[Uml-devel] [ot] google

[Andrew Sutton]

On Wednesday 04 February 2004 09:16 am, Jonathan Riddell wrote:
> On Wed, Feb 04, 2004 at 12:02:53AM -0500, Andrew Sutton wrote:
> > oh... and on a way off-topic note, i have my first of three potential
> > interviews with google tomorrow. wish me luck!
>
> Guid luck.
>
> I look forward to being able to say that Umbrello is used by Google :)

i can officially confirm two things...

1. google does not apparently use much in the way of uml so they almost 
certainly don't use umbrello :( makes one wonder how they design their 
software... hmm...

2. i will not be working working there :( i found out earlier this evening 
that i was apparently not a match. boo.

by the way, here's a question they asked during the interview. and nobody said 
i couldn't broadcast this, so...

suppose i give you an interface that returns to you the content of every 
document on the internet (i.e., the entire internet). how would you write a 
program to find the 10,000 most popular bigrams on the internet. a bigram is 
two consecutive words. for example, in "every good boy does fine", you have 
"every good", "good boy", "boy does", etc... and your solution should be 
constrained by modern hardware (assume limited hardware resource like 4 GB 
ram or something like that).

the other question was regarding the implementation of the google "safe 
search" feature (i.e, not returning porn sites in queries).

andy