On Mon, 2013-08-12 at 12:50 +0100, Peter Cock wrote:
> Hi all,
> 
> I'm currently trying out installation of tools from the Tool Shed
> on a secondary Galaxy server (not our production server nor
> my development setup). This is running galaxy-dist, current
> latest stable I believe:
> 
> $ sudo -u galaxy hg heads
> changeset:   10201:ebe87051fadf
> tag:         tip
> parent:      10199:8bf64d933704
> user:        Dannon Baker <dannonba...@me.com>
> date:        Tue Jul 02 10:48:31 2013 -0400
> summary:     Fix two more downgrade invocations to accept the
> migrate_engine parameter
> 
> changeset:   10200:fd4113962c32
> branch:      stable
> parent:      10198:0a06df7da177
> user:        Dannon Baker <dannonba...@me.com>
> date:        Tue Jul 02 10:48:31 2013 -0400
> summary:     Fix two more downgrade invocations to accept the
> migrate_engine parameter
> 
> I'm having some problems, e.g.
> 
> Tool shed repository 'package_numpy_1_7' tool dependencies
> 
> Name  Version Type    Status  Error
> atlas         3.10.1  package         Error   Error installing tool 
> dependency atlas
> version 3.10.1: [Errno 2] No such file or directory:
> u'/opt/galaxy-dist-shed-tools/atlas/3.10.1/iuc/package_atlas_3_10/597c0e54eded/env.sh'
> lapack        3.4.2   package         Error   Error installing tool 
> dependency lapack
> version 3.4.2: [Errno 2] No such file or directory:
> u'/opt/galaxy-dist-shed-tools/lapack/3.4.2/iuc/package_lapack_3_4/98a7d116c15f/env.sh'
> numpy         1.7.1   package         Error   Error installing tool 
> dependency numpy
> version 1.7.1: Unsupported action type 'template_command' in tool
> dependency definition.
> 
> Is this due to to galaxy-dist being too old for Tool Shed tools
> whose XML installation scripts use action template_command?

Yes.

> e.g. http://toolshed.g2.bx.psu.edu/view/iuc/package_numpy_1_7/c75482be1d3a
> 
> One way to catch this kind of error would be if the nightly Tool
> Shed tests included attempting to install package repositories
> (e.g. the new "Tool dependency definition") explicitly.

Oh that would be so nice! +1 for me.
And by the way, notifications about new orphan tool dependencies also do
not happen. I think that would be very useful for a few developers that
are waiting for numpy, blast+ and so on.

> (I think I've suggested this before - but it is worth reiterating)
> 
> Thanks,
> 
> Peter


___________________________________________________________
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:
  http://lists.bx.psu.edu/

To search Galaxy mailing lists use the unified search at:
  http://galaxyproject.org/search/mailinglists/

Reply via email to