| working on getting the simple chat client example going |
|
More...
|
over 23 years ago
|
| merging changes from cc |
|
More...
|
over 23 years ago
|
| working towards 0-8 release. I was just investigating why function calls to remote destinations appear to be quite slow when I just an acceptance test on local loopback. I ahve found that the connection manager win32 poll time out is effecting the speed at which things can be send. I.e. the poll only returns when the read of exception events occur, due to the current architecture. Decreasing the timeout from 10 seconds to 0.001 greatly increases performance however the load on the machine increases. I'm continuing to look into this. |
|
More...
|
over 23 years ago
|
| working on web site |
|
More...
|
over 23 years ago
|
| add the new revamped web-site. This is the same layout as the kioskcontrol site |
|
More...
|
over 23 years ago
|
| trying to track down the slow down reasons |
|
More...
|
over 23 years ago
|
| I currently adding unit test to catch the current slow down problem. XmlRpcConnectionTest contains the checkSendMethodCall which is currently timming out when it should be returning straight away. The test fails to that effect |
|
More...
|
over 23 years ago
|
| I just spent a few minutes profiling and I've discovered that the actual transport of data is really quick. What is actually killing performance is the hand over of data from the handleWrite to the function waiting on the queue to have data. Its occupying 90% of the time and typically is calling a single function call to take 2+ seconds. I'm seriously going to have to fix this. |
|
More...
|
over 23 years ago
|
| working towards the next release |
|
More...
|
almost 24 years ago
|
| updating the documentation |
|
More...
|
almost 24 years ago
|
| added code reference documentation |
|
More...
|
almost 24 years ago
|
| bring in changes from cc. Fixed blocking the connection manager whenever a new connection is received (problem in cc), add new flag remove which if set will cause the Connection to be removed as soon as possible from the connection manager |
|
More...
|
almost 24 years ago
|
| finished of refactoring, just trying to get the unittests working. |
|
More...
|
almost 24 years ago
|
| adding lockableDictionary test |
|
More...
|
almost 24 years ago
|
| adding new unittest and updating the old ones |
|
More...
|
almost 24 years ago
|
| finished merging code removing future dir |
|
More...
|
almost 24 years ago
|
| updated this example to use newly refactored lib code |
|
More...
|
almost 24 years ago
|
| working |
|
More...
|
almost 24 years ago
|
| removing unused files |
|
More...
|
almost 24 years ago
|
| I'm just adding unit tests |
|
More...
|
almost 24 years ago
|
| got the refactor working. I'm just adding unit tests |
|
More...
|
almost 24 years ago
|
| continuing to get poll on linux and other stuff going |
|
More...
|
almost 24 years ago
|
| refactored and remove notifier as it wasn't needed |
|
More...
|
almost 24 years ago
|
| got the unix/posix pollUsingPoll to work. |
|
More...
|
almost 24 years ago
|
| n |
|
More...
|
almost 24 years ago
|
| continuing to work on the future developent |
|
More...
|
almost 24 years ago
|
| continuing to work on the new xmlrpcstream single thread approach |
|
More...
|
about 24 years ago
|
| I've added a new directory to lib called future. This directory contains the files I am using to experiment with implementing a single thread to manage all sockets refactor. |
|
More...
|
about 24 years ago
|
| adding missing file |
|
More...
|
about 24 years ago
|
| updating manual |
|
More...
|
about 24 years ago
|
