Github user metlos commented on the issue:
https://github.com/apache/tinkerpop/pull/494
So in another words it is a matter of setting up the level of breakage you
want to allow between different Tinkerpop versions. I based the "minor increase
allows breaking changes" on th
Github user metlos closed the pull request at:
https://github.com/apache/tinkerpop/pull/494
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature
Github user metlos commented on the issue:
https://github.com/apache/tinkerpop/pull/494
Hmm, maven central seems to be taking its time - it still doesn't have the
latest revapi-java version available (which is why CI failed for the latest
commit).
But anyway, I've updated
Github user metlos commented on the issue:
https://github.com/apache/tinkerpop/pull/494
I'm near a new release of Revapi that contains some important fixes (like
wrong classification of a return type change to a covariant type). So
everything I wanted to do to this PR was to update
Github user metlos commented on the issue:
https://github.com/apache/tinkerpop/pull/518
Created #519 with the updates as suggested here.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have
GitHub user metlos opened a pull request:
https://github.com/apache/tinkerpop/pull/519
Make ElementHelper.areEqual(Property, Object) handle nullsâ¦
⦠so that it can be used correctly in equals() methods of Property impls.
Added test methods for additional equality
Github user metlos commented on the issue:
https://github.com/apache/tinkerpop/pull/518
Thanks for the feedback. I'm closing this PR then and will target the
reworked patch to tp32.
---
If your project is set up for it, you can reply to this email and have your
reply appear
Github user metlos closed the pull request at:
https://github.com/apache/tinkerpop/pull/518
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature
GitHub user metlos opened a pull request:
https://github.com/apache/tinkerpop/pull/518
Honor the equals() contract on Property impls by allowing nulls as valid
comparisons.
If user code put the properties into some kind of collection together with
`null` values, things would blow
Github user metlos commented on the issue:
https://github.com/apache/tinkerpop/pull/494
@okram the API changes seem to be due to your changes in 05bfb029. You
could try and fill out your first intentional API changes list ;-)
(Most of the below is just git extravaganza
Github user metlos commented on the issue:
https://github.com/apache/tinkerpop/pull/494
> org.apache.tinkerpop.gremlin.structure
> org.apache.tinkerpop.gremlin.structure.io
> org.apache.tinkerpop.gremlin.process.computer
> org.apache.tinkerpop.gremlin.proce
Github user metlos commented on the issue:
https://github.com/apache/tinkerpop/pull/494
`.travis.install-maven.sh` is a workaround I used in another project for
https://github.com/travis-ci/travis-ci/issues/4872. Trusty, which is the distro
used on Travis for Tinkerpop builds
Github user metlos commented on the issue:
https://github.com/apache/tinkerpop/pull/494
Yes, Java makes it somewhat difficult to properly declare "module
boundaries" - let's hope Java 9 will do that right.
But for now, it is often useful to declare certai
Github user metlos commented on the issue:
https://github.com/apache/tinkerpop/pull/494
K, finally the CI fails for the right reasons :wink:
https://travis-ci.org/apache/tinkerpop/builds/176007319#L3026
---
If your project is set up for it, you can reply to this email and have your
GitHub user metlos opened a pull request:
https://github.com/apache/tinkerpop/pull/494
TINKERPOP-1443 - Introduce API check into the build
This introduces API checks into the build and fails the build whenever a
"severe" change is introduced that would break either source
15 matches
Mail list logo