Tasty » Adam Sampson's concurrency bookmarks

IPC Protocols - MozillaWiki edit / delete

Protocol specifications from the Mozilla project (they plan to model-check using SPIN).

to concurrency ipc model-checking mozilla spin ... on 03 April

Eventlet Networking Library edit / delete

Python concurrency library using greenlets -- effectively "Twisted done right". This looks worth experimenting with.

to concurrency greenlets networking python software ... on 24 March

TransactionalMemory - GCC Wiki edit / delete

GCC has implemented the ICC transactions extension (-ish). Hmmm. Perhaps this'd be worth a lecture in 0801?

to ag0801 compiler concurrency gcc transactional-memory ... on 03 December 2011

Userspace RCU | LTTng Project edit / delete

I could imagine uses for this in simulation -- perhaps Tim's transactional stuff?

to concurrency rcu simulation software ... on 15 November 2011

Daniel's Blog: Android messaging and concurrency (for native code development) edit / delete

Rapid overview of message-passing within Android.

to android concurrency message-passing phone programming ... on 29 September 2011

Gilbert Dhuime edit / delete

Gilbert's occam-pi manual (in French).

to concurrency occam teaching ... on 22 September 2011

High-Performance Synchronization for Shared-Memory Parallel Programs | University of Rochester Computer Science edit / delete

Code for assorted synchronisation operations. TBB uses this as a reference.

to concurrency parallelism research software synchronisation ... on 12 September 2011

simonmar/monad-par - GitHub edit / delete

A parallel monad from Simon M.

to concurrency haskell parallel ... on 02 September 2011

Chapter 1. Boost.Lockfree edit / delete

A prototype Boost-ish lock-free algorithms package.

to boost c++ concurrency lockfree ... on 05 August 2011

The Trouble with Erlang Concurrency | Tim Fox's blog edit / delete

The specific point is that the actor model makes it difficult to distribute a service (which isn't a problem when you have explicit channels -- you just use an any-to-any channel). The general point is that providing one model isn't helpful when that model doesn't work for your problem.

to concurrency erlang language-design research ... on 26 June 2011