Whoa since when is Red Hat not free?

[hasu]

It is an interesting discussion. If we are continuing the discussion in this thread..

Originally posted by: n0cmonkey
Apples and Oranges. The tools for building software are plentiful and cheap (gcc and friends are free). The tools for building a truck from scratch are not. Think virtual, not physical.

May be I should start building the tools first. To me the difference between s/w and h/w are very subtle.

Originally posted by: drag
One thing you have to understand is that software is not like physical objects. It's just ideas, abstracted math. It only functions by being copied. You copy it from the cdrom to the harddrive. The software is copied from the harddrive to memory. Memory is copied to cache. Cache is copied to registers to be used to drive transistors and modify input data.

In fact as I said there is very little difference between s/w and hardware. You need time, raw materials and knowledge to build either of it. In case of software it may be compiler, existing pieces of software and in the case of a truck it may be hardware tools and other raw materials. To make those tools to build a truck you have to start from the iron ore, build a plant from scratch to extract iron and start build the tools - this can be a community project and finally you can have all the tools you want if you have enough knowledge and free man power to support you. Same is the case with gcc compiler. You need computer, electricity, paper and pencil, free man power and knowledge. So what is the difference? Difference to me is that in s/w it is easier to derive from a previous implementation and you can do it without getting your hands dirty and sitting comfortable in front of the computer.

Originally posted by: drag
However software itself is so cheap to replicate that it is essentially free. Effectively the cost of making one Linux kernel is the same cost of making 10,000,000,000 copies of that linux kernel.

So is the case with a truck or any other hardware. Price of the product reduces and gets closer to the raw material cost as the production increases. Why are we getting 100 DVD-R's for $15 now? If I build a car plant just to assemble one truck how much will that truck cost? Similar arguments goes with software too. If Microsoft had the team to make a single copy of Vista how much will that one copy of OS cost?

That may be the reason why the price of CPUs gets lower and lower as the time pass by. May be these Linux distributions should follow that. When they release a new version, sell it for some small amount of money -- may be for first 6 months then taper the price off as the time goes by and it would eventually become free may be in an year or so. That will give these companies required revenue and a reason to improve on the previous version.

 

[Nothinman]

May be I should start building the tools first. To me the difference between s/w and h/w are very subtle.

In some ways yes, but in duplication there's no comparison since software can be copied an infinite number of times instantly with virtually no cost.

So what is the difference? Difference to me is that in s/w it is easier to derive from a previous implementation and you can do it without getting your hands dirty and sitting comfortable in front of the computer.

And again the lack of material costs for duplication. If someone wants to download a copy of your software and change it all it costs is some disk space for them and some bandwidth for you and both are extremely cheap. But if someone likes your car and wants to duplicate that and mess with it they have to buy the physical parts or take yours which means you no longer have it.

So is the case with a truck or any other hardware.

That's BS, sure mass production lowers the cost significantly but it's nowhere near free.

 

[hasu]

Originally posted by: Nothinman
That's BS, sure mass production lowers the cost significantly but it's nowhere near free.

Even for building the car the raw material is free from the earth. If I get enough free man power and knowledge I can build a free car from free resources available from the Earth. And once the plant is ready anybody can improve upon using more free labor. The only difference is that in case of car if a little difficult compared to software.

 

[Nothinman]

Even for building the car the raw material is free from the earth.

Only if you can find a place to get those materials that isn't already owned by someone. And there is still a transfer of material that doesn't happen with software, you're taking the material from the earth and putting it into the car but with a copy of software there's nothing being physically moved.

 

[sciencewhiz]

Originally posted by: her209
The only reason I'll upgrade is for hardware compatibility, otherwise, its RH9 for me.

That sounds like a rooted server waiting to happen, if it hasn't already.

 

[hasu]

Originally posted by: Nothinman

Even for building the car the raw material is free from the earth.

Only if you can find a place to get those materials that isn't already owned by someone. And there is still a transfer of material that doesn't happen with software, you're taking the material from the earth and putting it into the car but with a copy of software there's nothing being physically moved.

