Re: Would be really nice if...

Haximus wrote:
> > Sometimes 'innovative thinking' isn't so innovative, nor in the
best
> > interest of the language. This is one such case.
>
> I will give you a little analogy of your way of thinking: the boss
sends
> you out for a selection of donuts and dutifully you head down to
Dunkin'
> Donuts or where ever. The first thing you notice is that the boss'
list
> matches the offering at the store identically, only one item they
have isn't
> on the list - a plain donut. Now, most people will say "give me one
of
> everything except the plain donut!" But not Fitzy, due to his
stubborn
> nature he must verbally list each and all the donuts specified in the
list,
> one by one, to his server. Meanwhile people are waiting in line
behind
> Fitzy, wondering wtf he doesn't just specify which ones he DOESN'T
wan't in
> order to save time.

Look if someone has 100 columns to include in a select list, they never type them. You yourself said this. No one types them individually so it is not an issue. The usage of select * is debatable. I have come across one argument which I consider potentially valid for its use. In all other cases it is a bad thing and can cause many problems when used in production code. Extending the ability to use this potential bug provides no benefit other than you seem to like the idea.

Comparing programming languages to natural languages is the worst analogy I have come across in some time. Just because there are thousands of possible combinations of words that can be used to order the same bag of doughnuts does not mean this is a good design for a programming language. The amount of redundancy in SQL is already a problem. You just propose increasing it further.

-- 
MJB