Hi.

Thinking as a business: how the hell can I create a 'system' with this OS, with these restrictions, with this little users and developers in the market and with next to no software available?

Mind you, I don't mind paying for it, as long as I'm not locked in to an OS I cannot adapt myself. If something breaks, who's going to fix it for me and by when? The risks are just way too high when I can use another OS, buy support for it, and can have access to the source to verify its workings and more importantly, have people all over the world who can write applications for it.
Sure, they are entitles to license is however they want, but "appears to be" does not cut it for business use, especially in the embedded world (which would be the only commercially viable possibility for this OS at this point). Do they want to grow besides a hobby-os and actually make money out of their time investment?
Sure I have :-)

First, I do work for free. I've been involved in many projects in the past which did not pay. I'm currently involved in an unpaid project which takes up a lot of time developing, and even some money I put into it. (it's for a humanitarian project in Algeria)

An no one said open source == free. I hope they will make money from it, but I think they are going at it the wrong way: how can you make money from a few thousand users....
Linux runs just fine on an Efika if you use the chip for what it is intended. I think instead that you're the hobbyist running MorphOS - nothing wrong with that of course. It's when you start running applications such as FireFox and OpenOffice on Efika that you run into problems - but this is not the OS's fault (you would have the same issues when OpenOffice ran on MorphOS).
In your opinion, not in mine.


Respect has nothing to do with the licensing: one is at a personal level (and they did a good job - never said otherwise), the other is at a business level (how are we going to viably use it/ make money with it).


Johan.

_________________
Johan Dams, Genesi USA Inc.
Director, Software Engineering

Yep, I have a blog... PurpleAlienPlanet

Except that VxWorks has a proven track record, one can buy support from them, and the source is available if you pay for it. (actually, the source is free, except for the kernel source - which costs and requires an NDA). Also, there are ports available to many platforms.
Of course, VxWorks is in an entirely different class than MorphOS, aiming at mission critical applications. In short, you cannot compare MorphOS with VxWorks unless we're running nuclear power plants with MorphOS.

Do know however, that VxWorks is getting a lot of competition from Linux in markets VxWorks used to be king, e.g., in the telecom sector.

There are parts of some nuclear power plants that actually do run (real time) Linux, as are a lot of medical appliances, especially those aimed at home nursing (with and without sensor networks).


Johan.

_________________
Johan Dams, Genesi USA Inc.
Director, Software Engineering

Yep, I have a blog... PurpleAlienPlanet

Well, Johan replied for me and covered most of the points I wanted to make, but:
This is not a dumbass attitude. Don't know how old you are or how much you've seen from the IT industry, but I've seen pretty much all important Computers/OSes after the 70s. Amiga, BBC, Atari, Spectrum, Amstrad, Commodore 64, Linux, BSDs, MacOS, BeOS, etc (well, I haven't used RiscOS :-P).
Apart from the commercial Unices, Windows and Macos (who all had lots of $CASH$ to support the platforms), ALL of the rest died or are irrelevant . Of ALL these OSes (not all are OSes I know, but let me make my point) that died because open source did not exist back then. IF Amiga or BeOS used an open source license in the '90s, things would be very different right now. But they haven't and things are as they are.

You might like to hide from the fact or think that everyone else is crazy to think like that, but the fact is indeed that "everything that is not open will eventually die".

Matt, I think you're wrong there, many companies that used VxWorks before, are switching to RTLinux, simply because development is cheaper on that. Sure, VxWorks and the rest of the platforms have their market, but eventually they will have to become more open. I don't mean GPL, GPL is a stupid one-way license, there are plenty more alternatives.
I wish they make lots of money on their OS, I just have doubts on their viability.
I won't discuss the quality of their work as I haven't seen it, but don't make the mistake to assume that it's a revolution in the OS industry. It's a light and efficient OS, but it's not the first nor the last. The fact that noone without an Efika/Pegasos can try it doesn't help its case either. Also, whether I've done anything as worthy is irrelevant. ALL of my work on Altivec has been TOTALLY open, and I dare everyone here to deny that. Anyone who does AltiVec work, could potentially benefit from my work. OTOH, MorphOS has been apparently using AltiVec internally for some operations, but apart from Grzegorz who works on MorphOS no one has even hinted of any "openness" in his/her work. Who's doing more good here? The MorphOS team? or me? No, wait you don't have to answer that, I already know your answer.
I'd like to hear a QNX user on an EFIKA, and their impressions. I think you might be disappointed to know that MorphOS might come second in terms of efficiency on the EFIKA.
There is no reason for me to be insulted by your comments, simply because it's obvious that you have absolutely no idea of how the industry operates. Also, the first thing in this thread was questions about installation, wrong again there. Lastly, it was never a question about respect to the MorphOS team, I respect what they do, but no, I won't be grateful to them for charging more than pretty much every OS out there -save the commercial Unices.