Yes, it is just that creating something for free is a little easier in s/w. But I don't see anything for free other than software. Some how the concept of "free" is just tied to software industry. Heck I don't even find someone to do just a free oil change in my car.

The case of freedom has been exactly opposite, Software never allowed much freedom (and is getting worse every year). VMWare had restrictions in performing performance comparison of their products. Now Microsoft has similar restrictions for .NET. If you buy a car you are free to use it whatever way you want, disassemble it and learn how it was made, do performance comparison or write a review. That's what I would expect from non-free Open Source Software and that's the ownership. That doesn't mean that I can buy an Accord and establish a new factory just to build cheaper Accord clones. That should be prevented by appropriate laws.

To me, open source software and free software must be two different things and in the long run, even open source can not be free. If it continues to be free then there won't be much progress in it. But it might take a while before reaching that point.

 

[drag]

Originally posted by: Nothinman

Even for building the car the raw material is free from the earth.

Only if you can find a place to get those materials that isn't already owned by someone. And there is still a transfer of material that doesn't happen with software, you're taking the material from the earth and putting it into the car but with a copy of software there's nothing being physically moved.

It's not even that.

Cars are not made with raw materials extracted from the ground.

You need to have iron removed from the earth. Then it needs to be refined to remove impurities. Then you need to have a place to combine it with exact measurements of carbon, tungsten, magnanese, and other trace minerals to turn it into usefull steel. A difference of a thousandth of a percent of the carbon content of the steel with produce different grades. Add more carbon to make it stiffer, to make it harder. Add to much and it becomes brittle. Some applications require that you need to have steel to be soft and take shocks, or your need to have special formulas to make springs.

Then you have lots of other types of materials that need to be created. Aluminium, leather, plastics, and all sorts of other stuff.

This requires a massive effort involving many thousands of people working in many factories just to get the materials you need before you can even begin to build a car. You needs ways to orginize all of that.

Then you need molds, cutters, weilders, and lots of other machinery present build the spare parts, and you need to have people there that are knowledgable enough to use that machinery and build the spare parts. Then you need to have people that know how to take those spare parts and create a car from them.

There is no way you can ever create a car on yourself by taking raw materials from the ground, not unless your willing to devote years of study and then years of labor to that single goal. And that is just to create a COPY of a car. Be sure to throw in another decade if you want to learn how to engineer one.

I think that it's very likely that engineering a car is actually a small cost compared to the overal effort that goes into creating a paticular model of car. I bet a car company is going to spend as much or more money on their legal department then they actually spend on their engineering.. since the overal orginization required to make automobiles on a large scale is exponentially larger then the nessicary workforce it takes to design one.


With software the act of design is, along with documentation, is the only significant cost. That is all it takes to create useful software.

This is why software is unique. The costs of making the software is effectively the same as it takes to utilize the software (ie a computer) and. depending on the distribution method, the costs of distribution and duplication are largely handled by the end users themselves (ie internet connection).

Then unlike most things it can effectly be handled in a largely distributed manner. With Art you can't do that, if you have people working in a group they have to work as a group. With software the indivduals can be working with fairly minimal amount of coordination as long as they have decent communications (generally IRC and mailing lists) and good guidelines.

Then unlike most things you can use software as building blocks for new software. Code reuse is a very important aspect of software development.



For example if Ford was to give their cars away for Free it would quickly put them out of business. The costs of materials and labor would kill them within weeks.

But with Redhat it costs them the same to sell their software to people that need it to replace aging Unix systems versus selling the software to those same folks PLUS giving away the software to every man, woman, and child on earth. It only costs them in terms of bandwidth.. and that is why things like bittorrent were created.



This entirely unique to anything else in human history. This is something new, something different. This is something that is very rare for societies to have to encounter and deal with something that is radically different as software.

It's not something that can realy be compared to automobiles, or cooking, or writing a book, or painting a picture.

 

[Nothinman]

Yes, it is just that creating something for free is a little easier in s/w. But I don't see anything for free other than software. Some how the concept of "free" is just tied to software industry. Heck I don't even find someone to do just a free oil change in my car.

