>132
http://c2.com/cgi/wiki?PrematureOptimization
>>136
I'm not sure what you're trying to say.
Very little selects against unproductivity in the computing ecosystem. I don't think "it's what people are doing" is necessarily a good argument because of this. In the professional/enterprise world, which dominates the culture of the open source world (in the sense that the open source world is mostly professional programmers in their spare time, not full-time open source writers), productivity is essentially selected
against. Your job security is improved by writing inflexible crap that takes a long time to develop. Secretly and subconsciously, few of us
care if we can be more productive.
There are occasionally miracles, like Rails. But it's no surprise that it was born out of web startups and otherwise people who had a vested interest in the success of their company and not its failure. (And, again, having an interest in your company's failure is the norm.) But claiming that Ruby isn't a product of fashion is pretty... stupid. It's like the poster child for fashion in computing. "rockstars" anyone? Come on.
And, qualitatively, a lot of what Rails offers would really be better in a language with macros. But the thing is, a Lisp just doesn't superficially look as nice/familiar as Ruby. It's like getting programmers to use high level language is a matter of giving them a "spoonful of sugar" to help the medicine go down.
The problem with Blub programmers' "bullshit detectors" is that the practical effect is "everything new (to me or otherwise) is bullshit." Which is mostly true. The problem is that they don't see how bullshit their day to day work really is. If their bullshit detectors were half as 'sensitive' as you say, they'd see that spending all day debugging incidental problems caused by C++ is
complete bullshit.