Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
Victor Boctor
@vboctor
Registration for sonatype in progress. Hopefully we will have a maven snapshot package in the near future :)
Victor Boctor
@vboctor
I was surprised that SimpleDateFormat doesn’t provide a method that returns true/false based on validity of the format rather than throwing an exception and having to catch it (which is bound to be slower). In .NET, there is Parse() which throws FormatException, and TryParse() which uses true/false.
Mike Noordermeer
@MikeN123
It's not as if someone ever thought about the implementation of SimpleDateFormat (I believe it was once sponsored by IBM). But the TryParse construct is not really common in the Java world, the new java.time library does not seem to use it either (which would be the number one reason to switch the whole library to JDK8 only...)
Catching an exception is not necessarily slow, at least it used to be slower
A TryParse() implementation would be difficult anyway, since there are no such things as 'out params'
Victor Boctor
@vboctor
@MikeN123 makes sense that the pattern wouldn’t be common due to lack of out params. However, it is unfortunate that two parsing methods in the same framework throw different very different kinds of exceptions - different classes and checked vs. runtime.
We got registered to start publishing our maven package OSSRH Issue. Hopefully we will have a snapshot package soonish.
Erik van Paassen
@evpaassen
Java 1.6 backfires again... #176
Asbjørn Aarrestad
@ahaarrestad
@evpaassen Java 1.6 does not backfire. we have to work withing the given environment, and that is just a challenge. I'm just a bit surprised we did not see this during testing, but perhaps ci is running java 8 in java 6 compliance mode? thus not detecting the problem?
Erik van Paassen
@evpaassen
CI is running with JDK7, currently. The project is compiled against the 1.6 language level though, but this isn't a compile error obviously.
Asbjørn Aarrestad
@ahaarrestad
@evpaassen missing some junit-tests?
Erik van Paassen
@evpaassen
@ahaarrestad No? The tests would fail on Java 6, but no one ran the tests against Java 6.
Asbjørn Aarrestad
@ahaarrestad
@evpaassen Can we ask ci to jun jdk6?
Erik van Paassen
@evpaassen
@ahaarrestad No, Travis CI does not support Oracle JDK6: http://docs.travis-ci.com/user/languages/java/
Asbjørn Aarrestad
@ahaarrestad
@evpaassen just read that page. Don't know If I agree to the argument, though.... Will we get the same effect using openJDK6 ?
Erik van Paassen
@evpaassen
If you'd assume OpenJDK6 behaves exactly the same as Oracle JDK6. But most likely, it will not. Testing with OpenJDK6 tests compatibility with OpenJDK6, as testing with Oracle JDK6 would tell you something about compatibility with Oracle JDK6.
Asbjørn Aarrestad
@ahaarrestad
I agree. that's why I don't agree with the argument in the travis-ci docs.
you may want to move on, but you can't
Erik van Paassen
@evpaassen
Java 7 will become EOL too in April (http://www.oracle.com/technetwork/java/eol-135779.html), so we'll have to keep an eye on Travis JDK support.
Asbjørn Aarrestad
@ahaarrestad
that site states that oracle is supporting (for paying customers) jdk 6 until 2018.....
Mike Noordermeer
@MikeN123
Yeah, so unless Microsoft is going to pay to provide all contributors with an SDK, we can consider it EOL.
It should be fine to test against OpenJDK6 though, apart from some crypto differences it should be the same.
Erik van Paassen
@evpaassen
should
Noble Brown
@Noblebrown
Open JDK6 is better than nothing, even if it's not as good as using the Oracle JDK. It's disappointing that Travis CI doesn't support Oracle JDK6. EOL or not, Java 6 was around for over four and a half years before Java 7 was released. You couldn't pry it out of many enterprise organizations with a crowbar.
sohankhush
@sohankhush
Does ews-java-api supports OAuth ?
Mike Noordermeer
@MikeN123
No. EWS does not support OAuth (well, actually it sort of does a little bit of OAuth, but it is not too useful, since you need full mailbox permissions)
Victor Boctor
@vboctor
EWS service in Office 365 supports OAuth providing the ability for an app to access your mailbox with seeing your password. However, it doesn't support fine grained consents so like the OData API which supports consents like read mail, send mail, calendar and contacts. Ews-java-api doesn't support OAUTH though.
I meant without seeing.
Mike Noordermeer
@MikeN123
since there is no 'user consent' scenario for EWS, even with O365, I think it's fairly useless
if an admin needs to give full access, he can create a service account as well
Victor Boctor
@vboctor
I think OAuth gives benefits like: 1. app doesn’t see password, 2. constrained access, 3. ability to revoke. Would EWS OAuth give you 1 and 3, but not 2?
It would be use for developers of services / applications requiring such access rather than admin.
Mike Noordermeer
@MikeN123
EWS with application impersonation also gives you 1, and I'm not sure EWS OAuth gives you 3. on a per user basis
So essentially using OAuth with EWS instead of app impersonation is just swapping the (app specific) username/password for an OAuth token
Asbjørn Aarrestad
@ahaarrestad
anbyody got experience publising to sonatype?
the ticket is ok, so I guess we can start publishing snapshots any day now?
Mike Noordermeer
@MikeN123
afaik it's just a matter of configuring the deploy plugin in the pom.xml
but yesterday we discovered HEAD is currently borked with some firewalling configurations (e.g., TMG)
has to do with httpclient 4 cookie handling
still have to figure out a workaround for it
Asbjørn Aarrestad
@ahaarrestad
i have a local deploy at the moment, but want to change to a central one when available ....
Erik van Paassen
@evpaassen
Does anyone have experience with retrieving the EffectiveRights property of an Item while using a FindItems call? I'm experiencing some strange behaviour from EWS: http://stackoverflow.com/questions/28195941/ews-finditems-call-returns-incorrect-effectiverights-values
Matthias Leibmann
@mattleib
Note that EWS does support OAuth in Office 365, for both code and client-credential flow. The only difference to the Rest APIs is the permissions. EWS only supports "full access to mailbox" while Rest APIs do only support the granular permissions specific to a class of resources (Calendar, Mail, Contact). We recently updated the permission strings in the Application registration portal to better signal what is available for EWS. From a OAuth protocol point of view there is no difference. Note that maybe you had some issues in the preview phase where we marked the "full access to mailbox" permission as "admin only", meaning they required an admin to consent. With GA we removed this restriction. All permissions are now consent-able by an end-user. More here: https://msdn.microsoft.com/en-us/library/office/dn903761(v=exchg.150).aspx
Mike Noordermeer
@MikeN123
Thanks for the update. The client-credential flow seems to be a recent change we have been waiting for a long time ;-) I spotted your blog at http://blogs.msdn.com/b/exchangedev/archive/2015/01/21/building-demon-or-service-apps-with-office-365-mail-calendar-and-contacts-apis-oauth2-client-credential-flow.aspx
Victor Boctor
@vboctor
I’ve created the wiki for the project. Feel free to start creating content or move content from the readme file to there.
Russell Bolles
@Helmsdown
Wanted to confirm that a snapshot build has not been published to https://oss.sonatype.org/content/repositories/snapshots
Victor Boctor
@vboctor
Victor Boctor
@vboctor
Any suggestions on the rhythm of pushing the snapshot package out? Should we do this manually weekly + ondemand when there is a critical bug?