drivers anyway, unless a hardware licensee appears.

All of this for a completely uncertain return.

Voodoo economics at its best.

Perhaps the "market research" method employed by AInc to decide that the max return/investment ratio from AmigaOS is gained by only making it available for and sold together with Teron motherboards distributed by Eyetech could fit in here somewhere. The insane lottery / unspecified surprise preorder / corporate fanclub membership thing.

>There are probably 3 - 10 times as many people who publicly >protest against the restrictions imposed on AmigaOS and its >hardware base as there are current Teron board users.

Pure speculation on your part. You have no numbers to back up this claim.

There are about 1000 protestors. My speculation was about the number of Terons in use by Eyetech customers today. You now say that this number is more than 600. Perhaps you could provide a more useful number? And what factual numbers - not "pure speculation", of course - did AInc use when they imposed the restrictions on all future versions of AmigaOS?

>There's probably twice the amount of discontinued Pegasos >1s out there, with a new model around the corner,

Nonsense. No more than 600 Pegasos I units were ever produced and the AmigaOne sales have already overtaken that number since quite some time.

While we're juggling numbers in the hundreds range and ignoring the essence, how many licensed Teron CX's were in use by Eyetech customers when it was chosen as the only hardware platform (apart from Amigas) for AmigaOS, and how many were speculated... sorry... expected to be sold? How many were delivered by Eyetech before they saw it being discontinued?

Moreover, I doubt very much that a sufficient number of those people would be willing to buy a copy of OS 4 to merit the development costs.

(Is this where I accuse the poster of "pure speculation", while not providing any disproving numbers? )

It'd be very interesting if you could please tell us what kind of minimum profit you require to support a piece of hardware, and what you think those numbers are for some often discussed hardware, like the Teron CX and PX, the Pegasos I, or a common Mac like a new PBG4.

You are talking about a discontinued product and a product which only exists in the form of an announcement.

If you think I'm going to bet the shop on that kind of nonsense, dream on.

I do NOT think you should do that. They're just options. Or rather, they could be options, if the current licensing/bundling/dongling wasn't compulsory.

But somebody's betting the shop for you and AmigaOS already, it seems.
A discontinued product? Teron CX. It was ageing fast already at its release in 2001, and later when it was presented as "AmigaOne SE". As the PX is basically the same product but with newer CPUs, it's not very speculative to say that it's not going to last for very much longer.
A product that only exist in the form of an announcement? I remember Alan Redhouse announcing the forthcoming smaller Terons already as "AmigaOnes" (i.e. they're supposedly licensed hardware that has passed AIncs rigorous testing with AmigaOS...), at Amiwest.

The thing is, your software and your shop has been bet on this hardware and this dealer. There can be no options. A totally unnecessary and scaringly counterproductive additional dependency for you and AmigaOS, IMO.

You lack key data such as:

1. development cost of OS 4

Since you separated this from point 3 in your list, this is irrelevant. It is developed (well, in development), and further sales can be nothing but good.

2. per unit license fee to Amiga

As a curiosity, it'd be fun to know what this is, since you felt like mentioning it in the first place. But the fact that it exists at all is of course bad, since it works against the plan that AmigaOS is made dependent upon: getting more licensees. And since you say "Amiga" [Inc.], am I wrong in presuming that this trademark tax would not be funneled back to you and AmigaOS development?

3. cost to support an additional platform

That's what counts of these points. And, since you'd presumably not make the port solely for ####s and giggles: 3b. Possible extra profit, a wider potential hardware install base and a more attractive AmigaOS from supporting (or allowing others to support) more platforms regardless of hardware vendor license status.

As you point out all of us here except you lack key data like point 3, so could you please share that information? I think it was Steffen Häuser that said thanks to the HAL, porting would generally be a matter of a couple of weeks. No, I don't think a specific platform was mentioned (input from MagicSN appreciated!). If we use the Pegasos I solely for sake of example, how many copies would need to be sold to recoup the cost?

[mikeymike]

@ shIva

to my first questions : it´s clear, that os4 is not intentioned to be a unix box, but things like ssh2, etc. are not really "only unix" anymore.

They might be as standard on MacOS X, which has a FreeBSD kernel IIRC, which makes me think it's at least part-unix box :-)  But otherwise, up to the present they're a unix-only thing.

for further versions a multiuser file system (afaik mufs is availible) is very important.

So do I, which is why I submitted the question I did :-)  Though some people vehemently disagree that AmigaOS should go in that direction.

documentation : i´d really like to see a documentation to the system like sun does it.

So just "Is there going to be any online documentation for OS4?" or is there something particular you like about how the Sun docs site does it?

There's developer reference and user reference.  You'd like to see both online presumably?  I don't see anything wrong with that, just clarifying your opinion :-)

[mikeymike]

@ Seehund

Ok, going straight to the point (what I believe is your point anyway), say for example if there was a HCL (which I think there should be) and it said "this won't work on the Pegasos or any Apple motherboards", don't you think the response would be "don't you go oppressing me!" from the particular sect of "I love MorphOS and hate AmigaOS"?