Because it's virtually free to make copies of software and distribute them. Most people create stuff for themselves like rebuilding a car, knitting blankets, painting, etc for fun. Building software is just another kind of hobby but it has the added benefit that you can give out virtually unlimited copies of the software for free whenever you want.

The case of freedom has been exactly opposite, Software never allowed much freedom (and is getting worse every year). VMWare had restrictions in performing performance comparison of their products. Now Microsoft has similar restrictions for .NET.

That's nothing new so I wouldn't exactly call it a regression. Companies like Oracle, Sun, MS, etc have a long history of publishing benchmark results of some pieces of software.

If you buy a car you are free to use it whatever way you want, disassemble it and learn how it was made, do performance comparison or write a review.

But you wouldn't be able to use that knowledge to build a competing car or you'd be sued by whoever owned those patents. There are a lot more variables that can affect software performance than there are with a car. When someone reviews a car's performance it's fairly simple to take their test and get similar results but with software things like library versions, kernel versions, other things running, etc can change the numbers.

To me, open source software and free software must be two different things and in the long run, even open source can not be free. If it continues to be free then there won't be much progress in it. But it might take a while before reaching that point.

No one ever said that open source and free software were the same thing and lots of places have gone to great lengths to spell out what they require in order for something to be considered free. Everyone usually points to the Debian Free Software Guidelines, OpenBSD and the gnu.org list of licenses compatible with the GPL.

 

[hasu]

Originally posted by: drag
Cars are not made with raw materials extracted from the ground.

Neither was gcc or any other free software. You take the physical computer for granted. Who will spend money for that and why?

With software the act of design is, along with documentation, is the only significant cost. That is all it takes to create useful software.

All the processes and planning which is required in designing a car goes into designing an IDE also -- especially now a days both are done on computer. And it takes 100's or 1000's of volunteers to build the new version of a free software. If you can get 100's or 1000's of people to work for free you can build whatever you want. It is a little easier in software because it is easier to collaborate such a work over the Internet and you don't have to get every one at the same time at one place etc. But if that trend was not confined only to software, we should have seen that kind of startups even in other areas at least on smaller scale. My question is why don't we see that kind of free stuff in any other discipline.

For example if Ford was to give their cars away for Free it would quickly put them out of business. The costs of materials and labor would kill them within weeks.

Even if you are ready to pay the cost of materials still it would kill them, because labor is not free in automobile industry.

It's not something that can realy be compared to automobiles, or cooking, or writing a book, or painting a picture

Let's hope that the new trend won't kill its own development

 

[Nothinman]

All the processes and planning which is required in designing a car goes into designing an IDE also -- especially now a days both are done on computer. And it takes 100's or 1000's of volunteers to build the new version of a free software.

I think you overestimate the manpower required to build most software packages. Sure there are some packages that require a lot of work like Gnome, KDE, the kernel, etc but they're the exception and are usually broken up into smaller chunks to make it more digestable. Using the Linux kernel as an example, the whole thing may be ~7 million lines of code but the majority of the work by far these days is in the drivers and each one of those is only a few hundred to a few thousand lines. Infact looking at the commits between Linux kernel 2.6.19 and 2.6.20 there was under ~760 separate authors according to git and that's not taking into account the same person submitting via different email addresses so it's even smaller than that.

And one person can do an amazing amount of development work if they really want. Just look at how quickly Linus got git into a working state after BitKeeper decided to revoke the kernel developer's free licenses. It's about scratching itches, in that case Linus knew none of the available VCSes would be able to handle the development model of the kernel so he went about writing his own.

Let's hope that the new trend won't kill its own development

What new trend? There has been open and closed source software since computers were created and they've both survived so far.

 

[drag]

The 'Free' as in Freedom part is what is required for "open source" software to work in the long run.

Otherwise companies aren't going to contribute back to it because it wouldn't be fair for themselves to do that.

Right now Microsoft is starting to surrender to the 'Open Source' camp. They've done a few things, open sourced a few things. Some things that matched the OSI definition of 'OpenSource' and other things which are 'shared source' initiatives were effectively your allowed to looki at the source code, but you can't actually use it for anything important.

