@ Daniel: Nick has been very generous to provide you a patience-filled response but I do feel what you said was not something that really made sense. I’ve taught data structures at a university but still I’d like to have a look at the bubble sort algo once to recollect if i am 100% accurate.

We in the real world care more about the jobs that we have to get done: the real world problems, and not the stories about what Dijkstra said in 1870s.

@ leonid: I think you’ve just been unlucky OR your country doesn’t have too many developers (which I doubt to be a fact). PHP is a simple language. Here, ’simple’ means easy enough for those with 0 years of experience as well as 10 years of experience. I’m also of the opinion that PHP is just a tool, look for a guy who knows the Internet and web based stuff (even on jsp/aspx) and he should be good for you as he’lll surely be able to pick up PHP in a week at max I’d say.

I find this a very interesting quote and when I think about it…if hackers keep us software designers, and network engineers on our toes for inventing new and different way to break the rules and our systems, perhaps we need to be as good as the hackers in coming up with our solutions.

The difference is, the hacker is there to get something done and damn the torpedoes. What many software designers do is get too involved with the “rules” and the purity of their solution. We get caught up in our own “educated” ways of doing things and the “hackers” come along and kick our asses. Why? Its certainly not because they’re dumb…its because they’re not getting hung up all the rules.

The obvious caveat to all of this is the “script kiddy” who uses the hackers tools to the work.

I just found that quote interesting and it got me thinking. Perhaps we as designers need to let go, just a bit, and start thinking about accomplishing a task and not the purity of its implementation, etc…

@Leonid: Fair enough – as long as there’s some leeway with the pen and paper test (in terms of looking for a logical approach, and something that would work if any syntax errors and trivial stuff were fixed) – then I have no problem at all with that. Also if some was was actually volunteering that bubble sort was the way to go and then proceeded with implementing it from scratch, then yes, you probably want to avoid that person, because they’ve got both the theory and the PHP application of it wrong. :-)

@Daniel: I agree that some theory is important as a basis for understanding why things are the way they are, and what the fundamental limitations are. But for comparing sort algorithms, that’s a pretty pure theoretical CS area, and it’s worth learning once, but it’s not really worth remembering all the details of every possible algorithm and their order (IMHO). Instead, all you need to know is that O(n log n) is the best we can do, that qsort is O(n Log n), and that PHP’s sort algorithm uses qsort: ( http://cvs.php.net/viewvc.cgi/ZendEngine2/zend_qsort.c?view=markup ). Once you know that, then you know that using PHP’s inbuilt sort function is the way to go, and you can explain why. But I don’t personally believe that knowing too much more detail than that is especially useful or necessary. And I agree that being able to define what a stack and a queue are is a reasonable question – but interviews are stressful situations, so even good people can make silly mistakes or fluff a definition – so having a fall-back question about the application might help. But as long as there’s a balance of application with the most important areas of the theory, then that’s probably the best approach.

@Abraham: Yeah, I know what you mean – I frequently have to have lots of reference sources open and combine them to get something done.

@Ben: I have a sneaking feeling that someone might have pre-warned Obama to give that answer, but it was amusing :-)

http://youtube.com/watch?v=k4RRi_ntQc8

[quote]What I cannot understand is why people with more than one Bachelor Degree in Computer Science recommend using bubble sort.[/quote]

This is perfectly valid: Dijkstra said that the only thing Bubble Sort had going for it was the name. No qualified developer should recommend a O(n^2) algorithm, when O(n log n) algorithms exist for the same job.

[quote]What is the difference between the stack (also known as FILO) and the queue (also known as pipe, also known as FIFO)?[/quote]

Again, a perfectly valid question. Stacks and queues are fundamental data structures. Your alternative is an entirely different question, which asks about practical implementation of such structures.

I cannot agree that just being able to produce code makes one a good developer, as you imply. It is the understanding of how the underlying data structures, algorithms and logic are working – exactly as the original article outlines.