nick (no@mail.com) on 2/17/08 wrote:
---------------------------
>Howard Chu (hyc@symas.com) on 2/17/08 wrote:
>---------------------------
>>David Kanter (dkanter@realworldtech.com) on 2/15/08 wrote:
>>---------------------------
>>>Howard Chu (hyc@symas.com) on 2/15/08 wrote:
>>>---------------------------
>
>>>1. You're not an average programmer - average programmers are not as good at optimization and are probably lazier
>>
>>Those folks should find another career then, one where their aptitude will allow
>>them to be excellent instead of just average.
>
>Take that to its logical conclusion, and you eventually will become average and
>have to stop programming. Sometime after that, there will be nobody left. :)

Heh heh. Not quite, but perhaps the original statement was too vague. You can pick some arbitrary set of metrics that can be objectively measured, such as # bugs per # lines of code, or percentage of real flops vs theoretically available flops. You can state "bug percentage of .01% or less is excellent" or "flops efficiency of 80% or more is excellent." If you eliminate all the programmers whose code is not excellent, then the remaining "average" programmer is still excellent. Being an average member of a population is certainly a relative measure that changes with the makeup of the population, but you can establish absolute measures of their products that are fixed regardless of the population. The two are completely separate concepts.

>But seriously, it is quite obvious that there is supply and demand at work, and
>you simply can't just brush that under the carpet without telling us how you plan
>to switch the world to controlled economies.

Funny you should mention that. I was thinking a good way to start would be to outlaw marketing.

There certainly is supply and demand at work, but it is being heavily manipulated by marketing which mostly consists of outright lies. In a free market with accurate information, the demand for junk products would dry up. There are already people objecting to the proliferation of bloatware - it's human nature to want good stuff and avoid bad stuff. In a networked world with ubiquitous, instantaneous access to information, it's *possible* to have an educated market that consistently avoids the junk. We're not there yet because we don't have review sites that are free from the financial influence of their product suppliers, and we don't have consumers that are universally informed enough to know to look before they buy. Education is a partial solution here too.

>So given that we actually have a lot of crappy programs and not-so-good programmers
>around, what this actually means is that software is so important to people that
>they are willing to pay for a poor product. Pricing them out of the market by reducing
>the number of programmers will not help them. What will help them is increasing
>the quality and productivity of the supply of programmers.

>Education is definitely one way to go. It is by no means the most cost effective
>or the only way to go. So increasing the quality and/or quantity of the poor and
>average programmers output (by means other than education) definitely has value.

Increasing the quality, ok. Making life easier for lazy programmers, I'd rather not. Those guys can go be lazy in somebody else's industry, or find work that they enjoy so much that they're always motivated to go do it.

Education may not seem cost-effective in the short term, but short term perspectives aren't ever going to solve anything.