That of automobiles. We have the car parts, the car design (engineering), the extensive testing, the car manufacture (the mass production), then regional distribution, retail sales.
After that there's all the services that go with it, with many specializing in some things and not others. Gas stations at the top, places like Jiffy Lube that do oil and fluids but don't touch brakes (and thus can't do VA inspections), body shops, tire speciality places, glass specialty places, etc etc. So in a sense, breaking down modern IT into its specializations can be done with any other industry (housing as well; the parts are standardized, as are the designs, paint, building practices; then you have local services for repair and maintainance, some of which you can do yourself just like cars).
there are three main differences between Software and most standard industries that have such large local support structures.
1) The price of duplication of software is 0. nothing. that in and of itself is an astounding concept never before seen.
2) The price of entry into the market is 0. You really don't have to pay anything to get into the market with a shareware or freeware piece of code. All it takes is time, which doesn't have to be "valued" every instance. You don't have to be certified, you don't have to be degreed, you don't have to be employeed and have everything go through some employer (and their lawyers and inspectors), you don't even have to have been through puberty yet.
3) Its possible, and commonplace, to NOT give the end-user the ability to fix problems. If your car has a problem, and you know something about cars, you can at least look at it to fix it before going to the shop. If your house has a need for a new extension, and you know something about hardware, you can file a construction permit, get the parts, pull out the tools, and do it all without having to consult a specialist.
When Microsoft products have a problem, you have to call them up. There is nothing you can do to try to fix the problem (except reboot). There is nothing you can do to fix the problem so future occurences don't happen.
Open Source is the alternative. When software breaks and you know how to write software, like with a car you can go under the hood and see if its something you can fix before telling the designer.
1. Hosting costs money. Bandwidth costs money. CD's cost money. CD Reproduction costs money. Servers costs money. These costs are negligable for a shareware product, but considerable for a full CD distribution that is very popular, such as Linux. Thus we have many mirrored servers. These cost more money. They exist at generosity of the community. Don't confuse giving or giving back with free, as in no cost whatsoever.
2. The price of entry for shareware is negligable, except for finding a host. The price for entering the retail market is considerable. Reproduction, packaging, advertising, and shelf space all cost money. The price for entering a speciality market, such as banking or health, can be considerable. Banking systems are so complex that they take several years of writing, followed by several years of testing, before they go fully live. Health software may require FDA approval. Air Traffic Control software requires an act of Congress.
3. For the average business man, is Linux any different? This is still a mysterious computer system that they hire specialists to take care of. For them, the task is non-trivial.
The more pertanent question here is: can the experts trained in a system take care of it? For closes source systems, the answer is kinda. For open source systems, the answer is kinda.
If we compare this to mechanics: some are general mechanics. Some mechanics specialize in braking system, and can turn rotors. Other mechanics learn how to make their own parts. There is an entire profession called Automotive Engineer. These are all different things.
How many people really understand the Linux kernel? How many build their own Linux distros? How many understand RPM packages? How many write their own software from scratch? How many people understand each of these processes completely? Most don't, even the experts. Once you exceed your area of knowledge, you are left either with a large learning curve, or the need to trust another expert.
Your argument does not hold for the common man, most buisness owners, or the expert who needs another expert. Open Source is as mysterious as Closed Source. However, you have a very good point about the experts. The experts, such as yourself, do need to know what is going on inside these things. It is this way that closed source represents a fundamental denial-of-knowledge-attack on the software designer profession. Thus, it is no surprise that software designers champion a system that they can understand and verify over a system that they must infer and guess.
For the record, I am for open source. However, I am limited in my support. Open Source should be championed upon strong engineering and system development foundations. This leads open source down a road of sustainability and reliability. I am against using open source as a way to thumb a nose at Microsoft. I think that good engineering and politics don't mix.
Actually, that still describes a ubiquitous industry
After that there's all the services that go with it, with many specializing in some things and not others. Gas stations at the top, places like Jiffy Lube that do oil and fluids but don't touch brakes (and thus can't do VA inspections), body shops, tire speciality places, glass specialty places, etc etc. So in a sense, breaking down modern IT into its specializations can be done with any other industry (housing as well; the parts are standardized, as are the designs, paint, building practices; then you have local services for repair and maintainance, some of which you can do yourself just like cars).
there are three main differences between Software and most standard industries that have such large local support structures.
1) The price of duplication of software is 0. nothing. that in and of itself is an astounding concept never before seen.
2) The price of entry into the market is 0. You really don't have to pay anything to get into the market with a shareware or freeware piece of code. All it takes is time, which doesn't have to be "valued" every instance. You don't have to be certified, you don't have to be degreed, you don't have to be employeed and have everything go through some employer (and their lawyers and inspectors), you don't even have to have been through puberty yet.
3) Its possible, and commonplace, to NOT give the end-user the ability to fix problems. If your car has a problem, and you know something about cars, you can at least look at it to fix it before going to the shop. If your house has a need for a new extension, and you know something about hardware, you can file a construction permit, get the parts, pull out the tools, and do it all without having to consult a specialist.
When Microsoft products have a problem, you have to call them up. There is nothing you can do to try to fix the problem (except reboot). There is nothing you can do to fix the problem so future occurences don't happen.
Open Source is the alternative. When software breaks and you know how to write software, like with a car you can go under the hood and see if its something you can fix before telling the designer.
Re: Actually, that still describes a ubiquitous industry
2. The price of entry for shareware is negligable, except for finding a host. The price for entering the retail market is considerable. Reproduction, packaging, advertising, and shelf space all cost money. The price for entering a speciality market, such as banking or health, can be considerable. Banking systems are so complex that they take several years of writing, followed by several years of testing, before they go fully live. Health software may require FDA approval. Air Traffic Control software requires an act of Congress.
3. For the average business man, is Linux any different? This is still a mysterious computer system that they hire specialists to take care of. For them, the task is non-trivial.
The more pertanent question here is: can the experts trained in a system take care of it? For closes source systems, the answer is kinda. For open source systems, the answer is kinda.
If we compare this to mechanics: some are general mechanics. Some mechanics specialize in braking system, and can turn rotors. Other mechanics learn how to make their own parts. There is an entire profession called Automotive Engineer. These are all different things.
How many people really understand the Linux kernel? How many build their own Linux distros? How many understand RPM packages? How many write their own software from scratch? How many people understand each of these processes completely? Most don't, even the experts. Once you exceed your area of knowledge, you are left either with a large learning curve, or the need to trust another expert.
Your argument does not hold for the common man, most buisness owners, or the expert who needs another expert. Open Source is as mysterious as Closed Source. However, you have a very good point about the experts. The experts, such as yourself, do need to know what is going on inside these things. It is this way that closed source represents a fundamental denial-of-knowledge-attack on the software designer profession. Thus, it is no surprise that software designers champion a system that they can understand and verify over a system that they must infer and guess.
For the record, I am for open source. However, I am limited in my support. Open Source should be championed upon strong engineering and system development foundations. This leads open source down a road of sustainability and reliability. I am against using open source as a way to thumb a nose at Microsoft. I think that good engineering and politics don't mix.