Let's just say I have enough industry experience that I couldn't care less to get into this argument. Picky semantic-play bored me decades ago. As Matt said, pretty much the only thing behind that would be a free version. I don't care.

Back to things that matter...

Having slight issues with the installer. After it completes and I boot hd:0 /boot.img I get the MorphOS loading screen, then a fairly blank grey "Workbench screen" with mouse pointer and date/time, no menus or anything. Anyone..?

EDIT: the guys on the mailing list solved this one for me (disk format/partition table mistake on my part). Running now, woohoo!

I'm glad to see the release we've all LONG awaited :)
I'll be downloading the demo tonight, dunno when I'll get a chance to play with it on the Efika though.

I am a little disheartened by the price. 240USD is a little out of my budget right now, especially for a desktop OS solution. And for those with multiple machines... ouch.

_________________
I <3 tiny computers.

Johan, your assessment is not wrong per se. However, it should be noted that MorphOS is not specifically targetted at the embedded space for many of the same reasons that you cited. The most realistic embedded use would be for multimedia devices which happen to include dedicated 3d graphics hardware, for instance. A project such as this would profit from certain fully implemented core features which are only available in a very limited number of embedded operating systems.

In general, while there are no technical stumbling blocks why MorphOS could not run inside, say, a sophisticated dish washer, there is

a) no compelling reason why a company should use MorphOS instead of certain proven and known-to-be-stable operating systems in this market segment (security of investment),

b) no compelling reason to wait for MorphOS to be adapted so it can fulfill the same role which other existing solutions already do (matter of timing),

c)noone who actively promotes it as an embedded operating system which implies that businesses will not even get to know about it (customer information or the lack thereof).

MorphOS is just not marketed for what you think it is :) As far as Amiga-inspired operating systems go, AROS is the obvious choice for many embedded use cases (simple kiosk systems, etc.) since it can be used free of charge and its source is freely available. Provided, that the EFIKA port will eventually see the light of the day.

Hi André.
Wouldn't it be in MorphOS's best interest then to try and reach those markets by taking the development beyond the hobby realm and involve more users/developers? The only realistic way to do that from my point of view is to open up. How else is this little known OS with ZERO market penetration or even visibility outside a select few going to compete against Linux (which can fulfill the multimedia aspects you mentioned) which has a lot more developers available, is tried and tested and has a huge momentum going which is growing every day?

Don't get me wrong - I truly believe MorphOS can be a major kick-ass OS for these appliances. I'm just wondering how MorphOS is going to show that to the rest of the world when it only runs on a select subset of hardware and is fully in control by a couple of people who see their OS as a hobby...

One more thing, the last thing I want to do is turn this thread in a flamewar. These questions that I ask are genuine. Two years or so ago I had a project where MorphOS could have been used - let's see where we can go in the future.

I do believe the only viable way to get MorphOS beyond its current state is to open up. And to counter some possible flames, as Konstantinos mentioned as well, GPL is not the only way to open up, and is not at all the best way for MorphOS.


Johan.

_________________
Johan Dams, Genesi USA Inc.
Director, Software Engineering

Yep, I have a blog... PurpleAlienPlanet

Johan:
I actually doubt that the MorphOS Team is even interested in
making a big operating system. I'm pretty sure they all have
their regular day-jobs and want to keep MorphOS as a hobby* -
nothing that they would live on..

* A hobby is something you keep for fun. When you make a
profession out of it the stakes are going up and you still
need food on the table..

