The TARDIS Data Packager program assists the annotation of data for registration with TARDIS.edu.au.
TARDIS works off a federated storage model. That is to say, data is stored at the owner's local lab or institution and registered on this site for search, download and citation.
Requirements for a TARDIS repository are simple.
Data, once organised using the packaging tool can be hosted on any computer with a web server and optionally on an ftp server with its ports open to the world. The URL to this data can then be registered on this site by anyone signed up.
Generally, if there's somewhere to host data that has a URL accessible by the world, then it's fit for TARDIS. This includes digital repository software such as Fedora or ePrints, which might exist at an institutional library.
Setting a TARDIS repository up from scratch is a relatively straightforward process. Free, easy-to-use web servers that work on any operating system such as Apache HTTP. For full-download capability, an FTP server is required. Similarly, there are easy and free solutions such as Filezilla Server.
Note: Setting up a TARDIS repository on desktop computers at an institution will likely require opening external access in the firewall, so that data is public. Contact your system administrator for help with this.
There are two types of Datasets able to be published, raw and ancillary. Each set of raw diffraction images (img/osc/adsc etc..) should be in its own directory. Ancillary files are for log files, processed files, and anything else to be included with the published experiment.
"images_sample" - a directory containing raw diffraction images
"ancillary" - a directory containing all other types of files to be included
Download the TARDIS Data Packager program and run it. The initial screen gives users the ability to describe the overall experiment (a collection of datasets). Also important is the "Destination for Data" field. Choose an empty directory for the annotated data to be output to.
Click "New Raw Dataset" to locate and annotate a set of diffraction images.
Click "New Ancillary Dataset" to choose the path to extra files, and describe the set.
Click "Package Dataset" to begin the annotation and organisation process. Metadata is automatically extracted from raw diffraction images, files are copied to their destination and an annotation file "METS.XML" is created.
Files are copied along with the annotation file (METS.XML) to the destination path chosen in the experiment section. All of these files should be copied to a publically web-accessible location, so a URL can be created and submitted to TARDIS.
The copying of files to a web server can vary widely and a system administrator should be contacted for assistance. For example purposes, this section outlines how one might copy files via FTP to their web server.
Open an FTP client (Transmit for the Mac is pictured here) and connect to the web server that will store the files.
Copy the entire directory holding the datasets and the METS.XML file across to the server.
Once on a web server, the data should be accessible via URL. To verify this, a good test would be to try and access the METS.XML file in a browser. A successful display of METS.XML in Firefox is shown below.
This URL will provide as the basis for experiment registration. Log in to TARDIS and click "Data" then "New Experiment". Paste the full URL to METS.XML in the "METS XML url" field. A password can be entered to protect the experiment from the public view. Optionally, FTP details can be entered for easy downloading of full datasets.
Experiment ingestion into TARDIS" local cache can take minutes. A successful registration is shown below.
Note: Data needs to be approved by an administrator before attaining a persistent handle or showing in the public experiment index.
TARDIS can 'officially' support more metadata formats on request.