One of things they did was 'shared source' their kernel for embedded developers.

But it's worthless to them. You can look at it and study it to find out how the kernel works. but you can't fork it or use it in your products. If Microsoft stops selling licenses for the binaries you can't go and then keep using the old source code instead. So despite these half-hearted features they continue to loose market share in that market to Linux like ever other properietary OS vendor despite Linux's warts. And it's not even the cost that is significant, the trend for developers are that they are more willing to pay for support themselves.


Its' the Freedom that makes it work and it is what has allowed development to progress to this point.

For example in the EU's FOSS study attempting to quantify the economic impact of Free and Open source software concluded that in a year 1.2 billion Euro is spent specificly on the development of Open Source software. That this money is largely spent by companies who's primary business is not selling software, but they require the software to operate their businesses (of which is worth 256 billion Euro). The cost of reproducing this by using closed source software would be 12 billion Euros.

Also that less then 10% of software developers in the US are actually employed to create closed source software. The vast majority of the programmer job market is for corporations that use software and need to make their own software for internal purposes.

So the logical conclusion is that the vast majority of software is developed out of needs and business requirements and Free and Open source software has the advantage of being vastly cheaper and more usefull for the people that end up needing to make their own derived software.


So in order for FOSS software to stall is because software by that time will become perfect and there would be no need for people to create new or refine old software.

In that case proprietary software developers are screwed also.

 

[Sunner]

I don't even understand how someone can compare cars to software.
Assuming you have a computer, which is a pretty fair assumption since you wouldn't have much use for software otherwise, there is really only one cost involved in creating software.
Time.

It takes time to write it, but once that's done, you have your finished product, and the cost of duplication, distribution, etc, is extremely small.
And the thing is, everyone has time, and everyone has exactly the same amount of it, it's just a matter of how they choose to spend it.

A car on the other hand...
It still takes time to design it.
It takes materials, physical materials that can't be duplicated, and that need to be extracted and processed.
It takes tools, big and expensive ones.
Every new car needs new materials, and needs the tools to put it together.

Arguably, the design stages of both are more or less the same, they just take time.
After that, there are no similarities at all, the cost of producing the software(as in, copying) and distributing it is virtually nil.
The cost of producing and distributing a car most certainly are not.

 

[hasu]

Originally posted by: drag
The 'Free' as in Freedom part is what is required for "open source" software to work in the long run.
Otherwise companies aren't going to contribute back to it because it wouldn't be fair for themselves to do that.
....

Its' the Freedom that makes it work and it is what has allowed development to progress to this point.

For example in the EU's FOSS study attempting to quantify the economic impact of Free and Open source software concluded that in a year 1.2 billion Euro is spent specificly on the development of Open Source software. That this money is largely spent by companies who's primary business is not selling software, but they require the software to operate their businesses (of which is worth 256 billion Euro). The cost of reproducing this by using closed source software would be 12 billion Euros.

Also that less then 10% of software developers in the US are actually employed to create closed source software. The vast majority of the programmer job market is for corporations that use software and need to make their own software for internal purposes.

So the logical conclusion is that the vast majority of software is developed out of needs and business requirements and Free and Open source software has the advantage of being vastly cheaper and more usefull for the people that end up needing to make their own derived software.

I agree to all you said in this post.

My point was since there is no monetary benefit for free software developers progress will be slower. Look at gambas as an example. How many years did it take to reach this point. The development started in 1999 (as wiki says). I tried it a couple if years back and really liked it. I could very well appreciate the product as an RAD tool because of my experience with Borland's Delphi.

 

[drag]

Originally posted by: hasu

Originally posted by: drag
Cars are not made with raw materials extracted from the ground.

Neither was gcc or any other free software. You take the physical computer for granted. Who will spend money for that and why?

They did and they continue to do.

For example both IBM and Redhat continue to improve on GCC.

One example is SSE Trees and autovectorizing. A great deal of this technology comes from IBM.
http://gcc.gnu.org/projects/tree-ssa/vectorization.html

Why did they do this?