That was always how I understood it to be. It's a project done for the same kind of fun as homebrew or reverse engineering. A technical challenge to create your own OS, perfect and suited to a targeted few platforms (after all, they hardly have the time or resources to bring mass support - neither Linux or OSX covers those bases either).

From what I've seen in the few hours I've used it, I have really enjoyed the experience. Yes... enjoyed! Never once as with Debian even running the lower demanding window managers has the Efika *ever* felt like the comfortable-to-run low power desktop machine that it appears it can be. That's a remarkable thing.

They have updated functionality and kept the core `feel` of the Amiga OS alongside a GUI which is far more consistent, professional, pleasant to look at and friendly than any KDE or Gnome variant I've ever seen. It's remarkably intuitive too - in itself a wonderful thing for a small team to have managed.

I play with operating systems daily whether old builds of obscure projects, up to the current builds of Windows Seven. Whether little touches like the implementing of the old-style BootPic, or the handling of USB drives being connected, it does it how I would've wanted to.

You seem to consider thinking of this as "hobbyist" some kind of insulting term. I see it as a team of enthusiasts who created something to show other enthusiasts just what could be done. I think they've done an incredible job so far. Aside from audio currently still being due for the Efika, it does almost everything I would want day to day for a light use low power computer. Once I find an IRC client and a word processor of any decent quality, it'll perform most of the features my desktop machine now chews through power for.
Wow.

Just to confirm, even Genesi no longer promotes MorphOS to it's partners. While there was some excitement about it the last time we tried, it was rejected eventually for exactly a) and b) reasons.

"Why can't we just use Linux? Linux is here now!"

Unfortunately there *is no market* for an Amiga-clone toy OS with such a slow development cycle. I don't think it even exists for AROS. However if someone comes up and asks us for MorphOS.. sure.

150 EUR might be a good price from a certain point of view. I like shiny toys. I could buy an MP3 player, for example.. it just depends whether I want to re-live my teens by listening to Nirvana or pretending the Amiga is still alive..

_________________
Matt Sealey

If MorphOS hadn't been promoted at all in regard to the Efika, myself and several others I know of would never have even considered using the platform. It would've been just another underspec base motherboard, with the added issue of being PPC based which narrows the options considerably.

I can only speak for myself and a member of my family but say that MorphOS was the single biggest draw the Efika had for us over say, a £60 x86-based board - and I doubt I'm the only one. Much cheaper, faster, with far more features, OS options and similar power use. There was a very real point where it came down to a decTOP - which satisfied a lot more of the criteria - or the then-reduced price Efika.

The single point the Efika had for us *was* MorphOS. That's a fact. It WAS the deciding factor. We wanted what we saw. Which, were it not for quite a bit of linking being made between it and the Efika as a low cost platform that ran it... well, we'd never even have heard of the Efika were it not for MorphOS. I've spoken to and showed people enough of it since it came out that *they* were interested in the platform as a way to run it specifically.

Not because we're `pretending the Amiga is still alive`, but because it is what it is. A very fast, useable, user-friendly, well featured desktop-ready OS. One with extremely quick boot times, that uses the Efika's features to their strength, and doesn't take hours to configure to do so. My Efika has seen more actual use (in terms of me actually having it out and running vs gathering dust as a glorified IRC terminal) since MorphOS was installed than probably the entire time I've had it combined. You tell them about the platform, they dismiss it. It's a low powered PPC board that struggled to run Linux. You show them the OS and they are impressed.

The defining thing I have realised since starting to use MorphOS - I haven't once thought about the hardware itself since. Not once. I was too busy actually doing useful things in a nice environment that I never once thought of it at all. Not as that thing that made me constantly excuse its hardware for not being able to *quite* cope with an IRC client and a web browser and accepting that as a given. But as a working, useful computer - funnily enough much like the Amiga itself - where the hardware became largely incidental compared to what was running on it.
That was my understanding of what MorphOS would be. They delivered it with a good few unexpected nice touches and polish on top of that. It makes the Efika less a device, more of a desktop computer. Apparently far beyond what Genesi intended it to be. It's a funny way to sell the hardware to be so down and so negative on such a great showcase for the potential. Maybe there is no market for this OS outside a small group - I personally took it for granted that was the intention of what they were doing anyway, a fun project for people who wanted an enjoyable little OS. I'll happily be part of that nonexistant market, because they've done more with it to promote the Efika as an actual computer than Genesi. It's a niche device in itself competing against thousands of others that are often far more capable of the tasks they are in demand for. I'd have thought every sale it could get via MorphOS would've been useful and appreciated.

