I'm curious to see what other's experience has been. My experience with Galaxie has been very negative. Not ready for prime time, and a disaster in a production environment.
There are good ideas in the software - the system for user-defined variables is great, and something they justly brag about when they're selling the software. It has some other nice features as well. I personally like the approach to summary reports, although I think Empower may be easier to use in this respect. The system for defining variables in your method is a place where Galaxie does truly shine, though. They basically have exposed much of the available data as a useful API with logic easily available to make your own custom columns for your reports. Although as with everything in Galaxie, there are little awkwardnesses everywhere to the way you use and share your variables that takes away from the pleasure somewhat.
But it has some other features that show an astounding lack of thought. For example, on every line in your sequence, you have to specify a RunID (Suffix) that gets used to build filenames for the results. When you set up your method, you can say whether you want to use this field or not. The field defaults to zero. whether you use the field or not in your method, if you fail to set some value other than zero (easy to forget), Galaxie's default behaviour is to halt, without emmiting an obtrusive error, and pumpt the system dry. Obviously this is lovely on overnight runs.
That is to say nothing of the tendency of Galaxie to freeze in the middle of a run for no apparent reason... we are a couple rounds of "fixes" later on this and so far so good, knock on wood... but now there's a new bug where it sometimes fixates on the last method you had loaded and if it was a shorter method, it won't collect data for the last part of the run on the longer method.
With Galaxie you do an awful lot of powercycling the computer and all the modules to get yourself back up and running when it has a snit.
One other thing I don't care for, although it's maybe just me... Galaxie doesn't use the degree of 100% database dependence that I would like... a lot of stuff like methods and chromatograms are just sitting on the disk in plain files. Overall, I would say it is not as convenient for organizing data as I would like.
They tried to make a superior user interface - everything you use on one screen for simplicity. This basically works, but I find especially when you have a lot of files open it's awkward and uncomfortable to try and figure out exactly what you're looking at.
Edit: One other thing: The software very much wants you to define either one or two analog channels and use those for all data collection. I think this is probably a fossil of Star, Galaxie's predecessor. The channels cannot be named in such a way as to identify on a report what wavelength was extracted. So you just have to know what channel is what wavelength, awkward for sharing reports with other groups. We were told this is the only way to extract chromatograms in Galaxie by the Varian reps, and that there was no fix. In reality, Galaxy has a very sensible system that allows you to extract an arbitrary number of chromatograms from the absorbance channel, a very convenient and easy-to-use feature favourably comparable to the method of extracting channels in the Empower Method Set (I keep coming back to comparing to Empower, which I feel is what Galaxie has the most pretensions of competing with and being comparable to). The extracted channels are conveniently labelled in default report formats and it's easy to add a $METHODCHANNEL field to your report.
My feeling was that you should ignore the analog channels entirely and just extract from the absorbance channel - it's more modern, powerful and elegant, plus convenient. The big downsides to relying heavily on extracted chromatograms from the absorbance channel is that it seems like Galaxie was designed in this area with the main thought in mind of being Star II, The Revenge Of Star, and the absorbance channel begins to have a suspiciously aferthoughtish feel. For example, there is a polyview-type view that lets you see the chromatogram as it is acquiring but it can be awkward to view and manipulate the chromatogram. There is a nice live view that is much easier to work with but it seems you can only use it is you have checked off to collect data from an analog channel. Also, it seems tentatively that some of our problems with bugs causing runs to stop for no reason have to do with using extracted chromatograms from the absorbance channel, although I don't know if I believe that.
My feeling on it is that it has the makings of good software, but they really should have beta tested it before releasing it. I really wish I never heard of Galaxie, and that's a shame because without all the little annoyances and major problems that come from bad programming and not enough testing, it could be a good experience. But given a choice between say Empower and Galaxie I will choose Empower every time.
Stephen