OSDev.org

octavio wrote: Instead of making a new OS that will dislike most of us,what about design a kind of standard like posix to make easily to
porting drivers and aplications betwen all OS and when one of us writes one driver ,others just have to port the code to their OS.

Already existing.

JAAman wrote: (not sure what MS charges for the windows source)

Even if you have enough money to buy Windows source, you can't get it that easy.

JAAman wrote: remember, open source does not mean free, it simply means the source code is availible to someone willing to pay what the owner charges (not sure what MS charges for the windows source)

The Open Source Initiative defines "Open Source" to include free redistribution of both binaries and source code. Sorry, no cigar.

Intel, HP, Sun, IBM, Adaptec, Compaq... Big players behind this project and such a bad looking webpage

The project is basically defunct, having suffered from the dot.com burst and being shot down by the FSF for "philosophical" reasons (that document still makes my blood boil, as it basically sets free software <=> Linux). Yet still, they proved it can be done, with reference implementations available at SourceForge.

Hi,

Solar wrote:Ahem... this is going to be lengthy, but I think I finally got the right words together to explain what I feel about GPL vs. PD.

I hereby appoint Solar as the official "Dude In Charge Of Licencing"! I think you've got the right words together too...

Solar wrote:Any kind of publication "protects your rights" in the patenting sense. By publishing it, it becomes "previous art", i.e. it is not possible for someone else to patent the underlying technology. All you have to prove is that you published the stuff before the patent was filed - you would have to prove that anyway, no matter the license.

Good!


Now might be a reasonable time to mention something I've been avoiding.

As you all probably realised, my proposal is based on my existing OS project. Some of you may vaguely remember that this OS project is an attempt at the impossible...

There were hints in my proposal ("binaries will initially be distributed for free, however there is nothing to prevent the OS from becoming a commercial product"), and I guess I've been dancing around the issue of the "closed source" modules long enough.

The main reason these modules are closed source is because it's difficult to sell something if people can get it for free.

What I propose is a full blown commercial product. I'm not talking about competing with Windows or Linux, but something more like SkyOS.

More accurately, I'm suggesting that the project be treated as a potential commercial product, so that (with a lot of luck), after years of hard work and years of giving it away free we have the option of charging a small fee for it.

Solar wrote:That fee might be enough to pay for the free download servers. It might be enough for some additional advertising in magazines, boosting market share. It might be enough to pay for a few pieces of patented technology that make computing so much more comfortable.

It might also be enough to implement a cash reward system (or a bounty system, like a few other projects), enough to purchase copies of standards where this can't be avoided, and maybe even enough to be taken seriously be hardware manufacturers.

It'd be very unlikely that this will work out, but it's even more unlikely if we don't try...

IMHO this isn't entirely impossible. The key to making it work is to provide features that aren't present in existing OSs - "seamless peer-to-peer distributed computing". This is what I've been working on for a decade. This is what can make the project stand out, and this is what people will hopefully be willing to pay a small fee for.

Now, some of you are probably wondering "how can this be a community project?". The answer is that I meant everything I've said previously. All of the open source components can be fully open source, and all of the closed source components can have open source equivelents. The only real difference is that the closed source (commercial) components will support secure distributed computing, while the open source (community) equivelents will be intended for educational purposes.

Anyway, have a good laugh (I expect that), but then take a few minutes to consider it...


Cheers,

Brendan

So you must form a company, and what about stocks and other stuffs? If you want to make money you must thing about that. So this isn't only technical problem, which gives more troubles

EDIT:
@Brendan:
And that part about mixing close and open source and commercial and free/educational purposes, I didn't understand it at all

Hi,

kataklinger wrote:So you must form a company, and what about stocks and other stuffs? If you want to make money you must thing about that. So this isn't only technical problem, which gives more troubles

This depends - in Australia it's possible to setup a business instead of a company, which has a lot less overhead and setup costs. It's also possible to setup a non-profit organisation (where income = expenses) which has tax benefits and less overheads. It may also be possible to use an existing company as a "trustee" to avoid setting up a new company. Of course there are probably other options and other countries are probably different...

I took a look at the SkyOS project, and I think they are a business rather than a company (it's hard to figure out exactly what the business structure is though).

In any case, it's something that would need to be researched properly, if the project ever actually reaches the stage where it's possible.

kataklinger wrote:EDIT:
@Brendan:
And that part about mixing close and open source and commercial and free/educational purposes, I didn't understand it at all

Ok, think of it as 2 seperate OSs where both OSs are made up of "modules". One of the OSs is entirely open source, using nothing but open source modules. The other OS is partially closed source, using a mixture of open source and closed source modules. Support for distributed computing would be entirely contained in the closed source modules.

This means there'd be hundreds of open source modules, and around 6 closed source modules, and most of the modules will work under both versions of the OS. That is the basic idea at least.

The public will always be able to get the open source (non-distributed) version of the OS for free, including full source code. To get the distributed capabilities they'd need the closed source modules, which (eventually) would involve paying a small fee.

Closed source modules would include:

• - the VFS module
  - the IPC kernel module
  - the scheduler kernel module
  - the boot manager

To support "trusted computing", there'd be at least one more closed source module for handling keys (the fritz chip), but I'm hoping consumer pressure makes trusted computing disappear.

For some hardware where non-disclosure agreements are required (e.g. video cards) there will be more closed source modules, but these modules would work under both versions of the OS and cost nothing to use.

It's like "shareware" or "crippleware", where you need to pay to use all of the features. The nice thing is that the crippled version would have the same features as a normal OS like Windows or Linux - I'm sure there's a good marketting slogan in there somewhere...


