4.19.2011

Personal freedom, low-level programming and misguided old farts

When I think about what I really miss when studying informatics than it is a good relation between those rather modern-thinking developers using megabytes of hidden sourcecode within a single call and those who appreciate direct control using languages more close to hardware. From semester to another the latter ones begin to cease and you wonder where the classic and quality values of typical development went. In essence, they just speak more but don't get any actually new code done. They use existing technology to create irrelevantly trivial tasks and excuse their behaviour as science. To be honest, I don't like the term science and associated possible meanings, but I LOVE creating new approaches for already done "solutions" and rethink them in a more clean and atomic manner. Of course, this takes time, is more critical and error-prone - but in the end I feel better because I found some way to be close to the computer, it's initial purpose and knowledge about what's going on behind. It maybe that my lecturers do already know that, but it's questionable whether they actually believe it's of use to know that. I've seen their publication, their ways to create software and they simply work always on the same upper level of absraction and don't want to back off just one step. Either they have traumas from earlier assembler or even C experiences or they really never learned enough about it. It's like they ignore them by default. I can't look into their heads and I also don't want to. So it may be reasonable and all that, beeing a sign of dwelling too deep into their reasons.
However, I feel bound to the other side. Even if I'll never be able to program like I do know, I can do it on my own, bringing light to where once was darkness - to me and maybe to others. I started to rewrite things from the standard C library with a different philosophy, thinking more about compactness for the bytecode runtime I'll write for my own language. It's all about putting the right things in the right places. And function calls are always external.

No comments: