Phase-1 Processing

	Goals:
		sky subtraction
		flat-fielding/calibration
		identifying and masking bad regions
		generate catalogs of objects for phase-2

	- highly parallel - N chips per node

	- typically treat each run separately

	- tips:
		- always keep a DIARY file with full history

		- scripts are verbose (like 'make') so always do:

		./scripts/somescript.pl >& tmp.log &	# put job in background
		tail -f tmp.log				# monitor log output till done
		grep failed tmp.log			# check for errors

		- check results regularly - use examine.pl and
		examinebatch.pl

		- nearly all scripts give 'usage' if invoked with no args

	- getting organised
		- read data to $mscratch/tapedir
		- make symbolic links in $mscratch/srcdir (with renaming)
		- extract headers
		- make inventory (lc-format catalog)
		- make names lists

	- typical DIARY fragment
.....
# load bias frames from master onto slaves
	./scripts/copydatatoslaves.pl srcimages srcimages names/BIAS.lst fits
# do overscan correction, trimming, flipping if necessary
	./scripts/prepareimages.pl names/BIAS.lst
# visually examine all the bias frames
	./scripts/examinebatch.pl names/BIAS.lst srcimages tmp
# make a median of all the biases (create new names list if necessary)
	./scripts/combineimages.pl names/BIAS.lst srcimages flats/bias
# clear off
	./scripts/deletedataonslaves.pl srcimages names/BIAS.lst fits
.....