This message was deleted.
# community-support
s
This message was deleted.
✅ 1
v
Is this a retype error, or did you actually miss the
t
of
jvmopts
?
And even with the
t
, it would be
org.gradle.jvmargs
😄
w
uh... retype error, and I'll have to check tomorrow if I tried opts or args. I appended to a line colleagues added (way way) earlier.
v
Well, another option would of course be, not to use internal classes 😄
w
if there were other options to decompile bytecode...
(both ways. Alternatives for decompiling bytecode and alternatives to decompiling bytecode...)
v
I just tested. With
org.gradle.jvmargs
in
gradle.properties
it does work as expected.
But still I'd recommend to eradicate the need to use an internal class.
Alternatives for decompiling or to decompiling I can hardly name, as I have no idea what you do with that.
w
we parse compiled test classes, looking at certain annotations, to determine which test databases the test set to be run will need. It's a heap of cruft that needs to be lit on fire and rebuilt from the ground up, but we don't currently have the time to do that properly...
it is jvmargs, in gradle.properties too, and it somehow does not work here 😢
v
If you really need to read the bytecode in the meantime, use some library like
asm
. If you want to invest in a slightly different but still awkward solution, write an annotation processor that writes sind file with the necessary information.
If you really need to use that internal class, can you share an MCVE where it fails despite jvmargs?
w
I ran gradlew with and without
-Dorg.gradle.jvmargs='--add-exports=java.xml/com.sun.org.apache.bcel.internal.classfile=ALL-UNNAMED'
now, with
--info
in both cases, and with gradle.properties's
org.gradle.jvmargs
containing the same --add-exports in both cases: in the beginning it's starting a Gradle Daemon with that --add-exports when it's got the -D, and without if it didn't get the -D
I'll see about the MCVE...
no, I won't. I replaced the offending code with other code using
asm
.
👌 1
Thanks for that hint!
👌 1