Today I ran into a new issues with installing a specific version of Freebayes. It turns out to be an issue with how git deals with submodules, something I'm only just learning about myself. In order to ensure that FreeBayes is built with the correct versions of all of the various submodules for the desired revision, it seems that two commands are needed:

git checkout 9608597d12e127c847ae03aa03440ab63992fedf
git submodule update --recursive

In the current version of tool_dependencies.xml, there is a git reset --hard 9608597d12e127c847ae03aa03440ab63992fedf which seems to work similarly to the checkout command above, however, it does not put the submodules back at the correct revisions. That is only now an issue since they were updated only recently.

I've attached a patch to the tool_dependencies.xml file that should once again allow FreeBayes to be built and installed automatically. It would be great if you could update the Tool Shed repo. Thanks.

--
Lance Parsons - Scientific Programmer
134 Carl C. Icahn Laboratory
Lewis-Sigler Institute for Integrative Genomics
Princeton University

diff --git a/tool_dependencies.xml b/tool_dependencies.xml
--- a/tool_dependencies.xml
+++ b/tool_dependencies.xml
@@ -4,7 +4,8 @@
         <install version="1.0">
             <actions>
                 <action type="shell_command">git clone --recursive 
git://github.com/ekg/freebayes.git</action>
-                <action type="shell_command">git reset --hard 
9608597d12e127c847ae03aa03440ab63992fedf</action>
+                <action type="shell_command">git checkout 
9608597d12e127c847ae03aa03440ab63992fedf</action>
+                <action type="shell_command">git submodule update 
--recursive</action>
                 <action type="shell_command">make</action>
                 <action type="move_directory_files">
                     <source_directory>bin</source_directory>
___________________________________________________________
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/

Reply via email to