Say if Hyperion/Eyetech never mentioned the dongle and just did what I said in my example in the HCL, would that make anything better?

And as for quality arguments, Apple and Sun are two companies known for restricting their OS to their hardware (except for Solaris x86, but that's not really a product Sun is trying to push).  Both companies have good to excellent reputations for OS-hardware stability and reliability.  Windows does not.  AmigaOS has always been on proprietary hardware and has an excellent reputation for stability/reliability there.  UNIX variants have pretty good hardware support, and IMO are generally more stable than Windows because of their simpler design and avoidance of "house of cards" style putting one technology on top of another the way Windows does.

I didn't see anyone complaining when Workbench 1.0 or anything released since until OS4, that it only ran on proprietary hardware.

You may think I've avoided your questions to a certain extent, I did, basically because it comes down to you thinking it's a stupid idea to do what Eyetech/Hyperion is doing in restricting the hardware platform and I don't.

Personally I think if AmigaOS4 and subsequent releases were successful enough, then it would be wise to open the OS to other PPC platforms provided the software didn't need significant reworking.  I don't see that it is essential to its success though, I think the direct opposite, such ambitious aims could be the death of the project at this stage.

From what little I know of MorphOS, it is effectively limited to a very small range of hardware, just that you're finding out in a different way and you feel you're not being "oppressed".

[olegil]

But the feature list on amiga.com was written by Ben Hermans, so why ask him when you don't trust the feature list he wrote?

Matt, you're not making much sense here...

[mikeymike]

I really hope this doesn't become quoted scripture for those arguing against the decision to limit the platform OS4 runs on.  The "answers" are someone's points of view on the situation at best, certainly not undeniable arguments.

[olegil]

The bit about MOS is interesting. It doesn't currently run on Macs either, but do we see Seehund nagging about it everywhere? No. Why is that?

[olegil]

mikeymike wrote:
I really hope this doesn't become quoted scripture for those arguing against the decision to limit the platform OS4 runs on.  The "answers" are someone's points of view on the situation at best, certainly not undeniable arguments.

And the AmigaOne isn't "POP" either, so the whole debate starts off at the entirely wrong track. IBM dropped their Linux work for the Teron/AmigaOne boards when MAI decided to go for PPCBoot instead of a variant of OpenFirmware. Since OS4 at present will not work on any POP system, it's a bit silly to demand that it should work on ALL POP systems. IMNSHO.

OS4 hasn't even been released AT ALL, so it's ENTIRELY too soon to demand ports. Really. We'll take up the debate again when Hyperion has released something to the end users. If it then seems a good idea to drop the dongle, I'm sure it will be dropped. Right now there's nothing to gain by it. Except that Seehund could buy a non-existant Boxer from Terratec and run a pirated non-existant copy of OS4 on it. And I fail to see where the gain in that is.

[Seehund]

@mikeymike

Ok, going straight to the point (what I believe is your point anyway), say for example if there was a HCL (which I think there should be) and it said "this won't work on the Pegasos or any Apple motherboards", don't you think the response would be "don't you go oppressing me!" from the particular sect of "I love MorphOS and hate AmigaOS"?

Maybe, though why such a sect with an expressed disinterest in buying AmigaOS would complain, and why anybody should care about them, is beyond me.

Allowing separately sold copies (in addition to licensed hw/sw bundles for people interested in that) would leave the door open for ports, either by Hyperion or "compatibility packs" from others if Hyperion wouldn't think a particular platform is worthwile, regardless of whom I the end user buy my hardware from.

Say if Hyperion/Eyetech never mentioned the dongle and just did what I said in my example in the HCL, would that make anything better?

No. But I don't understand what such a scenario has got to do with anything? Should shrinkwrapped copies be sold, then it would make even less sense to require independent hardware vendors to dongle their hardware for somebody else's OS. It wouldn't be a HCL, it'd more like a "HVCL" - hardware vendor and compatibility list. For someone who wants to buy AmigaOS it's not much practical difference from what he sees today.

The problem is not really the dongle per se. It's not the "Value Subtracting Reseller" trademark licensing per se. It's not the OEM bundling per se.

The problem is the combination and effects of these things, and that it's made the only option available if AmigaOS is ever going to be running (legally and payed for, that is) on a particular piece of hardware.

And as for quality arguments, Apple and Sun are two companies known for restricting their OS to their hardware (except for Solaris x86, but that's not really a product Sun is trying to push). Both companies have good to excellent reputations for OS-hardware stability and reliability. Windows does not. AmigaOS has always been on proprietary hardware and has an excellent reputation for stability/reliability there.

