This Week in D

[Chris Wright via Digitalmars-d]

On Tue, 01 Dec 2015 17:18:47 +0000, Ola Fosheim Grøstad wrote:

> On Tuesday, 1 December 2015 at 16:18:37 UTC, Chris Wright wrote:
>> Judging by who is labeled an SJW, I'm one, and mentioning the existence
>> of trans people in a context where their existence is relevant is
>> sufficient to be labeled an SJW.
> 
> Really? That sounds bad, hopefully this will pass. Sounds like the trans
> people are going through the same process as the gays did before them.
> People who have problems with it probably have some uncertainty about
> their own identity at some level.

Gay *people*. Not gays.

>> As for judging a programming language in three seconds...that's a bad
>> analysis of the situation. I left Nim in part because Araq was, shall
>> we say, less than friendly.
> 
> Oh well, but Araq is a mild breeze compared to the D citizens!!! There
> are plenty of people here that like to show off and prefer to go through
> the roof rather than having someone kindly bring them back to earth...

Araq is the leader of the project. His attitudes weigh much more strongly 
than those of other people in the community.

> But some of us are very reasonable!! Like I decided not to hit you back

Remind me to give you a cookie when you're in town.

> for wrongly claiming that my "O(N)" should have been "O(infinity)"

I believe you claimed O(N) searches would guarantee that the item reached 
the head of the list. His algorithm allowed you not to swap the position 
of the found element. He made this clear -- first he said that the range 
was inclusive, and later he even pointed out that, because it was 
inclusive, it would reduce the number of swaps near the head of the list.

Since it's possible for each search for an item not to result in a swap, 
it is not guaranteed that searching for the same item repeatedly will 
ever move that item to the head of the list.

> and
> that unqualified "big-oh" usually means "average complexity"

Context matters.

> (when lazy
> comp sci people use unqualified big-oh it always means worst case :-)
> But since I am going there anyway: average complexity analysis isn't
> something you can do on the back of a napkin

But doing an average case complexity for exactly one simple case where 
the only random variable is the random number generator is often pretty 
easy. And that's what Andrei did.