On Sat 5 May 2018 at 09:13, Michael Osipov <[email protected]> wrote:

> Am 2018-05-05 um 09:15 schrieb Stephen Connolly:
> > On Thu 3 May 2018 at 22:10, Michael Osipov <[email protected]> wrote:
> >
> >> Am 2018-05-02 um 10:41 schrieb Robert Scholte:
> >>> I don't see a new test[1][2], only rewrites to confirm there's no
> >>> regression.
> >>>
> >>> I understand the issue, but we just need to be sure that nobody in the
> >>> future thinks that File.toURI() is short for File.toPath().toUri()
> >>>
> >>> Robert
> >>>
> >>> [1]
> >>>
> >>
> https://github.com/apache/maven/commit/43b34598629f086523a333dc18665389643832a5
> >>>
> >>> [2]
> >>>
> >>
> https://github.com/apache/maven-integration-testing/commit/5e18bb18784585dfc822038f5229785d439c677b
> >>
> >> Robert,
> >>
> >> done some more digging and added unit and integration tests to it:
> >>
> >>
> https://github.com/apache/maven/commit/9d29bb4d91e9545a9b5bc2957646ad42d5add210
> >>
> >>
> https://github.com/apache/maven-integration-testing/commit/1f4912cda5f49b6ba36a2693bb5c1701ea4d9b86
> >>
> >> Though, the IT reasonably runs on Unix if UTF-8 is set as file.encoding.
> >> I have also updated the ticket with more description.
> >>
> >> I think that this issue should be pushed to 3.6.0.
> >
> >
> > Unclear why you think it should be dropped from 3.5.4. We’re not changing
> > an API, only fixing a bug...
>
> This fix introduces a change in behavior. It might be unexpected in a
> fix release. If we are still OK for 3.5.4 then I am fine too.


Well if we are worried about any change in behaviour then we shouldn’t make
any changes ever.

The question is whether this is “significant”.

One of my objections to semver is that it only focuses on APIs... so under
semver:

1. No breaking API method signature changes => no major version bump
2. No new API methods => no minor version bump
3. Must be patch

Now in the case of Maven, I argue for semver is a poor fit, largely because
we have multiple APIs, SPIs and UIs each of which have different measures
of breaking changes... but we only have one version number

We need to get 3.5.4 with the jansi fix out... simple bug fixes are fine
for 3.5.4 too.

then we focus on the next version (3.6.x or 4.0.x) and move on... but we
need to move the baseline to Java 8 at this stage and get Java 10 support
better (both of which i’d like to see in a 4.0.x so if we are missing those
then it would be 3.6.x)


>
> Michael
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [email protected]
> For additional commands, e-mail: [email protected]
>
> --
Sent from my phone

Reply via email to