User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:45.0) Gecko/20100101 Thunderbird/45.0
The flag was changed from "-addmods" to "--add-modules" (see
). I don't quite understand the difference, but I believe that
it has something to do with making it ignorable by other JVMs (I
tried to summarize Tom's thoughts on the topic in the meeting
minutes). Adding the flag for testing purposes is simple enough;
it is, however, disappointing that they've stopped producing an
installer-free JDK (the latest one that I can find is Build 158 on
Jan 3/2017 ).
Tom, is the Equinox update likely to make it into Oxygen M6?
Because it fails. I didn't try https://bugs.eclipse.org/bugs/show_bug.cgi?id=493761#c9
and I assume that makes it work, though not on all VM
implementations and versions. It's also a hassle that it's
distributed (by Oracle) as a *.exe that installs it. I tried
these complicated setups for an older version and that worked:
But when I tried to repeat that yesterday for the latest
download, Firefox complained that the 64 bit Windows exe
contained a virus and the above steps didn't work on the
At this point I really don't want to run the Java installer
and have Java 9 ending up being the system default (if that's
what really happens when it's installed) or that the installer
removes older versions (as the installer tends to do).
Presumably the Equinox launcher will be improved soon.
From the notes I see one of the questions is, why can't the
June release already include Java 9 (beta) support? If some
last minute problem arises that can't make it into Oxygen.0, a
respin like the one for Neon in October could be produced,
affecting only the projects that leverage Java 9 features (and
the folks that produce packages and catalogs)...
On 01.03.2017 19:34, Wayne Beaton wrote:
Why can't we currently test running on Java 9? I've been
doing it for months.
On 01/03/17 08:56 AM, Ed Merks
Thanks for being clear on that issue. Support for
compiling and developing Java 9 falls primarily on JDT (I
assume/guess). But support for running the IDE itself on
Java 9 and making sure that works correctly is a burden that
falls on us all. Given we can't currently even test that,
it seems to me it's not a great idea plan that way. Like
you, I don't feel comfortable planning an Oxygen release
date around a Java release date. More flexible and less
risky in my opinion would be an Oxygen.0a update that could
respin primarily any JDT updates that are needed and would
not be so much focused on the statement that the IDE itself
will run on top of Java 9. I.e., more along the line of
what Nick is suggesting...
On 01.03.2017 14:48, Thomas
Bug 493761 needs to be fixed for Oxygen.0.
The issue of launching the Eclipse platform on Java 9
is different than having the support for compiling and
developing Java 9 classes.
I worry about delaying Oxygen.0 to align
with Java 9 release because I can easily see the Java 9
release getting delayed again. I would not want to hold
up Oxygen.0 for that.
I assume this must be fixed in a general way that
works regardless of whether Java 9 is actually used or
not and regardless of the JVM implementation
provider. Is it hopeful that this will be addressed
in time that we may all test the state of Java 9?
Does it really make sense to plan the whole release
schedule around this assumption? And as Marc
suggests, is planning a release during the high
vacation season a good plan?
On 28.02.2017 18:51, Marc Khouzam wrote:
Considering the dates are so close between
Oxygen.0 and Java 9,
I think having an extra release is more hassle
for adopters than it is worth.
So waiting for Java 9 is a good plan for
One drawback I see is that more people will be on
vacation at that time,
but with proper planning, it should not be a
for moving the Oxygen.0 release to the JDK9
Would make things for adopters of that
release easier, too.
> Note that I've added a biggie to the
list: with all the talk about having an
extra release to coincide with the Java 9
release, we've neglected to ask why we need
a release in both June and July. The Java 9
release date (which by all accounts is
pretty stable) is about a month after our
planned Oxygen date. Is there any reason why
we shouldn't just move the Oxygen.0 date
(other than "we've always done it that way")
> Feel free to answer here. It would be
great if you can come to the meeting armed
with an answer to that question.
> See you on Wednesday.
> For all phone lines: Participant
conference extension: 710 then enter pin
> North America (toll free)
> Germany (local call anywhere in
> France (local call anywhere in France)
> UK (toll free) 0800-033-7806
> Switzerland (local call anywhere in
> SIP clients: call 710@xxxxxxxxxxxxxxxxxxxx,
then enter pin 35498.
>  https://wiki.eclipse.org/Planning_Council/February_01_2017
>  https://wiki.eclipse.org/Planning_Council/March_01_2017
> Wayne Beaton
> The Eclipse Foundation
> eclipse.org-planning-council mailing