New issues in notarization of jpackage-created Mac applications (Scott Palmer)

[James Elliott]

Ah, thanks for those links. You are right, it is a huge project! And it sounds like progress is being made on these fronts, although I have not seen much discussion of the need to be able to turn on the hardened runtime and pass in an entitlements file. I will hope this is all converging on a working state in a future release, where I will be able to switch back to letting jpackage handle things rather than doing low-level surgery on the application image myself prior to signing and notarizing it. If anyone wants to have me try things (or even find out some of what I’ve learned getting and keeping this continuous integration pipeline working over the last several weeks) hopefully they will know where to find me. ^_^

Cheers,

	-James

> On Feb 9, 2020, at 14:18, Volker Simonis <volker.simonis at gmail.com> wrote:
> 
> I think you should also look at https://bugs.openjdk.java.net/browse/JDK-8238534 <https://bugs.openjdk.java.net/browse/JDK-8238534> for which a fix is currently discussed on build-dev (https://mail.openjdk.java.net/pipermail/build-dev/2020-February/026703.html <https://mail.openjdk.java.net/pipermail/build-dev/2020-February/026703.html>).
> 
> The sheer number of different mailing list has always been a big  problem for externals looking at the project. But in the he end, it's a huge project :)
> 
> James Elliott <james at deepsymmetry.org <mailto:james at deepsymmetry.org>> schrieb am So., 9. Feb. 2020, 18:52:
> Thanks for drawing my attention to that related bug, Scott. I would add a comment on it that Apple has updated the notarization process so this symlink is now rejected at that stage, not just by Catalina itself, but I lack standing to do so. (It still strikes me as strange that OpenJDK development is so insular compared to other open source communities I am involved with. ^_^) But I am glad I have at least found this mailing list.