Because they want to make it easier for programmers to automaticly take advantage of multiple proccesors cores without having to rewrite their software. IBM wants to hardware, the future is lots of fast small proccessor cores. GCC is realy the only game in town when it comes to cross-platform compilers.


Another example is SSP for GCC. Originally developed by IBM researchers for GCC and was called Propolice. Redhat took up the torch for it and rewrote it so it was suitable for inclusion into GCC proper.


Why did they do that?


Because Redhat wants to sell services for supporting software. Stack smash protection greately reduces the chances for traditional programmer mistakes to cause security issues. Through the simple act of recompiling your software then improves the security of it by eliminating the exploitability of a whole class of buffer overflow flaws. This increases the marketability of Linux since improved security was always a bullet point.


I could go on and on.


The vast majority of software development is done for a purpose. The vast majority of software development is done by the people that need it, it's NOT done by people who want to market it to other folks. (although sometimes that is a secondary effec)

It's all done to fuffill a need. You need to be able to _justify_ the costs of producing software.

Having the software Open Source or Free does not diminish the need for developing and improving the software. In fact it is usefull because it dramaticly reduces the costs of development plus most of the time it actually results in superior code due to a number of secondary effects.

The reality of the situation is that most software development is done by or by the order of the people that use it.

With software the act of design is, along with documentation, is the only significant cost. That is all it takes to create useful software.

All the processes and planning which is required in designing a car goes into designing an IDE also -- especially now a days both are done on computer. And it takes 100's or 1000's of volunteers to build the new version of a free software. If you can get 100's or 1000's of people to work for free you can build whatever you want. It is a little easier in software because it is easier to collaborate such a work over the Internet and you don't have to get every one at the same time at one place etc. But if that trend was not confined only to software, we should have seen that kind of startups even in other areas at least on smaller scale. My question is why don't we see that kind of free stuff in any other discipline. [/quote]


If you don't see it it's because your not looking or you are blind.

Science? Mathmatics? Medicine? Pretty much any research feild is similar.

Even for things like machinists and engineering. People share ideas and concepts all the time. This happens in almost every other feild. Doctors, lawyers, engineers, machinists, artists, mechanics all get together and talk about stuff. They publish trade journals and help each other out when they can.


And software is nothing _but_ concepts. It's just abstracted math.

All of those disciplines have always benifited very strongly by openness. Sharing ideas, sharing concepts. It's when you start to close them off is when they start to suffer.

Sure patents are there to make it economicly viable.

But you have to realise that the intention of patents is TO HAVE THINGS OPEN. NOT to close them off. In order to get a patent, say for a car engine part, you have to publicly publish your inventions and everything about it. In other words patents are a attempt to legally require you to be OPEN. To make your ideas and concept public domain.

Patents were designed to eliminate the need for hiding things, for keeping things secret... in other words 'trade secrets'. That and in europe any invention was automaticly a state assit and essentially inventors and scientists were state servents, so it was a huge marketing feature for USA at the time.

That's right. FREEDOM as a incentive.


Due to a veriaty of reasons, however, the system breaks down for software. Patents are a very artificial things that were never designed to work with software. I won't go into the whys and hows right here, (if you want to, I can) but with software patents they are used to keep things closed. People use them as a excuse to hide source code, to hide ideas and to enforce incompatabilities.. which is entirely opposite of why we have patents in the first place.

 

[Nothinman]

My point was since there is no monetary benefit for free software developers progress will be slower. Look at gambas as an example. How many years did it take to reach this point. The development started in 1999 (as wiki says). I tried it a couple if years back and really liked it. I could very well appreciate the product as an RAD tool because of my experience with Borland's Delphi.

And for counterexamples look at Compiz/Beryl, Xen, the Linux kernel, etc. They've all progressed as fast or faster than their commercial counterparts. Monetary benefits are only one motivator and when you're doing something for your own use it's not even usually on the list.

 

[hasu]

Originally posted by: drag
They did and they continue to do.

For example both IBM and Redhat continue to improve on GCC.

