core dump

[roger64]

@DiapDealer

You're right. Sigil is guilty of nothing. It's merely a fact that, I use jre for ODTImport and I need to get rid of this link every new jre version to get Sigil running. No more than that. I have just to be aware of this, I do not consider to be "bitten".

[KevinH]

The OpenJDK package includes a jre. One that should work properly.

[DiapDealer]

Originally Posted by KevinH

The OpenJDK package includes a jre. One that should work properly.

That's what I figure, too (though I've not tested it with the plugin in question).

Why use the unsupported aur package that you have to remember to "patch" after every update, when you can very likely use the supported OpenJDK package (from Extra) and forget about it? *shrug*

[Doitsu]

Originally Posted by DiapDealer

Why use the unsupported aur package that you have to remember to "patch" after every update, when you can very likely use the supported OpenJDK package (from Extra) and forget about it? *shrug*

@roger64: I've installed OpenJDK (jre8-openjdk and jdk8-openjdk) on my Arch Linux machine and never encountered any problems with it. You might want to give it a try.

[DiapDealer]

Thanks for confirming!

[eschwartz]

Originally Posted by DiapDealer

For further reference, the following note/warning has been added to the jre 9.x.x aur package's installer script:



Note that in the event where the jre 9.x.x aur package is an absolute requirement, removing the above symlink is sufficient to restore Sigil (and other programs)'s ability to run.

The Arch-supported OpenJDK 9 package does not install a Java NPAPI browser plugin. The extra package "icedtea-web" provides a Java browser plugin for OpenJDK (and doesn't appear to crash qt5 webkit programs).

Oh wow, that again.

This is... arguably a bug in the jre package's inclusion of broken stuff, but I don't really expect that AUR maintainer to make the sensible call there... one more reason why people should stick with the tried-and-tested packages in the official repos. Including broken java web plugins is sort of questionable judgment.

[DiapDealer]

Originally Posted by eschwartz

Oh wow, that again.

This is... arguably a bug in the jre package's inclusion of broken stuff, but I don't really expect that AUR maintainer to make the sensible call there... one more reason why people should stick with the tried-and-tested packages in the official repos. Including broken java web plugins is sort of questionable judgment.

Webkit isn't entirely blameless in this either. It's often looking to load plugins' shared libraries from hardcoded paths--even when web plugins have been disabled via the api.

Othewise, I entirely agree. I don't install much of anything from the AUR anymore.