Starting with tag: [TAG meep-1.1.1 stevenj@alum.mit.edu**20090824195548] [rewrite of polarization code, moved everything into neew susceptibility class; C++ interface has changed; main user-visible change in libctl code is removal of store-polarization-energy feature; libctl interface names have changed, but old names are retained for backwards compatibility csliang@mit.edu**20090821185915] [reconciled with Meep 1.1 csliang@mit.edu**20090822050319] [making timing stack bigger so that we can avoid hacks in step.cpp stevenj@alum.mit.edu**20090822162335] [remove unused var stevenj@alum.mit.edu**20090822171526] [first pass at needs_W_notowned support for update_pols (not tested yet since no anisotropic susceptibility, but make check on old stuff passes) stevenj@alum.mit.edu**20090822173522] [make sure to recompute boundaries when W allocated if notowned W needed stevenj@alum.mit.edu**20090823051050] [added mechanism for polarization to have internal data that needs to be communicated at chunk boundaries (not used yet, but will be needed for multi-level atom) stevenj@alum.mit.edu**20090823173124] [rename poldata -> polarization_state (more descriptive) stevenj@alum.mit.edu**20090823173616] [added simpler way to get Drude model stevenj@alum.mit.edu**20090823175723] [tentative support for anisotropic Lorentzian susceptibility (untested, but at least it doesn't break the isotropic case) stevenj@alum.mit.edu**20090823194710] [catch user error: specification of Lorentzian with too high a frequency is inherently unstable stevenj@alum.mit.edu**20090824022522] [check whether dispersive material couples TE/TM stevenj@alum.mit.edu**20090825153650] [add test case for new exp(i beta z) feature stevenj@alum.mit.edu**20090826031706] [merged new beta test stevenj@alum.mit.edu**20090826033516] [use boolean for stress_dct_integral argument (thanks Alexander) stevenj@alum.mit.edu**20090830202907] [tentative 1.1.2 version bump stevenj@alum.mit.edu**20090830210218] [added Lorentzians+whitenoise materials for Alex thermal stuff stevenj@alum.mit.edu**20090902223023] [properly synchronize trivial_sigma array across all chunks as intended; thanks to Alejandro for the bug report stevenj@alum.mit.edu**20090904220039] [rename omega -> frequency in susceptibility for libctl (maintaining backwards compatibility), to avoid confusion about 2pi factors stevenj@alum.mit.edu**20090911005131] [deallocate any trivial offdiagonal susceptibilities stevenj@alum.mit.edu**20090929191735] [with special-kz, we can use real fields by default even for kz nonzero stevenj@alum.mit.edu**20090929194022] [fix incompatibility of Casimir with guile 1.6, thanks Peter stevenj@alum.mit.edu**20090924170506] [add MANIFEST file stevenj@alum.mit.edu**20091003003201] [fix MANIFEST for susceptibility class stevenj@alum.mit.edu**20091003003415] [remove dup DOCMP loop, thanks to Song Chua for the bug report stevenj@alum.mit.edu**20091012061323] [added option eig-match-freq? in eigenmode source to use specified frequency via Newton, fixed wraparound bug in eigenmode source field pattern stevenj@alum.mit.edu**20091009173925] [remove obsolete TODO, thanks to Marcin Stolarek stevenj@alum.mit.edu**20091013122353] [use libctl_quiet feature (new) to suppress duplicate ooutputs from main.cpp in parallel execution, thanks to Hila Hashemi for the bug report stevenj@alum.mit.edu**20091013195325] [resolve merge conflict (had already removed obsolete comment in this version) stevenj@alum.mit.edu**20091014211646] [more stable 2-norm in bicgstab (avoid overflow) stevenj@alum.mit.edu**20091126013431] [end_global_commun. should reset mycomm_save to prevent double-free in end_divide_parallel stevenj@alum.mit.edu**20091209180734] [initial stab at storing P in susceptibility class (known_results segfaults) stevenj@alum.mit.edu**20100621194709] [bug fix in new susceptibility code: don't subtract_P until P-data is allocated ardavan@mit.edu**20100621205824] [include a Lorentzian polarizability in one of the tests