[GSAS-II] GSAS-II news

[Toby, Brian H.]

Summary of GSAS-II news items: 
	Software server fixed
	Linux 3D graphics fixed
	Multicore processing
	Scripting interface

* There has been a lot of e-mail traffic over the past week as there were problems with people outside ANL accessing our web server to install or update GSAS-II. This was hard to figure out, because some commands worked and others did not and of course everything worked for us (from inside the firewall). Fixing this required at least two of our best IT folks for most of yesterday. Sorry to everyone inconvenienced. If you are still having problems, please contact me.

* Another bit of news is that yesterday we also fixed a longstanding bug in Linux openGL graphics that messed up plotting structures.

* We have also implemented multicore processing for powder refinements, but at least for now, this is turned off by default (and today Bob found a bug with multicore refinements with constraints that I have not yet tracked down). If you want to try this out, set configuration variable Multiprocessing_cores to the number of simultaneous computations to be done. Speedup is greatest for refinements with lots of reflections and one gets perhaps half of the theoretical benefit of the multiple cores (two cores runs at 75% of the single core time rather than 50% but with enough cores we have seen times drop to ~35%). Note that the optimal number of cores to use seems to be the real number of cores, not the bogus hyper-threading number that Intel claims; Setting Multiprocessing_cores to -1 will pick what we think is a good default. See https://www1.aps.anl.gov/files/download/CXS/Meetings-Workshops/GSAS-II/2017Sept/G2mtg20170911.pdf for more information.

* As you have likely noticed, GSAS-II has a new appearance (where the tree and data windows have been combined) with what we call the “2 frame” version. Bob has been busy updating tutorials to match this and has been fixing bugs and cleaning up behavior. 

* University of Chicago undergrad student Jackson O’Donnell spent the summer working on adding a scripting capability to GSAS-II. There is more to be added to this to make it fully featured, but this is an excellent start. If you are interested in using GSAS-II from shell or Python scripts, please contact me and I will point you towards documentation. 

Other stuff: we have improved checking of constraints to prevent most errors (if you do manage to generate a constraint error, please tell me how) and we have broadened use of sequential refinements where refined variables change across histograms. For this and other news (as presented to our local user community) see: 
https://www1.aps.anl.gov/Science/Scientific-Software/Meetings-Workshops/gsas-II. Their development requests are being added here: https://subversion.xray.aps.anl.gov/trac/pyGSAS/roadmap. We have lots to do. 

Brian