Re: Thoughts on Kommander
Rob Kaper made some points against kommander in his recent blog post. I must admit that i quite like the kommander idea. It is the quick shell hack in a kde version. I can’t see what’s wrong with quick shell hacks. They are just that. That it’s quick doesn’t mean that it has to have horrid interface. And recall the concept: kiss - keep it simple, stupid. I see no point in creating big apps with kommander. That doesn’t disqualify it for the small ones tho. Claiming otherwise is like claiming that you should write a full-blown C++ app to recode that bunch of html’s into utf-8. Iow, overkill.
I don’t believe that kommander guys want it to be (ab)used for large and complex projects, other than as a customization tool. And about being a killer feature, yes, i can see it as such. It doesn’t appeal to me either, i’d just hack the thing in python or perl or ruby or anything with full qt/kde bindings. But there is still a big market for kommander.
As for the RAD… There is a way to substantially speed up development, and it leads through use of higher level languages. C++ just doesn’t cut it, honestly. You must pay too much attention to too many details. And you don’t really gain much control that would be of some use to you. Programmer time is really more valuable than machine time these days. And waiting for g++ and ld to get the thing together is a pain in the neck. You can test much faster with an interpreted language. And the runtime speed hit is not that bad either. Not for small-to-mid sized programs. And it can cut the development time significantly.