Pie-thon Update
Slides from my OSCON presentation.
Spotted in the audience: Jim Hugunin, Patrick Michaud, Guido van Rossum, Chip Salzenberg, and Larry Wall.
Sam's talk at OSCON
Hello to everyone coming here from Sam Ruby’s talk! Well, we had the test cases working last night and suddenly today they all fail because the entire python_group got disabled in the parrot source tree. Sigh :/ Anyway, if you weren’t there, you...Excerpt from pirate blog at
Any chance you could use your pull with O’Reilly to get them to have a conference east of the Rockies any time soon?
Posted by Christian Romney at
Ken: no. As I said, my focus was more on language compatibility.
Christian: here you go, though it looks like I may have overshot slightly. Oops!
Posted by Sam Ruby atSam Ruby: Pie-thon Update
was hugely sorry to miss Sam’s session...Excerpt from del.icio.us/tag/python at
Another notable in the audience: Rasmus Lerdorf.
So you had the creators of PHP, Perl, and Python in attendance at your session — I wonder how many other speakers can claim that level of interest :)
Posted by Dan Scott at
Random thought: do you have a feel for how much method dispatch via the byte-code interpreter has an effect on overall performance?
Perl5 and CPython use several C function calls and a hash lookup. Apple/GNU Objective-C uses a C function call and sparse arrays. David Stes' Objective-C (POC) uses an inline cache with a fallback to a C function and a sparse array. C++ uses a single-instruction indirect-jump through a static vtable.
I came up with a technique for building a dynamic, single-instruction indirect-jump vtable and I’m wondering if it’s still relevant.
Posted by Ken MacLeod at