*********
Draft
*********

We have identified two critical path criteria and two niggling criteria:

Critical Path
	1. Analyses agreement--Do the analyses on www12 return the same result as those on production? We will check this for all four workflows using the following methods:
		1.1 Performance-upload a file created using a pulser and check the results. Check two other datafiles as well--not from a pulser but real cosmics.
		1.2 Flux-- use the same file to check results. Check two other datafiles as well--not from a pulser but real cosmics.
		1.3 Lifetime--
			Analyze up to five data files on 12 and production (same files).
			Go to the scratchdir for each compute and diff the file lifetimeOut
			Rinse, rather, repeat four times (different data inputs)
		1.4 Shower--
			Analyze up to five data files on 12 and production (same files).
			Go to the scratchdir for each compute and diff the file eventCandidates
			Lather, rinse, repeat four times (different data inputs)

	2. Administrivia--Do all of the site features work?
		2.1 data file searches
		2.2 poster searches
		2.3 graphic uploads
		2.4 save plots
		2.5 deleted data files
		2.6 re-run analysis
		2.7 logbook
		2.8 site look and feel

Niggling
	3. Performance--of the server and site itself--not the detector
		3.1 Update the message to users when analyses are running. Something like: "You can close this window and pick up your results later."
		3.2 Does anything odd happen during testing criteria 1.1-1.4?
		3.3 Do the SWIFT and VDS versions on 12 both work reliably?

	4. Enhancements--are there any documented enhancements that we should attempt and wait for.
		4.1	Find these in Bugzilla.