Gábor Boskovits <boskov...@gmail.com> writes: > 2018-03-01 18:11 GMT+01:00 Ricardo Wurmus <ricardo.wur...@mdc-berlin.de>: > >> Hi Guix, >> >> we have a problem with jar manifests. When we use the Class-Path >> property to ensure that an executable can find its dependencies on the >> class path at run-time, we end up with a manifest like this: >> >> --8<---------------cut here---------------start------------->8--- >> Manifest-Version: 1.0 >> Class-Path: /gnu/store/i28vi94r8z9f0x02zgkrv87w16ibmqkw-java-htsjdk-2. >> 10.1/share/java/htsjdk.jar >> Created-By: 1.8.0_151 (Oracle Corporation) >> Main-Class: picard.cmdline.PicardCommandLine >> --8<---------------cut here---------------end--------------->8--- >> >> Note that the Class-Path property is broken into two lines. This means >> that the reference scanner will miss it and grafting will fail. >> >> > Could we modify the reference scanner to find these lines? > Would there be any serious drawback to that?
The reference scanner needs to be fast. The scheme version is heavily optimized to do quickly replace references for the purpose of grafting. The C++ version will eventually be replaced with the Scheme version as the daemon gets replaced. An alternative to recording full references in the manifest file is to install a “lib” directory that contains symlinks to dependencies. The manifest can then contain relative paths to these symlinks. That’s what I did in the package definition for dropseq-tools. -- Ricardo GPG: BCA6 89B6 3655 3801 C3C6 2150 197A 5888 235F ACAC https://elephly.net