Ok, let me try again...

Some early quotes: (emphasis mine)

This basically implies that you want to see wide spread adoption, so the authors can make money for the past/future development. We argued that this won't happen unless this and that would happen - one of the main requirements that it would become more open (as in source, not GPL), and that there is a possibility to take MorphOS there, focusing on what it's good at (fast,multimedia,...)

Some later quotes: So, doesn't it make sense then to open source the system so that more people can join the hobby? I know for sure I'd like to check out the kernel of this little OS (maybe I can even use it as a teaching tool in my Operating systems classes).

Oh, I have nothing against it being a hobbyist system, I just thought it might have the potential to grow bigger, like, say, Linux did. This does not have to mean that the MorphOS developers have to give up their hobby, or their daytime jobs, or whatever.

Besides, the current developers might give up on their hobby in favor of something new and exciting (thats what happens with hobbies). If you have the source, at least others can pick up easy and make it their hobby project.
It's not an "underspec base motherboard", it's an evaluation board for the 5200B embedded chip. I don't understand why anyone would want to use an EFIKA to replace their Desktop PC. The 5200B is wonderful for those embedded applications where you need relatively lots of computing power, where a normal microcontroller wont' do, yet are limited in power supply. It can be used as a thin client (even running some applications itself), but it just isn't made to run a desktop oriented OS with the latest FireFox and OpenOffice.
Well, good for you. If it satisfies your needs and those of a few others, thats fine. After all, thats what a hobby is (I have many). The problem is, a lot of those hobbies don't make sense when you try to make money - which is what a company is trying to accomplish.
I have to disagree with you here. It is not desktop ready. There are not nearly enough applications for MorphOS to make the platform desktop-ready.
Linux is literally light years ahead of MorphOS. Of course MorphOS will boot faster since it omits a lot of the things Linux is doing which MorphOS is not. There are graphical, multi-threaded Operating Systems for microcontrollers that will boot in a few seconds, but these cannot be compared to Linux either...
Oh come on, Efika can do this just fine. I even created a GUI for http://projects.powerdeveloper.org/project/efika/338 based on a web interface using a Webkit based browser. You can also see some GUI demo's in the blogs.
Maybe there is. And all I was trying to argue is that it would be possible to explore these options if it were open source.
It is an embedded system - Genesi has been promoting it as such. MorphOS could have been part of it if it had been available 2 or so years ago. since it is a non-existent market at this time, what could possibly be lost by open sourcing the system?
No it's not. It is actually positioned very nicely and has enough computing power for what it's targeted at. Besides the performance per Watt is a factor that will become more and more important (it is already), and the 5200B and successors are great in this regard.
The problem is that a couple of hundred even 5000 sales don't mean anything for a company - they are targeting millions!
With 5000 sales you cannot possibly recover the development and manufacturing costs of the system, let alone run a rewarding business. As I said before, MorphOS could have been part of the equation some years (maybe even 1 year) ago to drive a market of possibly millions of devices.

Let me summarize this rant...

MorphOS == hobby-OS (Great); maybe open sourcing it will allow more people to enjoy the hobby.

MorphOS == Not to make money (Great); why charge large amounts anyway? Why the difficult licensing (one per system, with keyserver...)

MorphOS == Real Business (Great); Open it up so we can see what it can do, and maybe we can use it to sell millions of devices. Maybe not.

MorphOS is better then Linux - no
MorphOs has features which set it apart from Linux - yes - how can we capitalize on these?


Johan.

_________________
Johan Dams, Genesi USA Inc.
Director, Software Engineering

Yep, I have a blog... PurpleAlienPlanet

To STKD, Johan covered me completely, I only have ONE question:

Why is a hobbyist OS incompatible with the Open Source philosophy? Why can't they open the source and let others share the fun (I know I would like to).