One example is SSE Trees and autovectorizing. A great deal of this technology comes from IBM.
http://gcc.gnu.org/projects/tree-ssa/vectorization.html

Another example is SSP for GCC. Originally developed by IBM researchers for GCC and was called Propolice. Redhat took up the torch for it and rewrote it so it was suitable for inclusion into GCC proper.

Basically the support for open source comes from business community because there is a viable economics behind their business. They are not doing all this to create some free products.

you don't see it it's because your not looking or you are blind.

Sharing ideas or publishing papers is entirely different from releasing the final product or service for free. There is no economics supporting pure science research and that is one thing which makes it difficult to explain the importance of basic science to others. So we cannot compare sharing of ideas to free open source products.

Even for things like machinists and engineering. People share ideas and concepts all the time. This happens in almost every other feild. Doctors, lawyers, engineers, machinists, artists, mechanics all get together and talk about stuff. They publish trade journals and help each other out when they can.

That happens in proprietary closed source software also.

Sure patents are there to make it economicly viable.

But you have to realise that the intention of patents is TO HAVE THINGS OPEN. NOT to close them off. In order to get a patent, say for a car engine part, you have to publicly publish your inventions and everything about it. In other words patents are a attempt to legally require you to be OPEN. To make your ideas and concept public domain.

Patents were designed to eliminate the need for hiding things, for keeping things secret... in other words 'trade secrets'. That and in europe any invention was automaticly a state assit and essentially inventors and scientists were state servents, so it was a huge marketing feature for USA at the time.

I agree

Probably we should list all the original ideas and inventions (not necessarily patented) developed in proprietary world and free software world to see who had contributed more. Proprietary projects are always properly funded and some of the open source projects are also funded (again by companies whose intentions are different).

 

[drag]

Originally posted by: hasu

Originally posted by: drag
They did and they continue to do.

For example both IBM and Redhat continue to improve on GCC.

One example is SSE Trees and autovectorizing. A great deal of this technology comes from IBM.
http://gcc.gnu.org/projects/tree-ssa/vectorization.html

Another example is SSP for GCC. Originally developed by IBM researchers for GCC and was called Propolice. Redhat took up the torch for it and rewrote it so it was suitable for inclusion into GCC proper.

Basically the support for open source comes from business community because there is a viable economics behind their business. They are not doing all this to create some free products.

So, what is the difference?

The system works because it's Free. They want software because they need it to get stuff done. Why other purpose does software have?

Also you notice that the places were open source, or at least Linux OS is popular or gaining popularity is were the software development is a nessicity. Enterprise business, scientific computing, embedded computing, and such things. Were people tend to rely on software as 'products' it is not as popular, and about the only place that happens is on the desktop.

you don't see it it's because your not looking or you are blind.

Sharing ideas or publishing papers is entirely different from releasing the final product or service for free. There is no economics supporting pure science research and that is one thing which makes it difficult to explain the importance of basic science to others. So we cannot compare sharing of ideas to free open source products.

The software isn't the end product. Software is a tool to accomplish other things. The end products are usually things like financial accounting systems, or automating assembly lines, or other such things. What people use software for. It's what you can do with the software to make money.
Or the end products is support and services, like what Redhat does.

It's closed minded thinking that says the only way to make money is by restricting access to something.

Like I said for 90% plus programming that happens in the US at least is purpose-driven development, not profit driven. This is the same for proprietory software as open source software. It's just that OSS tends to be better and cheaper.

I don't see proprietory software dying any faster then open source software is.

Even for things like machinists and engineering. People share ideas and concepts all the time. This happens in almost every other feild. Doctors, lawyers, engineers, machinists, artists, mechanics all get together and talk about stuff. They publish trade journals and help each other out when they can.

That happens in proprietary closed source software also.

Sure patents are there to make it economicly viable.

But you have to realise that the intention of patents is TO HAVE THINGS OPEN. NOT to close them off. In order to get a patent, say for a car engine part, you have to publicly publish your inventions and everything about it. In other words patents are a attempt to legally require you to be OPEN. To make your ideas and concept public domain.

