I'm getting much more complicated errors. Here is one that relates to multiple repositories, where the first stage is firing off when one of the repo's (vault.exe) fails. There are lots of UX issues shown up in the video of this issue. However it's getting noticeably better over the last few weeks.
See http://www.screencast.com/t/FQ4Fr1ST the "Repo's Revolt edition".
Summary of issues
- Private messages list is still difficult to use as everything is listed "A new system event was logged". It's next to useless information. I have 50 of them.
- The time workflow between repository changes (and errors that might be triggered) to the execution of the first stage isn't working as expected. I'm getting the first stage firing without all my repo files being available.
- There are lot's of views where the logging information is inconsistent for trouble shooting problems.
- After upgrading the software the first set of builds that run seem to "automatically" fail. My ever reliable Codeplex SVN repository failed, then I manually ran it and everything just worked. That inconsistency is a concern.
- There are still lots of problems around repository connections and error handling and how far you need to search within the UI to find these issues.
Noticeable improvements
- Lots more " + Help" buttons.
- Improvements on the forms to include more practical information instructions on how to enter data, so you don't have to constantly open the docs. COOL.
- Improvements in "build initialising" hangs. I am still getting them when my server goes into backup meltdown.
- Much easier to move around the configuration history and build history, and more information populating correctly.