[X] My Assistant

[djellison]

Emily's asked me to comment on PDS usability etc as there is something of a review going on as to how it might be improved. We're just the informed-amateur's, but any input might be useful. If anyone has anything specific, then scribble it down here and we'll see what sort of consensus we can form.

Doug

[djellison]

- compressed form (e.g. ZIP or GZIP).

That was something I mentioned specifically to Mark ( Mark Rose from AMES who's doing this ) - and that the small-bodies node has it down well, with all the DI data being zipped day-by-day, and the stardust encounter data being zipped up etc. For example, a single 9 meg DI ITS image zips up to approx 1/3rd of that.

I suppose I shoudl chip in with my suggestions which I emailed in before starting this thread...

"Deep Impact data at the Small Bodies node was great, as data is available, by instrument, on a day by day basis, with each day zipped (tar'd actually) up to download as one file, the same is true of all the sets at the SB node and it saves a lot of trouble, time, and hassle - but I think that everywhere that has a ZIP available, should also have that available as a folder to browse for specific content.

A classic example of this is the Geosciences Node for MER data. The mobility reports are available on a sol-by-sol basis as a single ZIP file, however those interested in mapping only want one file from each of those, a single png file that is about 5kb. Fortunately, Jennifer at the MER Notebook is looking into pulling those files out for me to save downloading >5GB per rover, for approx 500kb of data. But it does demonstrate that the combined availability of a ZIP of appropriate archives whilst very useful, should also be accompanied by a browseable version of its contents. The two together offer the best of both worlds. Those who want the lot can grab the ZIP's, those who want something specific can dive in and grab just the 2mb they want, instead of 1872Mb they don’t.

I am not sure of the structure for Cassini releases, but something which the MER releases might benefit from is splitting them into a release-by-release structure (i.e. one folder for each data set, for each 3 month release). Were the contents of these 'zipped' also, then when a new release arrives, people could simply download that, and add it to their own archive.

The difference between these areas is something of a problem however - node-to-node there are huge differences in structure, delivery format etc etc. Standardising across all nodes would be of great benefit, with all data sets available to browse online, grab appropriately 'sliced' ZIP's, and navigate via FTP.

I think there would be some excellent value in having a compiled distribution of WGET for all suitable platforms, with verbose instructions and example batch files to grab select file types and data sets, in bulk, for download. Perhaps an online tool to generate WGET command lines given spacecraft, instrument, data type, date range etc? This would be a huge asset.

I think an emphasis on attaining and maintaining a local copy of specific data sets would be valuable, and an ability to be more specific in this, i.e. only one type of data product per folder 'above' the date, orbit or sol number. For example, Navcam release have, in each sol, many different data products ( terrain wedges, IDD reach ability and so on and so forth) - WGET has solved this problem for me in an instant by specifying specific codes within files ( i.e. *RAD* etc ), but the ability to navigate through.

If possible, any directory navigable via a browser should be navigable via FTP also, as whilst WGET and other downloading programs are useful, FTP probably remains the easiest, most reliable way in which to grab large quantities of data within a directory structure"

Doug