|Hudson Continuous Integration Server|
|Mailing List • Forums • IRC|
|Community Meeting 20-Feb-2012|
Add or comment on agenda items here, just prefix the item with your name so we can identify you on the call:
- What is the status on the next release on the 2.x line?
- [susan] M2/M3 Release
- [susan] Website and wiki
- [susan] Initial Code Contribution
- [susan] Issues migration
- [Winston] Hudson performance in a enterprise installation, defining best practice, dedicating a release for performance analysis.
- [Henrik] What are the core team's priorities besides the eclipse migration? upcoming features?
- [Henrik] Out approach to scripting Hudson seems half-hearted at best, how do we improve this? (We need a unified way that plugins can easily attach themselves to)
- CLI works: but is limited in its API plus requires special access through firewalls etc.
- old "REST" api: I can configure a project, but it generally lacks in api
- new "REST" api, I personally find it confusing, eg. there is a difference bewteen project/<id> and projects/<name> in the latter I can get the config, nodes doesn't support configuration at all.
- [Henrik] Collaboration on testing the hudson UI. Right now we have several testing "frameworks" in play for hudson. Right now we have
- hudson-test-framework (part of core): Don't know it but assume it is for plugin developers writing their own tests
- hudson-test-harness: Extensive "white-box" testing of hudson
- hudson-test-ui: Small selenium based testing package primarily for cascading projects
- plugin-tester (https://github.com/hudson-plugins/plugin-tester): My prototype for testing a plugin in hudson using selenium tests. Similar to hudson-test-ui except my version tests against multiple hudson cores, and allows each test class to specify which plugins should be installed.
I suggest we look into unifying hudson-test-ui with my prototype, but where should it be located and which plugins should be tested?