On Sun, 18 Mar 2018 14:05:23 +0100
Julien Lepiller <julien@lepiller.eu> wrote:

> * gnu/packages/web.scm (java-eclipse-jetty-security-9.2)[arguments]:
> Ignore one failing test.
> ---
>  gnu/packages/web.scm | 4 ++++
>  1 file changed, 4 insertions(+)
> 
> diff --git a/gnu/packages/web.scm b/gnu/packages/web.scm
> index ceb7d9b7c..ed4fef740 100644
> --- a/gnu/packages/web.scm
> +++ b/gnu/packages/web.scm
> @@ -6104,6 +6104,10 @@ infrastructure")))
>      (inherit java-eclipse-jetty-security)
>      (version (package-version java-eclipse-jetty-util-9.2))
>      (source (package-source java-eclipse-jetty-util-9.2))
> +    (arguments
> +     `(#:test-exclude
> +       (list "**/ConstraintTest.*")
> +       ,@(package-arguments java-eclipse-jetty-security)))
>      (inputs
>       `(("util" ,java-eclipse-jetty-util-9.2)
>         ("http" ,java-eclipse-jetty-http-9.2)


Can you fix the test or write a comment on why the test fails?

I git-bisected this one and found out:

f2785bd657c55cd36f436b1f6ee1af5d72683162 is the first bad commit
commit f2785bd657c55cd36f436b1f6ee1af5d72683162
Author: Ricardo Wurmus <rekado@elephly.net>
Date:   Sat Mar 10 18:35:31 2018 +0100

    gnu: icedtea-8: Build "out" reproducibly.
    
    Partially fixes <https://bugs.gnu.org/30730>.
    
    * gnu/packages/java.scm (icedtea-8)[arguments]: Add phases
    "patch-keystore" and "strip-jar-timestamps".
    [source]: Also patch DIST_ID in "configure" script.

I'm too tired to look into this now.

Maybe Ricardo has an idea on why this one fails?

Björn