Patents were designed to eliminate the need for hiding things, for keeping things secret... in other words 'trade secrets'. That and in europe any invention was automaticly a state assit and essentially inventors and scientists were state servents, so it was a huge marketing feature for USA at the time.

I agree

Probably we should list all the original ideas and inventions (not necessarily patented) developed in proprietary world and free software world to see who had contributed more. Proprietary projects are always properly funded and some of the open source projects are also funded (again by companies whose intentions are different).

Ok. I'll start:
TCP/IP.
Unix.
Internet.
World Wide Web.

All created by in situations were they were able to freely redistribute the source code and AT&T had no financial purpose behind original creation of Unix since they were forbidden by their monopoly status to sell computers or computer software. Of course the greedy corporations killed the golden goose, but that is why Linux is now slowly killing off propriatory Unix systems. Linux is the new Unix. (along with the BSDs, of course)

Your turn.

 

[hasu]

Originally posted by: drag
So, what is the difference?

The system works because it's Free. They want software because they need it to get stuff done. Why other purpose does software have?

Also you notice that the places were open source, or at least Linux OS is popular or gaining popularity is were the software development is a nessicity. Enterprise business, scientific computing, embedded computing, and such things. Were people tend to rely on software as 'products' it is not as popular, and about the only place that happens is on the desktop.

Basically free software is a huge IT infrastructure maintained by many companies (like RH and Novell) being paid by their clients such as IBM and HP. That's perfectly fine, which does not need programmers write code for free. They get paid for it anyways.

I now realize the reason for difference. All through the discussion, I have been talking about Linux becoming a Desktop OS. All I said was if Linux has to compete with other proprietary desktop operating systems (such as Windows and OS X) then Linux cannot be free, some one has to take the tab. End users for Linux would then be common people like me.

 

[drag]

I don't know the figures on it, but most all Linux developers get paid for what they do.

It works out pretty well for some folks. Other people do it as a hobby. whatever floats their boats.

For instance I can pretty much garrentee you that if you understand networking and become a expert at Samba and are a contributing programmer that you WILL get people to hire you. People will line up to give you money.

(once you start dealing with the ugly side of microsoft compatability.. that is making windows do stuff that Microsoft doesn't want; like file and print server compatability with non-microsoft servers then you'll understand why there are so few Samba expert programmers to go around.)

What do people generally use personal computers for?
Do work stuff at home.
play video games (and high-end 3d games are actually the minority of things. Flash and puzzle games out perform those, ironicly)

Then occasional multimedia stuff. Like make a dvd copy of a home recording or a slide show type thing.

Gnome, Novel and friends are doing their best to make a good business case for Linux desktop there.

For video games there are lots of small things, but it's not good at the big things. Which isn't so bad since big pc gaming is a dying industry anyways. Open source would probably help a lot.. Like WOW would still make a crapload of money if they based their stuff on existing open source software. Most people would be suprised by the amount of open source stuff that goes into your average video game. But the licenses are very typically non-copyleft licenses so people that work on code that is used in commercial games will never receive any benifit from other people using it. There are lots of interesting things that could happen, but they aren't going to for a long while yet.

You have some video-game like things being worked on, but they are probably being held back by the proprietory nature of desktops and 3d hardware/APIs nowadays.

People point out second life as a big deal, but I dont' think it's a big deal at all. Them open sourcing their client stuff and they are talking about open sourcing the servers stuff, but they may amount to something after all, but you have things like http://www.opencroquet.org/


Their faq has funny parts:

Why do we need Croquet?

Answer: There are a number of reasons why we need something like Croquet today. First, the personal computer user interface was cast in amber 20 years ago. This was for various reasons:

* The desktop GUI originally developed by Alan Kay and his colleagues was an extremely successful approach to interacting with the personal computer.
* The emergence of software monopolies removed any encouragement to innovate on the platform. Compare a modern PC of 2004 to the first Macintosh shipping in 1984 and the major difference you will see is color.
* The fact that these dominant systems were created in early bound languages made it impossible to easily modify the foundations of the system either by the developers or by third parties.

