News for hackers. Stuff that matters.
Re:You're Doing It Wrong: CS in the real world by acqq on Saturday June 12, 2010 @03:20PM (#1426394)
The article is really poorly written. The specific example is very specific, the claims that "nobody else gets it" are in fact unsubstantiated. All the effects he mentions are very known among right professionals. And the title text is not about "relative latency" you mention but about the cost of paging, once something gets paged out.
Re:You're Doing It Wrong: CS in the real world by stcredzero on Saturday June 12, 2010 @03:48PM (#1426425)
The specific example is very specific

Sounds good to me.

the claims that "nobody else gets it" are in fact unsubstantiated

I'm pretty sure that the entire shop at my last consulting job would all claim it was news to them. Most people don't have to deal with issues like the ones discussed in the article, so it's perfectly understandable that it would be news to them. I don't have to deal with such issues on a daily basis, though I think I would have been able to do the same analysis if I had his job.

And the title text is not about "relative latency" you mention but about the cost of paging

I didn't know "relative" was so associated with memory and paging that you could jump to a conclusion about my specific meaning. (And actually, I meant to be quite general.) AFAIK, the cost of paging has something to do with the latency of resident memory vs. latency memory that was paged out.

Re:You're Doing It Wrong: CS in the real world by Confusion on Sunday June 13, 2010 @08:50AM (#1427368)

  I'm pretty sure that the entire shop at my last consulting
  job would all claim it was news to them
On the other hand, I'm pretty sure that the same doesn't hold for the actual target audience of this article.
Re:You're Doing It Wrong: CS in the real world by mfukar on Sunday July 11, 2010 @02:25PM (#1505892)
I'm willing to forgive his "poor" writing on account that he's right - most CS students and graduates don't have any clue about cache effects on algorithm implementations, and that's just wrong.