Re: [boost] 1.31.0 Win32 Graph/graph failures?

From: Jeremy Siek (jsiek_at_[hidden])
Date: 2004-02-02 11:21:34

Hi Beman,

On Feb 2, 2004, at 9:56 AM, Beman Dawes wrote:
> Hi Jeremy,
>
> >Ok, I know what caused this. There's a bug in that STL
> >implementation... they have the wrong
> >return type for std::set::find. I've changed it... though I don't have
> >VC++ handy, so you'll have to tell me if it worked.
>
> Tests running now...
>
> <rant>
> As an aside (and NOT for 1.31.0) it would be nice to hold the 
> execution time down on some of the graph tests. Both bfs and dfs take 
> a long time. For example, dfs takes around 45 seconds on my 3.0 Ghz 
> machine. Multiply that by 7 compilers, and you get the idea. 45 
> seconds of time on a 3.0 Ghz box is an vast number of computations. 
> Probably more test cycles than all the programmers in the world got in 
> a day in 1965. We ought to be able to write 100% coverage test cases 
> that none-the-less run quickly.
> </rant>
>

Ok, on HEAD I've changed the graph size and number of iterations used 
for those tests.

> OK, test finished, and it passes the three compilers in question.

Great.

Cheers,
Jeremy
_______________________________________________
Jeremy Siek <jsiek_at_[hidden]>
http://www.osl.iu.edu.edu/~jsiek
Ph.D. Student, Indiana University Bloomington
C++ Booster (http://www.boost.org)
Office phone: [phone]
_______________________________________________