Also,
Wrt to boot times, how fast is MorphOS booting on EFIKA? 10 secs? 20sec? It is possible to lower the boot times of Linux significantly (I have it down to 9secs to a full KDM manager though on an Athlon X2, most time is taken by udev btw, also down to 25s from 50s on a G4 1Ghz). Steps are being taken to make this even faster, we've started a fast (written in C) udev replacement. If one gets rid of old UNIX-crap and doesn't use shell scripts for booting, Linux is in par with MorphOS and all those fast embedded RT OSes. Just give us a little time to further our designs and prepare an alpha version at least.

Hi Johan,
first of all, opening up is a very broad term. Essentential parts of MorphOS are already open source and managed via community development. Furthermore, the act of becoming a member of the main development team and receive access to certain parts of the closed source is not an impossible endeavour. So, MorphOS is already "open" to a degree.

If "opening up" strictly refers to making its source code freely available to everyone, please note that this measure will in fact guarantuee that a) it is fully set into the hobby realm and b) development is made harder because it would be unlikely that there was any cashflow in the short term, if ever.

Consider how long it took Linux to develop from a pure "hobby" project with no commercial support into a free operating system whose commercial development is driven by services-oriented business models.

Case in point, I would like to refer you to the open source AROS which, as surprisingly as this may seem, has not been known to be an innovation drive in the loosely related Neo-Amiga scene. In fact, AROS has existed far longer than MorphOS but offers noticably less functionality in its current state today. (To be fair, MorphOS does use some AROS code.)

Apparently, the overall situation is too complex to argue that a mere license chance will have drastic effects.

I will refrain from commenting on the current state of Linux ;-)

Here we are talking about the true core of the problem. Hardware support, that is. I can assure you that the developers are well aware of this. As has been mentioned previously, MorphOS is being ported to the MacMini and possibly additional Mac models.

This is obviously only a stop-gap solution, but there is no doubt that it represents a major leap for MorphOS compared to the current state of affairs.

As for the hobby label, this was used previously by some to describe that they do not make a living off MorphOS and work on it in addition to other full-time work. However, it is safe to say that MorphOS was not initially started to create a non-profit operating system. Of course, profit is nowhere to be seen on the rader, but it is not like the core developers would reject the very idea if this opportunity were to exist in the future.

Put differently, if a delusional optimist works on a project which is intented to make him a millionaire for three years until he gives up without earning a dime, he is a commercial entrepreneur. If somebody with substantial doubts about the profit probability does the very same, that person is considered a hobbyist.

"Hobby" can be a very inprecise label. Linux is the best example.

I did not perceive your comments as a flame attempt at all :)


André

Hi Johan,
you have a surprisingly narrow view of the word "company" :-)

The majority of workers in the Western hemisphere works for small- and mid-sized businesses of whom most have no realistic chance of selling millions of anything. Ever. Yet, they miraculously survive, some even flourish.

Now, I do understand where you are coming from. But we are debating about a niche project which is not presently aimed at world domination.

For the sake of argument, let us do the math...

5.000 x 150 USD = 750.000 USD of revenue

At a modest average salary of 5.000 USD per month, you could pay 12 full-time workers and still have 30.000 USD for test hardware, administrative expenses, etc.

5.000 sales per year are unrealistic? Well, YellowTab sold ~100,000 copies of ZETA at 100 USD in Germany.

Granted, the scenario above may not make up for past work, but it is close to impossible to calculate how much development time was spent by each individual developer anyway. As for your "rewarding" comment, this is a highly subjective matter. Some people find it incredibly rewarding to be able to work on a project that they are passionate about.

Steve Jobs, whose Mac computers are still a tiny minority in the global scheme of things, is quoted as saying: "Being the richest man in the cemetery doesn't matter to me ... Going to bed at night saying we've done something wonderful... that's what matters to me."


See AROS.

Even if you ignore all the time spent on development, you actually need to invest money to be able to develop an operating system such as MorphOS 2.0. The price of MorphOS 2.0 is first and foremost a contribution to future hardware purchases and will help financing various support services (file hosting, mailing lists, etc.).

Source code validation would only be available to serious inquiries from companies who have the resources and commitment to pay a substantial sum for a commercial source code license. Nobody expects this to happen.

I agree 100%. For instance, Linux is a robust solution for servers of all sizes, whereas MorphOS is simply no serious solution for a server whatsoever. Obviously, the choice of an operating system depends on each specific use case.


André