about the interview: so... I got there about 5 minutes late, but it wasn't a big deal. interviewed with "one of the developers" first, and I liked him. first question: implement the fibonacci sequence. I had a pretty nifty solution, and while stepping through it for him fixed one bug. some box+pointer diagrams. and I mentioned heap vs. stack memory, which is good to know, some. second question I bombed. "trace through this code and find the memory leak". very subtle. a good question. [it was a stack implementation that "leaked" when objects were popped. it just decremented the storage pointer without 'nulling' what had been stored. so anything that was in the array would remain uncollected.]
then he was going to ask a question about notify vs. notifyAll and I told him I didn't know anything about java2 threading. and we chatted about this and that. apparently postgresql 7.3 upgrades nicely other than having to change foo = NULL to foo IS NULL in queries (which is how I already code).
discussed stuff I'd learned about postgresql recently (vacuuming and how to design tables for that and such). and he asked a question about design patterns; I told him what I knew of them, so he asked a "simple" question. I recognized the pattern, at least. "Singleton". though I couldn't remember why they were good. but we discussed implementation a little and I got to name another design pattern there. [Factory] and I got to ask a question about singleton that I had been wondering about and got to mention that I thought the java folks had read the book and used its naming scheme rather extensively. maybe bonus points there.
then the second interview, lead developer... he hadn't actually looked at any of my code. had printed out some of it. So I talked him through it some. found one bug. and he explained to me java2 threading. at which point, using that, we verbally reworked my answer some. and then we discussed a small problem based off of the connectionpooling stuff. went over the projects on my resume, to some extent, what I did at poemranker, what I'd learned at metazoa, what I'd learned recently.
I think I spent about 20 minutes with each.
next week I should get a call as to whether I get a second interview.
They say they need developers ASAP. I need a job sooner than that. la la la. :)
then he was going to ask a question about notify vs. notifyAll and I told him I didn't know anything about java2 threading. and we chatted about this and that. apparently postgresql 7.3 upgrades nicely other than having to change foo = NULL to foo IS NULL in queries (which is how I already code).
discussed stuff I'd learned about postgresql recently (vacuuming and how to design tables for that and such). and he asked a question about design patterns; I told him what I knew of them, so he asked a "simple" question. I recognized the pattern, at least. "Singleton". though I couldn't remember why they were good. but we discussed implementation a little and I got to name another design pattern there. [Factory] and I got to ask a question about singleton that I had been wondering about and got to mention that I thought the java folks had read the book and used its naming scheme rather extensively. maybe bonus points there.
then the second interview, lead developer... he hadn't actually looked at any of my code. had printed out some of it. So I talked him through it some. found one bug. and he explained to me java2 threading. at which point, using that, we verbally reworked my answer some. and then we discussed a small problem based off of the connectionpooling stuff. went over the projects on my resume, to some extent, what I did at poemranker, what I'd learned at metazoa, what I'd learned recently.
I think I spent about 20 minutes with each.
next week I should get a call as to whether I get a second interview.
They say they need developers ASAP. I need a job sooner than that. la la la. :)
Comments
public class Fib {
protected static int[] fibs = {0,1};
public static synchronized fib(int n) {
if (fibs.length >= n) {
int[] fibtemp = new int[n];
int i=0;
for (;i<fibs.length;i++) { fibtemp[i] = fibs[i]; } for (;i<n;i++) { fibtemp[i] = fibtemp[i-1]+fibtemp[i-2]; } fibs=fibtemp; } return fibs[n]; } } cute, no? ;)