Croquet was built to answer a simple question. "If we were to create a new operating system and user interface knowing what we know today, how far could we go?" Further, what kinds of decisions would we make that we might have been unable to even consider 20 or 30 years ago, when the current operating systems were first created? We decided that it was time for an existence proof that innovation could still continue and succeed on the personal computer. We felt that the very definition of the personal computer and its role needed to be shifted from a single-user closed system to a next generation broadband communication device.

It's funny because it's true. Stuff like beryl or whatever is interesting, but 3D desktops can go far far far beyond just finding creative ways to display windows.

Coming from most people this stuff would be very pompus, the first time I saw it I thought it was a joke. But they are the same people that are responsable for a lot of the important developments in the internet and web stuff. (and no, not web2.0 )


As far as multimedia goes what is aviable in Linux and is open source software is usually much more capable then the typical sorts of apps people tend to use for stuff on Windows, not taking into account professionals. But they need refinement and simplification for more people to get used to it.

 

[Nothinman]

For instance I can pretty much garrentee you that if you understand networking and become a expert at Samba and are a contributing programmer that you WILL get people to hire you. People will line up to give you money.

That's a pretty ambitious statement, the only Samba developer that I know of for sure that gets paid for working on it is Jeremy Allison and that's only because of the big stink he made over the Novell/MS deal.

 

[drag]

Originally posted by: Nothinman

For instance I can pretty much garrentee you that if you understand networking and become a expert at Samba and are a contributing programmer that you WILL get people to hire you. People will line up to give you money.

That's a pretty ambitious statement, the only Samba developer that I know of for sure that gets paid for working on it is Jeremy Allison and that's only because of the big stink he made over the Novell/MS deal.

The thing is is that Everybody and their Mom uses Samba.

If you have a OS that is not Windows and you need it to communicate with Windows for whatever reason (say your building a NAS) then you'll need Samba to do it.

So I wouldn't expect you'd get hired to contribute to samba, but you'll probably be able to find a job quite easily because you can and do hack on samba.

As far as big companies that depend on Samba for whatever reason, like Apple, they don't contribute jack back themselves.

 

[Nothinman]

The thing is is that Everybody and their Mom uses Samba.

Yes but that doesn't mean that everyone working on Samba can get paid to do so.

If you have a OS that is not Windows and you need it to communicate with Windows for whatever reason (say your building a NAS) then you'll need Samba to do it.

Maybe. Or you could require your own client software be installed on Windows like Novell used to do. Or you could use some other SMB software like Dave.

 

[hasu]

Excerpts from the article
> Linux Weekly News has just published one of the most interesting
> analysis pieces on Linux that I've seen in ages. In it, LWN executive
> editor Jonathan Corbet addresses the common misconception that
> Linux, and other major open-source projects, are maintained by volunteers.

Read on .... Who really writes Linux?

 

[drag]

That's just one of about a billion different open source projects, it's talking about the Linux kernel which is very special and unique in a large number of ways.


In that E.U. study I linked to previously claimed that they found that 2/3s of software is still written by individuals, were as the rest is written by 'firms' and 'other institutions'.

 

[Nothinman]

Excerpts from the article
> Linux Weekly News has just published one of the most interesting
> analysis pieces on Linux that I've seen in ages. In it, LWN executive
> editor Jonathan Corbet addresses the common misconception that
> Linux, and other major open-source projects, are maintained by volunteers.

Read on .... Who really writes Linux?

The article also says that 32.7% of the changes could have come from unpaid volunteers. 1/3 is a pretty big percentage, especially when you note that the largest contributer only accounted for 12.8% on their own. And I'm sure he uses the phrase "may have" for a reason, it's very difficult to tell with 100% certainty that a specific patch was written because a company paid the person to write it. It's very possible that someone gets paid to work on part Y of the kernel and still occasionally works on part Z in their free time. And it's also possible that someone posts patches via a generic Google, Yahoo, etc address even though they're working on a specific company's dollar.

Also as drag points out Linux != OSS. It just happens to be one of the most prominent projects.