stevenj@alum.mit.edu**20100621235447] [more generel internal susceptibility data structure stevenj@alum.mit.edu**20100622010706] [multilevel atoms will need dt to initialize internal data stevenj@alum.mit.edu**20100622012127] [added cent2yee_offsets stevenj@alum.mit.edu**20100622221331] [incorporate multilevel-atom.cpp into repository and update meep.hpp ardavan@mit.edu**20100623030252] [got multilevel stuff compiling (untested) ardavan@mit.edu**20100623195456] [boilerplate constructors/destructors for multilevel_susceptibility ardavan@mit.edu**20100623200219] [support transition-specific sigma factor in multilevel atoms stevenj@alum.mit.edu**20100805231008] [correct usage of Gamma, GammaInv in multilevel code stevenj@alum.mit.edu**20100809194751] [allow vec.hpp macros to be used outside of meep:: namespace stevenj@alum.mit.edu**20100809202515] [multilevel-atom Scheme front end (compiles, untested) stevenj@alum.mit.edu**20100809210009] [various bugfixes in multilevel code (runs without crashing now) stevenj@alum.mit.edu**20100811033008] [when partitioning processes, make sure each process groiup gets its own random seed. stevenj@alum.mit.edu**20110209224619] [am_really_master can now use my_global_rank stevenj@alum.mit.edu**20110211191556] [version bump stevenj@alum.mit.edu**20110216180020] [bug fix in polarizabilities for parallel case (thanks to Ardavan for the bug report) -- when setting up the communications in boundaries.cpp, we also call the internal_notowned_ptr routines for chunks from other processors where the polarization data is NULL, in which case we should just return NULL stevenj@alum.mit.edu**20110219041334] [bugfix in norm2 function of cw solver (caused failure in MPI, among other things) stevenj@alum.mit.edu**20110219043319] [more convenient noise_amp def, that includes omega_0 factor stevenj@alum.mit.edu**20110223212704] [comment about using gaussian random #'s alexrod7@mit.edu**20110309202408] [bug fix in structure copy constructor - miscopied siginv alexrod7@mit.edu**20110309203125] [first stab at supporting real coordinate stretching in PML alexrod7@mit.edu**20110309205845] [use_pml is only called internally, does not need default arg alexrod7@mit.edu**20110309210158] [more rational argument order in use_pml stevenj@alum.mit.edu**20110309211000] [cleaned up mean_stretch API, added optional Rasymptotic and mean_stretch args to pml(...) constructors stevenj@alum.mit.edu**20110309212550] [use non-unit stretch factor in PML tests stevenj@alum.mit.edu**20110309213121] [it doesn't make sense to ever use a mean stretch < 1, and could cause problems if kappa goes through 0 stevenj@alum.mit.edu**20110309221358] [subtlety: real stretching (kappa) must scale more smoothly than absorption stretching (for fixed R-asymptotic) in order to have the same length scaling stevenj@alum.mit.edu**20110309224617] [fix tolerances in PML test for stretched case stevenj@alum.mit.edu**20110310042306] [bug fix - transition sigma was not initialized properly in < 3d csliang@mit.edu**20110321164853] [slight optimization csliang@mit.edu**20110321164927] [add transition-time synonym for 1/transition-rate csliang@mit.edu**20110321180045] [bug fix - when libctlgeom cannot find the normal vector, it returns normal == 0, so we must handle this case stevenj@alum.mit.edu**20110401205755] [honor --enable-portable-binary stevenj@alum.mit.edu**20120525193007 Ignore-this: f15e4f7d9e1ce3561dac934ecf75619b ] [initial stab at ldos (power-spectrum) calculation stevenj@alum.mit.edu**20120626194023 Ignore-this: bbbcecada13fc8a48408c3e00580515c ] [whoops stevenj@alum.mit.edu**20120627142621 Ignore-this: 761222b9edb186190062773cedc16361 ] [whoops stevenj@alum.mit.edu**20120628001000 Ignore-this: 3330030288b8bad51ca5f3d399465ea5 ] [sign fix of ldos (not that it matters if we normalize by another calculation) stevenj@alum.mit.edu**20120628001056 Ignore-this: 194426c7e0bfdc4047a3424cd969ed56 ] [fix eigenmode source to use correct principle-of-equivalence