Shibboleth Developer's Meeting, December 20, 2013
Attendees:
Call Administrivia
Dial-in attendee identification.
Next call is next Friday. Any reason not to meet ?
60 to 90 minute call window.
Brent
Daniel
Ian
FYI:
- Ohloh (Oh no!)
- MDA users
- UKf tooling
I'd like to get decisions on:
- Requirement for Servlet 3.0 in our V3 stack.
- Cleanest option is to just say yes. This would restrict deployment platforms for the V3 IdP.
- It might be possible to fudge it somehow, though. How much effort do we think that is worth?
- Final option would be revisit the decision to move to Spring Framework 4 and stick with Spring Framework 3 for the life of IdP V3, but I don't think we want to do that.
- Use of
@Override
. I haven't heard any objections to it on the list, but let's be clear about what we want to do; some selection or all of the following:- Change (reverse) the recommendation for the "insert
@Override
in new code" option in Eclipse? - Add a code style recommendation to include
@Override
where appropriate? - Should we do any work on any of the old code, or leave it to individual discretion?
- Should we use Checkstyle's
MissingOverride
check: "Verifies that thejava.lang.Override
annotation is present when the {@inheritDoc} javadoc tag is present."- Would probably involve quite a lot of warnings.
- Note that it is a purely textual check; Checkstyle doesn't seem to know when a method actually does override.
- There is also a Java compiler preference in Eclipse (settable either at workspace or project level) which defaults to ignoring missing
@Override.
- We currently set this explicitly to
org.eclipse.jdt.core.compiler.problem.missing
Override
Annotation=ignore
in the parent project's preferences. - We could change this per-project setting or remove it to allow the personal (workspace) preference to take effect.
- Change (reverse) the recommendation for the "insert
Rod
Mostly busy elsewhere. I'd like to get direction on what we do about SVN resources.
Scott
First half of week spent on RelyingParty and ProfileConfiguration code, ended up creating some interfaces, not sure it's really worth it now that I see more of the design
Syncing settings from V2 into V3 profile config classes
Worked on relying party and profile config selection actions, and rewired testbed to use them in place of mock actions
Last half of week spent reviewing Daniel's check-in and refactoring the serializer code a bit to handle our init model, and do some package reorg'ing
Tom
Other