Available for streaming and download, Google Native Client presented at UW’s computer science lecture series. Covers the restrictions on x86 code, new alignment rules, and performance on various benchmarks. 5% overhead, that’s nothing compared to many other sandboxing techniques.
Native client is 50KB download? Wild. It really is just a gatekeeper, runtime library separate.
Of course, I would love to get away from x86. LLVM, or ARM, or even Amd64. x86 makes me a sad panda.
Raymond Chen writes in “We’re using a smart pointer, so we can’t possibly be the source of the leak“. The most immediate cause is a subtle misuse of CComPtr, using operator= which performs an AddRef on a return value that has already been AddRef’d, leading to one AddRef too many.
The less immediate failure was a poorly designed API.
Well, not poorly designed. Unfortunately designed.
Continue reading “On the value of consistent API design”
Personal picks for PDC 2009 thus far. Haven’t watched them yet, so no reviews.
Small complaint on behalf of those who were physically in attendance: I imagine it’s really annoying to find related topics always be in the same slot. Concurrency Fuzzing and C++ manycore; Axum and F# Parallel/Async; I would think it would be annoying to have to choose between those pairs.
Of course, I always watch online after-the-fact when the videos are up, so no negative consequence for me.
Just a quick post, haven’t shared anything in a while, but found this discussion on loop variables and closures interesting — well, inflammatory, actually.
Continue reading “Loop variable closure semantics”