Combuster wrote:@bontanu:
I doubt you are aware of the exact parameters of his OS as it differs significantly from the norm here. As such it also differs much from the criterium on which the standard "Dont expect to make money out of your OS" is based.
Bearing much resemblance to an embedded OS, it has to find a place in a market area where there are tens of thousands of other OSes, of which the vast majority is making money. Succeeding there is much easier than succeeding in a saturated market which is dominated by three huge players.
I am aware of this "dream" parameters. He made no screenshots or binary release available and hence we are kept in ignorance here on purpose by him IMHO.
And NO in embedded market you can not succeed with an OS. OS is free or cheap there because there are a lot of open source embedded OS available with all sorts of features and styles. The only problem is interfacing with unknown proprietary hardware.
You stand more chance with a desktop OS than an embedded OS. Think about it... in fact 2 such event occurred in time: GeOS and BeOS did challenge Windows and were attacked and killed. Both were very good GUI kind of OS.
In desktop world the hardware is changing more slowly and it is more uniform than in embedded world hence a general GUI like desktop hobby OS has more chance in there (but of course still very low because of Linux/ Windows combo)
They key of making money in embedded world is having a lot of money to build a very cheap device in mass production and the social and commercial relations to advertise and place it in a good shelf in a lot of supermarkets.
You need a minimal budget of at least 1.000.000 euro for this. I know because I have worked in this area and produced such devices hence I have first class experience.
You can make money with an embedded OS if you have a contract to design a custom one for a very specific custom hardware / client and you are able to do it fast (3-6 months maximum).
A pre defined OS is of no use here because the hardware changes rapidly from one product to another in a search for the cheapest production costs.
Think about it: if you find 100.000 boards dumped in China with a 4 euro lower price than what you had yesterday then it makes sense to tray to adapt your OS+applications to it in a hurry because you gain an extra 400.000 euros in profit.
Hence C is king in embedded world contrary to popular belief that ASM is suited there. Ironically ASM is suited on x86 desktops.
Also it make sense to build your own chips if the product line is stable like in routers/ firewalls/ gateway/ p2p filters. And as an start-up you will have huge problems and need huge money to start making your own chips in order to compete well established ones.
More details about my criticism of his "dream" DINS OS you can find in his DINS "talk news" thread...
As for myself, I do OS development out of hobby, and in the hope I can help other people by trying things out and getting new information into the world. Consider it a form of academic research.
And of course for the kick when things like hardware accelerated graphics start to work.
That is a nice reason
However take care on that adrenaline rush or "kick" when you make things work for the first time.... Remember it will pass away... Hence establish yourself firmly into constant joy
Just kidding: I have to try that hardware accelerated GFX kind of "kick" soon
