[OT] n-way union

[Andrei Alexandrescu]

Georg Wrede wrote:
> Andrei Alexandrescu wrote:
>> I am sorry, at the first two reads I did not understand your 
>> algorithm. I noticed that it is more complicated than the canonical one
> 
> Canonical one. Righhhhhtttttt......
> 
>  From the outset I thought that you did have a solution in mind. And I 
> thought that the pause of a couple of days before any seroious attemnts 
> only meant that everybody else understood it, too. So, nobody wanted to 
> spend 5 days inventing something which would only be dismissed "because 
> Knuth did it better 50 years ago".
[snip]

I don't even know how to start answering this and particularly what 
follows (30% of I can't even figure where it comes from and where it 
goes), but allow me to retort with just a few short and clear points.

1. You were not set up. I did have a solution in mind, but it's not from 
Knuth or some other classic. I didn't actually find it anywhere (which 
is why I consider the problem interesting to discuss in this group), but 
it's rather simple and clear because five people in this group came to 
it independently.

2. I'm not sure - are you reproaching me I didn't sit down to understand 
your algorithm (and pardon me, it wasn't the clearest exposition there 
is and it still has more than a few kinks to be ironed out)? And are you 
comparing the situation with your professor etc.? And am I supposed to 
thank you for not making a biggie out of it? Well your instructor was 
supposed to read your work, and I am not by any stretch of imagination. 
That was rather inappropriate.

3. If you are alleging that if somebody implemented your algorithm, it 
would run faster than BCS's, you are simply mistaken. Yours simply does 
more work than necessary, and no amount of buts and ifs can cover for that.


Andrei