i see some sources where they mention it as being "unoptimized".
is there some technical meaning behind this, or do they mean that they didn;t spend too much time on it?

Optimization refers to seeking out areas of a program that behave more slowly than desired, or scale poorly, and improving them. Optimization should be one of the LAST things that you do in a program's development cycle, since you won't know where the performance bottlenecks will be until that point. Optimizing before then may end up in no overall performance gain.

great - thanks agin.

Optimizing is not worthwhile if you save 60 seconds of runtime off something that runs one hour. The user will figure the runtime = lunch time
and run it accordingly.

Or if you save few seconds on a short program that runs once a month in batch.

Saving one half-second on a function that is called thousands of times in a program that runs all the time, IS definitely worth doing.

The trade off is cost of programmer time vs the time savings staff will experience.

The only environment where optimization is everything is game development

yes, this IS for game development that I asked. spot on!

so, how is it different for game developers? what do they do differently?

They probably optimize over and over once their codebase has stabilized. Once you optimize out one bottleneck, you profile again and look for other bottlenecks. Then you optimize thouse out. And so on and so forth until it would become prohibitively difficult to code and/or your performance is acceptable.

a program may be optimized to take full advantage of a processor's ability. for example a program may use i486 assembly calls. such a program is i486 optimized. it may run on on a i386 only if it is backwards compatible and has a set of i386 calls to replace the i486-specific call (run unoptimized of course). when you run ./configure, AutoConf will determine which arguments to pass to gcc (or other compiler). these arguments are specific to your system. so on my P2, the following arguments to gcc are present "-mcpu=pentium2 -march=pentium2". im not sure if that makes the program backward's compatible with pre-i686 archeitechtures or not, but if it does not, then somewhere along the line, this binary would crash on a P1.