in reply to Re^4: Need for (XS) speed
in thread Need for (XS) speed

You sure put a lot of effort avoiding what surely has an O(1) cost of microseconds before a costly O(N) operation (according to the OP).

Replies are listed 'Best First'.
Re^6: Need for (XS) speed
by BrowserUk (Patriarch) on Mar 05, 2010 at 16:28 UTC

    It's not about performance. It's about reliability. The trap is infinitely more valuable than no trap. It really surprises me that you can't allow yourself to understand that.


    Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.
    "Science is about questioning the status quo. Questioning authority".
    In the absence of evidence, opinion is indistinguishable from prejudice.

      I hadn't read your post. I got tired of answering your questions going in 10 different directions. I have now.

      I'd use SvPOK over segfaults as an error check.

        I'd use SvPOK over segfaults as an error check.

        You are, of course, welcome to do that. But there is no good reason to.

        If SvPOK isn't set, you haven't pack'ed the scalar passed. And that isn't a data error that might occasionally arise at runtime. It is, (can only be!), a program logic error that will happen the first time, and every time, the function is called, until it is corrected.

        I hadn't read your post.

        So you respond to posts without reading. Too tired to read. Not too tired to flame. That explains a great deal, 60-odd levels of it.


        Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.
        "Science is about questioning the status quo. Questioning authority".
        In the absence of evidence, opinion is indistinguishable from prejudice.