This Linux is a new and exciting adventure for me.

When I write an application and distribute it, I believe that a binary for this new application is only for ONE type of Linux implementation.

It is possible to build various binary versions for various implementations.
That seems to complicate the release and distribution process.

Are Linux applications distributed with source so that users can build a version for their own particular implementation?

If a Linux applications is distributed closed source, how do various different implementations get the correct binary?
(I suspect it must be from the author.)

What are the usual considerations about this?

'implementation' is not used here as a distro, but the hardware platform.

You have asked some very poignant questions about the very essence of open source software and about nature of Linux distributions. I'll try to briefly address each of your questions.

First let me express a concern. I hope that this isn't a homework assignment. The questions sound a bit like homework. Asking people to do homework without clearly stating that it is homework is a violation of the LQ rules.

So, as you say at the bottom of your post, implementation is intended to mean the hardware platform. We refer to this as the architecture. Applications that are compiled into binaries are restricted to the hardware platform, or architecture, for which the binaries were compiled. This is also true of Windows and every other operating system. On the other hand if you create an application from interpreted scripts then these should run unchanged on different architectures. Scripts can be written in any of several languages including but not limited to bash, csh, tcsh, Python, and Perl. The code from these languages is 'compiled' at run time one line at a time. Each line is executed as soon as it is compiled, then the shell reads and compiles and executes the next line in the script.

This is the way of the computer world. There are emulators that attempt to create the illusion of a hardware platform so that you can run, for example, software written for Atari computers on a new Intel based computer. However software developers generally release a separate version of their product for each architecture that they want to see their product run on.

Google the phrase "open source". There is a mountain of information about the intended goals of the open source software movement.

In general the typical end user will not have the skills to read source code, much less to customize it.

The answer here is the same as how the end user gets the correct version of Linux for their architecture. When you visit a Linux mirror site you will see a separate installation kit for each of the supported architectures.

The only considerations are in cases that a given hardware platform will run several different binary versions of the product. For example, in Linux you will see installation kits intended to run on 32 bit Intel CPUs and other installation kits intended to run on 64 bit Intel CPUs. The 32 bit installation will work on a 64 bit machine so the end user has to decide which installation kit to use.

Thank you stress_junkie for this detailed reply.

No, this is not homework.

I have a very strong C background, PC hardware architecture, and other application experiences. Previously, I have used Linux, but not done detailed programming. As Linux is becoming used in embedded systems, my goal is to use this background and become deeply entrenched in Linux development (as much as possible). I hope to be active in kernel driver development. This plan is only possible because of the open source initiative and forums like this one.

I am close to buying a Linux Box and will put it on my network and do serious development. Now I am using Cygwin. I need the Linux box soon.

Thank you,
...Jim Hughen

You may not have to purchase any hardware to run Linux. You can install Linux in a free partition on your Windows computer. Many of us dual boot Windows and Linux. This can be an economical approach.

Just to clarify, that is true of shells eg bash, ksh, csh etc, but not of Perl.
In fact the real perl executable reads and compiles your prog/script up front and then runs it as a an executable env, which is why, like C, you can have the 'main' at the top instead of at the bottom.
See here for the details: http://www.perl.com/doc/FMTEYEWTK/comp-vs-interp.html

Not sure if the same applies to Python.