| 1 |
505071e6 |
Steven |
pingpong: working and complete
|
| 2 |
9b4fbfe5 |
Steven |
|
| 3 |
505071e6 |
Steven |
auction: working but doesn't have the timeout, end of auction stuff
|
| 4 |
9b4fbfe5 |
Steven |
|
| 5 |
505071e6 |
Steven |
bank: incomplete
|
| 6 |
9b4fbfe5 |
Steven |
|
| 7 |
505071e6 |
Steven |
telephone: versions first through sixth are complete and working
|
| 8 |
505071e6 |
Steven |
seventh is incomplete/not-working
|
| 9 |
9b4fbfe5 |
Steven |
|
| 10 |
505071e6 |
Steven |
im: single/fox: working
|
| 11 |
505071e6 |
Steven |
single/wx: working but incomplete
|
| 12 |
505071e6 |
Steven |
distributed: very incomplete
|
| 13 |
9b4fbfe5 |
Steven |
|
| 14 |
9b4fbfe5 |
Steven |
fib: a fib computation after Doug Lea's Fork/Join framework The goal
|
| 15 |
9b4fbfe5 |
Steven |
of this is to show speedup on jruby, which has real
|
| 16 |
9b4fbfe5 |
Steven |
concurrency. Right now, this is flakey. First, it doesn't bound
|
| 17 |
9b4fbfe5 |
Steven |
concurrency (Doug's whole point) which gets out of control in the
|
| 18 |
9b4fbfe5 |
Steven |
original version. The conservative version does weird things for
|
| 19 |
9b4fbfe5 |
Steven |
me on linux. But so does the pure thread version (though the
|
| 20 |
9b4fbfe5 |
Steven |
dramatis version is weirder than the thread version). |