Apple, Sun and Genesi sell their own proprietary hardware. They have control over everything from development decisions, component choices and future product lines to the sales and marketing of all this. They make money on selling their own hardware.

AInc/Hyperion/AmigaOS is not comparable to this. The Amiga hardware platform is dead, and nobody makes hardware with AmigaOS in mind. There's no longer any "Amiga" to sell. I say that it's silly to remove the main advantage that this gives AmigaOS and its users - the freedom to pick and choose from a third party hardware market. There's no way one can expect and depend upon this hardware market to come to you and pay for a trademark license and cut off part of their production to an artificially separated "Amiga market".

Quality? First off, I think the choice of the Teron CX evaluation boards as licensed "AmigaOnes" as well as the choice of Eyetech as the only distributor should show in a painfully obvious way just how little this "certification" and AInc's quality, user support and AmigaOS-compatibility assurance/testing is worth. To be precise, it's absolutely worthless.

Quality assurance of hardware is a task for the respective hardware vendors. Personally, I'd trust a hardware vendor, like e.g. Apple, to handle this bit much better than AInc. It's their business. Compatibility with a particular OS is the task of the OS vendor. Either it's fully compatible or it's not. If your OS is fully compatible with hardware X, then say so in the HCL. Quality and compatibility does not improve by trying to control just from whom and under what trademark your OS customers are allowed to buy this hardware X.


Let's play with the improbable; another licensee appears. He wants to sell Apple PowerBook G4s as "AmigaBooks". AmigaOS is ported, and people who want to run AmigaOS on a PowerBook have no option but to buy a hw/sw bundle from this dealer.
Just in what way is the quality of the PowerBooks sold by this dealer any better than PowerBooks sold anywhere else?
What extra means and authority, compared to Apple or other dealers, does this third party have to guarantee and affect the quality?
How is AmigaOS more compatible with PowerBooks renamed to "AmigaBooks", than with the exact same machine sold elsewhere?
What extra control does this dealer have over hardware development, supply et c?

None, none, it's not, and none, respectively. But this dealer would be the only option available if you want to buy AmigaOS.

You may think I've avoided your questions to a certain extent, I did, basically because it comes down to you thinking it's a stupid idea to do what Eyetech/Hyperion is doing in restricting the hardware platform and I don't.

I'd just like to clarify one bit, because sometimes it looks like I haven't made myself clear. Some people seem to think that I want AmigaOS to automagically run on whatever hardware you can throw at it. Of course not. That's against the laws of nature. I think it's a stupid idea to unnecessarily restrict the market for the product in question (AmigaOS), by trying to restrict another market over which the OS vendor has no control anyway. To me it seems like it's what's unnecessary and what's counterproductive that you and I disagree on, mikeymike.

Personally I think if AmigaOS4 and subsequent releases were successful enough, then it would be wise to open the OS to other PPC platforms provided the software didn't need significant reworking. I don't see that it is essential to its success though, I think the direct opposite, such ambitious aims could be the death of the project at this stage.

Delaying AmigaOS 4.0 even further just to have drivers for more hardware available at its initial release would be stupid, yes. Get it out in the wild yesterday, and let people play with it and develop for it. I do however see a growing (over time) potential hardware install base as absolutely essential. This cannot happen with the current distribution policies, and things need to be done Right and not Absolutely Screwed Up Beyond Comprehension from the beginning. If the powers that be see no options at this moment, they should of course still not just give up from the get go and close the doors to any options. I don't think a strategy based on "we have so many resources and infinite budgets, and AmigaOS is a guaranteed success anyway, so let's do things the hard way now and do something sensible later" is an option.

I really hope this doesn't become quoted scripture for those arguing against the decision to limit the platform OS4 runs on. The "answers" are someone's points of view on the situation at best, certainly not undeniable arguments.

Of course. Thinking "I READ IT ON TEH INTARNET SO IT MUST BE TRUE!!11" is not wiser to do when reading my site, than it is when reading amiga.com or any other site. Does this really need to be pointed out?

BTW, again, it's about limiting the markets of the supported platforms (with the secondary effect of limiting the number of supported platforms).

[Seehund]

@olegil

Since OS4 at present will not work on any POP system, it's a bit silly to demand that it should work on ALL POP systems. IMNSHO.

Agreed. Have you seen such "demands" somewhere?

OS4 hasn't even been released AT ALL, so it's ENTIRELY too soon to demand ports.

It has however been announced what hardware it'll run on at its initial release and what dealers that users are allowed to buy this hardware from, and that this is a policy that'll regulate our hardware market for "all future versions of AmigaOS".

It is not too soon to "demand" that a commercial OS that's dependent on hardware from third parties will have the option available to be ported to more hardware, without extra requirements imposed on that market. Obvious and essential stuff like this should really have been taken care of when the AmigaOS product was in its planning stage. "So, how will we sell and get people to buy our product?" "Let's worry about that when it's released. In whatever way we... or somebody... will do that."