On Thursday, June 28, 2012 04:13:26 PM Matt Newville wrote:
I think a key issue will be how well Bruce (and I?) can make the switch so that Demeter is using Larch...
I have mentioned this to Matt privately, but I will say so here as well. Matt first began talking about Ifeffit2 around the same time that I started working on Demeter. One of my original design goals was to make it easy (well, perhaps I should say "tractable") to make the switch from Ifeffit to Ifeffit2. I think I have done that. The interaction between Demeter and Ifeffit is completely abstracted using a template system and funneled through a single choke point in the code. Once I understand Larch's syntax and create the appropriate set of templates, switching from Ifeffit to Larch should be as simple as flipping a switch. That's the theory, anyway. I estimate that it will take me about a month to write and test a complete set of Larch templates. Then we will have to figure out packaging issues on the various systems. It seems within the realm of possibility that the adventurous could begin testing Demeter+Larch by the end of the year. But please don't hold me to that! That said, no one will be happier than me to have Ifeffit's memory limitations go away! B PS: Even the non-adventurous should be using the current version of Demeter+Ifeffit, as I told Mr. Yadav yesterday. -- Bruce Ravel ------------------------------------ bravel@bnl.gov National Institute of Standards and Technology Synchrotron Methods Group at NSLS --- Beamlines U7A, X24A, X23A2 Building 535A Upton NY, 11973 Homepage: http://xafs.org/BruceRavel Software: https://github.com/bruceravel