Cheers,

Brendan

Maybe a website with a "project of the month" or something could be done? What I mean is some website where quite mature OS projects could join and then one of them is selected each month or maybe 3 months or so, to be the "current" one.
The current project would then get help with stuff like drivers etc. and can keep it's creators original design goals. And when the time comes for a new project to be the "current" one as many members as possible try to boost that projects development.

This was just something that I came up with right now, but any opinions?

Brendan wrote: This depends - in Australia it's possible to setup a business instead of a company, which has a lot less overhead and setup costs. It's also possible to setup a non-profit organisation (where income = expenses) which has tax benefits and less overheads. It may also be possible to use an existing company as a "trustee" to avoid setting up a new company. Of course there are probably other options and other countries are probably different...

Well you can run a business but there can be only on owner (or two maybe). If more then two men develop that os (or whatever else) they should own a part of that business (it's like investment if they are not getting payed for the job they are doing) so it is not possibile to do that this way.

It may be a good idea about non-profit organisation, but yet again it depends on where the HQ of organisation is.

Excuse me while I mingle (again)... I think the Business classes are finally paying off.

Forming a stock-based company is a bad idea IMHO. The company would be controlled why whoever has the majority in shares, and the whole idea of a company-on-shares is that of "shareholder value", which can significantly collide with "best interest", believe me.

What you should have an eye on is liability. In Germany for example, if you don't form a company and do business as a private person, if you go bankrupt you are liable with all your possessions. Picture a single patent lawsuit and you can imagine how you can ruin your life that way.

So you'd be looking for some company / business front with "limited liability", i.e. when the company goes bankrupt, too bad, but the persons behind it remain unharmed (except for whatever they invested in the company).

Once you have decided on that, find out what kind of steering is OK with you. Again I can only quote German law, but some company forms (o.H.G.) require unanimous decisions from all owners... can be quite a pain in itself.

Then be aware that founding and registering a business or company of the desired form can cost quite something. The business form I'd have desired for my Pro-POS project would have cost 25k Euro to make official...

This is BSc 101. Any first-semester student should be able to give you the basics for your country. And be aware that you're approaching an international effort, which mightily complicates matters. (Again, the example is German law, which doesn't know such a thing as "releasing into the Public Domain"...)

All in all, a mighty PITA, and IMHO far too much headweight for a project not even decided worthwhile. Look at the vote results again...

Solar wrote: What you should have an eye on is liability. In Germany for example, if you don't form a company and do business as a private person, if you go bankrupt you are liable with all your possessions. Picture a single patent lawsuit and you can imagine how you can ruin your life that way.

So you'd be looking for some company / business front with "limited liability", i.e. when the company goes bankrupt, too bad, but the persons behind it remain unharmed (except for whatever they invested in the company).

Once you have decided on that, find out what kind of steering is OK with you. Again I can only quote German law, but some company forms (o.H.G.) require unanimous decisions from all owners... can be quite a pain in itself.

Then be aware that founding and registering a business or company of the desired form can cost quite something. The business form I'd have desired for my Pro-POS project would have cost 25k Euro to make official...

The law is Serbia is similar, put prices is lower. To start private business is almost free(less then 100 Euros), but you cannot to engineering & constalting I don't know is this true for SE as well. Costs for limited liability is about 500 Euros and for stock-based company 5k or 10k Euros.

@Solar: The more I listen to your way of things, the more I like the BSD license. ;D

I wonder how much it is here in Canada? I'm going to look it up

@kataklinger: Is that just the cost to regester the name and such, or is that just to get a permit of some kind?

Hi,

I agree with Solar - a stock based company wouldn't be appropriate. In Australia there's requirements like the company must do "everything in the best interest of the stockholder", which typically means ripping consumers off for as much money as you can for as long as you can, while minimizing the cash spent doing it.

I'd prefer that all cash was invested back into the project - for example, spent on marketting, resources, developers, etc. The company/business/whatever would be setup for an annual profit/loss of $0, with all income and expenditure publicly disclosed (e.g. on a web site). This helps consumers know what their money pays for, so that they don't mind paying the fee as much (and also helps to distinguish the project from large money hungry organisations like those stock-based companies).

This still leaves problems with liability.

I would be prepared to accept all liabilities (as long as the risks are minimized as much as is practical) until the project reaches the stage where it can reasonably afford to become a limited liability company. If the worst happens, I'd re-licence all of the closed source stuff as open source and upload it anywhere I can so that the project can continue while the lawyers ruin my life.

I'm thinking this won't be too likely, as the project won't be noticed much until it's reasonably large anyway, and there won't be any real risks until it's ready to be used (which may take years, or might never happen).

I guess I should also point out that I am already a registered business in Australia - I ran an electrical contracting business for 5 years, and while I have closed this business down I've kept the framework such that it can be restarted without any paperwork. Australian businesses don't have overheads and requirements like companies (company directors, AGM's, taxation nightmares, etc), so running it with an annual profit/loss of $0 isn't a problem.

Solar wrote:All in all, a mighty PITA, and IMHO far too much headweight for a project not even decided worthwhile. Look at the vote results again...

I don't think the vote results matter too much - my proposal varies significantly from what people were originally voting for. Despite this, it'd only take one or 2 people to start it off, and hopefully it would grow as it progressed from there.


Cheers,

Brendan

Brendan wrote: In Australia there's requirements like the company must do "everything in the best interest of the stockholder", which typically means ripping consumers off for as much money as you can for as long as you can, while minimizing the cash spent doing it.

Well it depends onf what are the interests of the stockholders, maybe it isn't money