The problem with opensource

The largest roadblock preventing OSS from becoming ubiquitous is not in the philosophy (not at all, fantastic philosophy) but rather in the medium. Modern, decent programming languages are simply too complex for Joe Average, reducing the contribution base for OSS down to a minority of proficient coders and crippling one of opensource's main selling points. For OSS to achieve its ideal, we need to develop simpler, more intuitive, more accessible, yet equally powerful forms of coding, such that any reasonable person using a piece of software is not limited in their ability to patch or modify it in any significant way. This is, obviously, not possible with the current state of computer science. We must overthrow tradition and embrace the future. WE MUST RISE UP AND FIGHT FOR OUR IDEAL. VIVA LA REVOLUCION 4CHAN.

>>21

Yes, however I would argue that the reason they suck at "usability" (read: GUI) is because it's not an interesting task to a hacker.

Furthermore it's not something the programmer himself would actually use - to the hacker usability = simplicity which more often than not means writing CLI tools. Note how the most popular tools for writing Unix GUIs are special purpose scripting languages such as Tk and widget libraries such as Qt, GTK which enable you to throw up a GUI in the least possible time.

Here's a thought I just had: what we need is a package that enables non-programmers to take a program and develop a GUI without having to look at any code. Something like Visual Studio only language independent. It would be necessary for the GUI developer to be able to poke the program from different directions to determine its behaviour. Perhaps the language independence could be achieved by an Emacs style extension language to enable it to work in different modes.

Hell maybe it already exists and I'm just ignorant. If not though maybe it's something I should write :)