Site operation discussions
Language design isn't just about big ideas, it's also about little details...
Usability testing the little details might show there is One True WayTM or it's all a waste of breath.
There is an interesting article in Time magazine on the IPod Nano, in particular the quote It was a gutsy play, and it came from the gut: unlike almost any other high-tech company, Apple refuses to run its decisions by focus groups..
I believe that great design in the end always comes from the gut feeling of the designer. Listen to your customers, but ignore what they say.
Find out what people actually do rather than what they say they do; measure how quickly & how accurately & ... rather than how quickly people think they do something & how accurately people think they do something... ; make the trade-offs based on data rather than gut feel and persuasion.
(We've got beyond guessing where the hotspots are in code, now we profile.)
Seems silly to me, who is to say which is the cart and which is the horse? I can't think of an advantage either way. I use a lot of SQL, often with many joins that make my FROM clause as long as my SELECT. If you have a short FROM, then finding out what an alias refers to is easy. If you have many tables, then you have to search within the clause anyway so it doesn't matter if it is before or after the SELECT. Of course I don't even know if aliases exist in this case, I only know about SQL. But also, besides Yoda, who says, From the cupboard get me that tomato soup?
Thanks for the links Erik. Ever since I worked my first job at a grocery store I have been saying that customers don't know what they want. It's nice to know that the folks with diplomas on their walls agree with me. I have also always meant that to mean that there is no substitute for old fashioned ingenuity. If you want a great idea, you can't just ask someone (or a focus group) to give it to you.
"None of us is as stupid as all of us" ;>
You're all familiar with that concept, right?
The IQ of a meeting is computed based on the IQs of the meeting's participants in much the same way as the net resistance of a parallel resistor network. In other words, assuming a meeting of n participants, each with an IQ of IQk where 1 ; then the IQ of the meeting IQm is computed as follows:
IQm = ___________________
--- n 1
--- i=1 IQi
And you said IDEs don't influence language design...