I like writing programs and releasing them as free software. I do like the thought of helping others with my software, but also I write them for selfish reasons . I want to find out something, play a bit with technology, do some experiment. Most of these experiments never leave my harddisk, but sometimes they could be useful to others. Even while 'only' experiments, I do try as much as I can to make the programs easy to install, easy to use etc., in short, they should Just Work(tm).

Too often I see programs announced with remarks like 'I know it's buggy and unstable, but here it is!'. Well - who would be enthousiastic to use such a program! So I try to do a bit better; it's almost unavoidable that there will be problems anyway, because people's environments differ from mine. But I'll do my best. (And still then, I found out the hard way that there are significant difference between Python 2.2/2.3/2.4, and PyGTK... Arghh.)

I said I like writing programs. Well... writing programs for me is a bit like running - it's a masochistic joy. It's especially nice afterwards.

Actually, there is another reason I write free software - I like to see other people using my stuff and writing about it (to feed my arrogance of course). Now, it's really easy to find a thousand links to your software in google. If you're in debian and freshmeat, mirrors and feeds will quickly make your program pop up in a million places. But with a little better searching I can find reviews for my program in Russian or German, or my Bonobo-Tutorial in Korean. That is just cool.

So I always try to give my programs a name that's not too generic and easily googable - don't call it, say, reader or eyeball or... LaTeX [actually, searching for latex gives you pretty good results]. Just don't call your uptime-improver v1a9ra.

And my final word of wisdom for today: there's no need to check for the return value of g_new and friends - they will never return NULL.