Forward Scattering
The Weblog of Nicholas Chapman
- Posts
- 67 from 28 Apr 2013 to 13 Aug 2026
- Posting
- now and then (4 in the last year)
- Kind
- personal blog
- Feed
- http://forwardscattering.org/rss.xml
2026 4
2025 5
2024 1
2023 2
2021 1
2020 1
2019 3
2018 1
2017 4
2016 18
- Watch out - Qt headers suppress warnings with Visual Studio
- Functional Programming and Performance
- Visualising the topology of the rotations with the help of Portals
- Slow Visual Studio implementation of aligned malloc
- Electronic painter experiment #3
- Electronic painter experiment #2
- Electronic painter experiment #1
- Azure website is so slow
- Stack overflow internal compiler error in vs2015
- std::map performance bug #3 - slow find
- std::map and std::unordered_map performance bug #2 - slow count
- An apparent performance bug with some std::map and std::unordered_map implementations
- Special relativity, faster-than-light communication, and absolute time and space
- Things that suck about C++ #1
- Fast sorting on the CPU
- Hash functions, one-way functions and P = NP
- Permutations and sorting
- Slow pscp upload speeds
2015 12
- Finding annoying programs that steal window focus
- False Sharing Curiosities
- Indirection isn't slow, unpredictable indirection is
- A proposal for C++: bit_cast.
- A little Winter demo
- Closure support in Winter
- Unicode string handling in Winter
- Don't use LLVM's BBVectorize pass
- Auto-vectorisation and parallelisation in Winter
- Fuzz Testing
- Finding the fastest Turing Machine that computes a given function is uncomputable. (And the consequences for superoptimisation)
- A limitation of LLVM with regard to marking sret functions as pure/readonly
2014 8
- Reassociation optimisations in Visual Studio 2012
- Some wave simulation animations
- RDTSC doesn't actually measure the number of clock cycles elapsed
- Second Quantisation - a Quantisation Too Far
- More on Kolmogorov Complexity
- Decidability of finding the fastest Turing machine to compute a function with a finite domain
- Intuitive Quantum Electrodynamics
- Problems with JITing and LLVM #1 - large offsets