The LAP Tree is a repository of tools, i.e. ready-to-run installations of processing components used by LAP. Key design desiderata are replicability (versioning) and relocatability (in terms of filesystem locations), such that central installation on compute nodes is completely avoided. Instead, the LAP Tree is realized as a version-controlled SVN repository, where (in principle) individual components can be checked out by an arbitrary user and into an abritrary location, for immediate use through Galaxy. The general philosophy is to include all ‘non-standard’ binaries and libraries in the LAP Tree, including command interpreters (like Perl, Python, and Java) that transcend the ‘basic’ operating system.

At present, all users use the trunk of the LAP tree (which is checked out to /projects/lap/development/tree/trunk/) and updated manually (on an as-need basis) by LAP developers. The mid-term goal is to enable users to request individual revisions of components directly in Galaxy, and then dynamically populate a user-specific instance of (relevant parts of) the LAP Tree. To approach this vision (sometime in mid- to late-2015, hopefully), we will need to be more careful about making explicit the provenance of tool versions and inter-tool dependencies, such that we can statically ‘reason’ over requested versions in a given workflow.

To maintain an inventory of tools, versions, and licenses integrated in LAP, the list below needs to be extended with full and up-to-date information (mostly copying from the build directories).

• Java (7u65); OBCL: validated

• Perl (5.20.1); GPL 1.0: validated

• Python (2.7.6); PSFL 2.0: validated

• Ruby (2.2.2); BSDL:

• langid.py

• NLTK Punkt: validated (twice)
• NLTK Tokenizer: wrapped (not validated)
• Tokenizer Segmenter : validated
• REPP Tokenizer (trunk): validated

• NLTK Tagger: validated

• HunPos (1.0): validated

• HFST: validated

• VislCG: validated

• Giellatekno:

• Oslo-Bergen Tagger;

• MaltParser (1.7.2): validated

• Bohnet & Nivre (2012):

• VislCG: validated

• RESA Alignment: to be validated
• Corpus Import: to be revisited
• CoNLL Export: validated
• RDF Export:
• CG3 Export: