Age | Commit message (Collapse) | Author |
|
With the 19.3.0 release, GraalVM supports two Java versions: Java 8 and
Java 11. This also affects all sibling packages, including GraalPython,
so the former graalpython-bin package is split in two:
graalpython-jdk8-bin and graalpython-jdk11-bin. To leave old installs
functional when they update, this package is retained, but only as a
“virtual” package with no contents except for its dependencies.
Since this package only used to provide the Java 8 version, it declares
a dependency on graalpython-jdk8-bin now (users may, if they want to,
install graalpython-jdk11-bin additionally or instead). Users should
explicitly install the real package; I will remove this one eventually.
|
|
|
|
|
|
The top-level license files turned from symlinks into proper files
(oracle/fastr#90 [1]), so now we have to copy those as well.
[1]: https://github.com/oracle/fastr/issues/90
|
|
|
|
|
|
|
|
The first non-RC release changed the source file name pattern a bit, and
standardized the license file names to end in .txt, but otherwise the
package format remained stable.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
With -t, the target is always treated as a directory, including the last
component. Instead, swap the arguments back into SOURCE DEST order, and
use the -T option to ensure that DEST really is a file.
|
|
Upstream thankfully changed the URL format to include the version, so
existing files from builds of previous package versions should no longer
be mistaken for the current file (this was a possible problem with rc2).
|
|
|
|
The java-runtime-common package already adds /usr/lib/jvm/default/bin to
the path, but this makes graalpython work even when another JVM than
java-8-graal is used as the default JVM.
|
|
This is a binary package (we don’t build graalpython from source), so it
should have a -bin suffix.
|
|
PKGBUILD copied from truffleruby with slight tweaks (different GitHub
organization, no docs/ directory).
Similar to Ruby, launching graalpython with --polyglot doesn’t work, but
accessing Python from R works. For example:
eval.polyglot('python','[i for i in range(1,20) if i%2 is 0]')
|