Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: help needed: Fwd: [epp-dev] Galileo SR1 schedule

Hi Markus,
We did fix bugs that is in the same vicinity of the code:

https://bugs.eclipse.org/bugs/show_bug.cgi?id=286498 (rev 10521)
https://bugs.eclipse.org/bugs/show_bug.cgi?id=286413 (rev 10511)
https://bugs.eclipse.org/bugs/show_bug.cgi?id=280721 (rev 10513)

perhaps that somehow caused a regression in the EPP builds. This happened in mid August. Can you try using our archived headless site:

http://www.eclipse.org/downloads/download.php?file=/tools/buckminster/archivedsites/org.eclipse.buckminster.site.eclipse.headless_1.1.350.r10510.zip

and if that makes a difference, also test:

http://www.eclipse.org/downloads/download.php?file=/tools/buckminster/archivedsites/org.eclipse.buckminster.site.eclipse.headless_1.1.350.r10522.zip

That way, we can see if it is a regression. It would also help narrow it down a bit.

Thanks,
Thomas Hallgren


On 09/18/2009 09:55 PM, Markus Knauer wrote:
Hi Thomas,

no, unfortunately it happens on *all* platforms that I tested (win32,
linux64) and as reported on macosx.x86_64.

Regards, Markus


2009/9/18 Thomas Hallgren<thomas@xxxxxxx>

Hi Markus,
Does this apply to this particular platform only or is it apparent in other
platforms as well? The reason I ask is that I don't think that macosx.x86_64
is one of the platforms that are included in the Galileo repository and IIRC
the Galileo repository is the source of input for the build.

- thomas




On 09/18/2009 06:19 PM, Markus Knauer wrote:

Hi Thomas,

I just checked what Eric wrote below and found out that he is not only
right, it is a real problem in the packages that I haven't seen in the
past:

In the .product files the packages define their own product and add this
parameter to the eclipse.ini - at least that is the plan and it worked all
the time.

    <launcherArgs>
       <programArgs>-product
org.eclipse.epp.package.java.product</programArgs>
      ....

Is this something that could be caused by Buckminster?

Or could it be caused by the additional Buckminster "build" step before
the
"perform" that I had to introduce in order to build the WTP intro pages
and
the capabilities plugins?

And why is the EPP repo build now creating those binary artifacts as can
be
seen in

http://build.eclipse.org/technology/epp/epp_repo/galileo/epp.build/buildresult/org.eclipse.epp.allpackages.feature_1.2.1-eclipse.feature/site.p2/
?

Thanks and regards,
Markus


References:
- Repository build:
https://build.eclipse.org/hudson/job/epp-repository-build/228/
- Output: Build 20090918-0704 from
http://www.eclipse.org/epp/download.php
- Exemplary product file:

http://dev.eclipse.org/viewcvs/index.cgi/org.eclipse.epp/packages/org.eclipse.epp.package.java.feature/epp.product?root=Technology_Project&view=markup




---------- Forwarded message ----------
From: Eric Rizzo<eclipse-mail@xxxxxxxxxxxx>
Date: 2009/9/18
Subject: Re: [epp-dev] Galileo SR1 schedule
To: Eclipse Packaging Project<epp-dev@xxxxxxxxxxx>


First thing I notice about today's macosx.x86_64 build of the Java package
is that the branding (About dialog, Welcome screen, splash screen [is that
even branded for packages?]) seems to be from the plaform, not the Java
IDE
branding.

Should we file bug reports for this stuff?

Eric

_______________________________________________
epp-dev mailing list
epp-dev@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/epp-dev








Back to the top