This tools has been passing its nightly tests fine,

However, there now appears to be a silent installation failure
(the install hasn't worked but there was no install error):

Automated tool test results

Tool test results
Automated test environment
Tests that failed
Tool id: nlstradamus
Tool version: nlstradamus
Test: test_tool_000000
Fatal error: Exit code 127 ()
/bin/sh: 1: NLStradamus: not found
Traceback (most recent call last):
line 171, in test_tool
    self.do_it( td, shed_tool_id=shed_tool_id )
line 102, in do_it
    self.verify_dataset_correctness( outfile, hid=elem_hid,
maxseconds=testdef.maxseconds, attributes=attributes,
shed_tool_id=shed_tool_id )
line 782, in verify_dataset_correctness
    self._assert_dataset_state( elem, 'ok' )
line 606, in _assert_dataset_state
    raise AssertionError( errmsg )
AssertionError: Expecting dataset state 'ok', but state is 'error'.
Dataset blurb: error

On the other hand, over on the main Tool Shed, with what should
be the same tool configuration, the install works (although there
is a new sniffer bug on the Galaxy side, see separate email):

Here's the tool_dependencies.xml file,

<?xml version="1.0"?>
    <package name="NLStradamus" version="1.8">
        <install version="1.0">
<!-- Although v1.7 used a subfolder in the tar-ball, v1.8 did not -->
                <action type="shell_command">g++ NLStradamus.cpp -o
NLStradamus -O3</action>
                <action type="set_environment">
                    <environment_variable name="PATH"
This downloads NLStradamus v1.8 from this folder,

The C++ tool is compiled as described in the README_C.txt file, using
g++, and included in the $PATH.

The older slower Perl implementation is not installed.

My hunch is a subtle regression in the <action type="download_by_url">
(note in this case the tarball does not contain a sub-folder).

Please keep all replies on the list by using "reply all"
in your mail client.  To manage your subscriptions to this
and other Galaxy lists, please use the interface at:

To search Galaxy mailing lists use the unified search at:

Reply via email to