fields stevenj@alum.mit.edu**20120629172256 Ignore-this: d7b9b337278016adb3ef47bb5bf9406 ] [allow manual specification of direction (for volume sources) in eigenmode-source stevenj@alum.mit.edu**20120629174506 Ignore-this: ea0d57fee7b7cfda42c1a43ab8ae0423 ] [add eigenmode-source example stevenj@alum.mit.edu**20120629180447 Ignore-this: e486d4c30fbfe7a35dda0ae55984a063 ] [make eig-match-freq the default stevenj@alum.mit.edu**20120629180552 Ignore-this: 185bec03fc1ab5ee1608f8bbab6d022 ] [all-components default is not working yet, need to specify it explicitly stevenj@alum.mit.edu**20120629181321 Ignore-this: 81d7ed36f670f8a6e3239dc7d931a9eb ] [all caps for ALL-COMPONENTS constant stevenj@alum.mit.edu**20120629181528 Ignore-this: 756e51bcd4b5e8ac48c0f81c82b0c165 ] [eig-kpoint is no longer required for match-frequency?=true (uses a reasonable guess) stevenj@alum.mit.edu**20120702170128 Ignore-this: 9d8ec63a4fca651314489a697e3e819b ] [LDOS feature can now optionally output Fdft and Jdft stevenj@alum.mit.edu**20120706185914 Ignore-this: 76ae3ad3b880b22e17dc6a12ad9d8b7c ] [more LDOS fixes (try to get absolute normalization right, not just relative normalization) stevenj@alum.mit.edu**20120707201434 Ignore-this: f0a570950b77b7288fa32a1f4db0f8fd ] [in dft_ldos, when computing Jdft, only FT the real part of the current for real fields (fixme: still not quite right if the spatial amplitude A is complex) stevenj@alum.mit.edu**20120708124357 Ignore-this: 63628f736590db141363dba371ef18d5 ] [print both right- and left- flux in waveguide example stevenj@alum.mit.edu**20120716170016 Ignore-this: d625ebbc5a5bb6ee8f2a7776de3ed39c ] [disable multilevel-atom support in libctl interface, since it is not yet working stevenj@alum.mit.edu**20120718140757 Ignore-this: 8cddf01d9eec05f79fcabb65c236579c ] [updated NEWS for 1.2 stevenj@alum.mit.edu**20120718142557 Ignore-this: 564ac0aa3292f195ff73d93d641ebfd5 ] [add epsilon-input-file feature, similar to MPB stevenj@alum.mit.edu**20120718183911 Ignore-this: 7024d307a18ef740bf7616575c38c35e ] [require libctl >= 3.2 stevenj@alum.mit.edu**20120718183913 Ignore-this: 14668ba8ac4d61694986c8fb9f5627be ] [add test program for anisotropic dispersive materials ardavan@ab-initio.mit.edu**20120718191900 Ignore-this: f0e6a09b8caa5e6af331a807b0b5db11 ] [fix up aniso_disp Makefile stevenj@alum.mit.edu**20120718192713 Ignore-this: d8ec010dba497f4a5b42dcb9aad6a885 ] [fixes for Guile 2.x (given libctl 3.2) stevenj@alum.mit.edu**20120719192546 Ignore-this: 5a3358e4e9f3b9466ef9279b57d2adc ] [fixes to work with guile 1.6 stevenj@alum.mit.edu**20120719222815 Ignore-this: 7277bce0365b14983443af7ca0c2363d ] [stress tensor test ardavan@ab-initio.mit.edu**20120719213123 Ignore-this: d00afbe198d84716d08e465e228f2d3b ] [fixed stress tensor test (volume list takes a component,not a direction) and added to test suite stevenj@alum.mit.edu**20120719230020 Ignore-this: 26aa93a19b46ea6ee040c107ae58f2e3 ] [copyright-year bump, tentative release date stevenj@alum.mit.edu**20120719231751 Ignore-this: 3f867b7995636f934462dfb4b4699fa4 ] [whoops, add stress_tensor to 'make check' tests stevenj@alum.mit.edu**20120720005841 Ignore-this: df15b9f375fd5fd5354a64754338fcf5 ] [fix to work with MEEP_SINGLE (i.e. in single precision) stevenj@alum.mit.edu**20120720014202 Ignore-this: 84d7aba99b2ed2137f7e9ddb5cecbaf4 ] [run 'make check' with at most 4 processes (since Meep aborts if we use more processes than the grid diameter, and 4 processes should be a rigorous enough test by default); aniso_disp can only use 1 process, so do this with divide_parallel stevenj@alum.mit.edu**20120720023937 Ignore-this: 8f55aa52e8460f059b62d05a7715faa0 ] [update tentative release date stevenj@alum.mit.edu**20120720024145 Ignore-this: e4e6f96caa2eaa6074bb56034c46ae83 ] [TAG meep-1.2 stevenj@alum.mit.edu**20120720172301 Ignore-this: 547a9b9eb4c362976662f430f423a3e7 ] [whoops, copyright update stevenj@alum.mit.edu**20120720174448 Ignore-this: 632b5c016e1442cc3303f620c0a44d11 ]