TODO

Polishing and Cleanup of Psi4 Prior to Release

Development Team

• Fix failing test cases (or remove if they cannot be fixed)
  • 120 (fnocc1, intermittent failures)
  • 121 (fnocc2, intermittent failures)
  • lingering fnocc issues
    • CDS will investigate why long int works and ULI don't (by which fnocc2 test case was closed)
    • Brandon will produce failing example input
  • 122 (Ryan)
  • 123
  • 124
  • 125
  • 126
  • 127
  • 128
  • 129 (frac, reduce number of points along curve)
  • 130 (cc54, see if it fails again)
  • 132
  • 133
  • 138 (mrcc-optfreq, Brandon)
  • 168 (pywrap-freq-e-sowreap, remove from make tests) (Lori)
  • 190 (various dfomp tests)
  • EFP energy test cases
• Deprecate initializing modules from Process::Environment::Wavefunction. Pass wavefunction objects into modules instead (Daniel Smith)
• Remove transqt and transqt2
• Remove unused/attic code
• Remove libchkpt
• As many of the 1.0 Milestones as possible (e.g., PCM-UHF fails to converge, Finalize ChemPS2, etc.)
  • make sure (dirty) isn't in the outfile header along with the git info. It's not supposed to be there when the source clone is clean.
  • make sure Additional authors gets additionally populated in code and also website
  • update copyright anywhere to 2016
  • new set of output.ref
• Delete/update obsolete build information (INSTALL.rst, INSTALL_CMAKE). See email chain CDS 8/13/15 for more related to-do
• Incorporate upgraded CI code and new DF-MCSCF code (Daniel Smith)
• Update capabilities list and associated driver wrapper functions
  • Ugur's DF/CD-MP3 code
  • UHF DF-MP2 117
  • review energy argument names for occ/dfocc methods
• Check with Ugur about any new features he's added
• Review and prioritize existing Git Issues
• Create release notes
• Tickets marked confirm-or-deny need user (as opposed to developer) experience
• R/U/RO HF/KS settle discussion in 103 also touched upon in priv101 about going to reference R, RO, U, CU, or auto that chooses according to mol and mtd. also removing energy/grad/freq arguments rhf, uhf, rohf. Agreement summarized in 117 AJ 118
• Record SAPT terms as PSI variables (Jerome)
• Record SAPT total interaction energies as PSI variables for methods introduced in Trent's paper (sSAPT, deltaMP2, etc.) (Lori) 37
• Get Psi to print the name of the aux basis it is using (Lori)
• Stop or warn if using a nonsensical fitting basis set (e.g., fallback QZ fitting basis not ok if primary set is 6Z) (Lori)
• Fix PK algorithm for conventional SCF (Jerome)
• Split quicktests into quick and medium tests for faster quick tests (Sam, York)

Documentation and Usability Team

• Fix examples in user manual that do not actually run (David, Lori) 149
• Document how to create a user-defined basis set (Lori) 43
• Decide what version number we're working toward. Notes and in particular, first reference therein Resolution: use versioning as described in Notes
• Review what docs go where. build-docs on wiki? class-docs in doxygen? all else in sphinx?
• Review binary installation instructions at here (Dom)
• Review cmake build instructions at here (Dom)
• Look through entire users' manual for any obsolete / incorrect / incomplete information 109 116 - [x] Review first half of manual for clarity (Mutasem) - [x] Review second half of manual (theoretical methods and later) for clarity (Jeff) - [x] Review all test cases in manual for accuracy (Trent) 149 - [x] Ensure short introductions to theory and associated references are in manual (Wallace) - [ ] Review Programmers' Manual part of manual - [x] update post-wfn-pass
• Complete updates to psi4public wiki regarding any user or compilation information
• Provide input for Website team regarding organization and content of user forums
• Finalize CheMPS2 integration 150 (get working in binary, and add Interfaces page to documentation)
• Add documentation and test case for frozen cartesian optimizations (Rollin)
• Add QM/MM documentation (Andy)
• Make sure X2C is documented (Francesco, Prakash, or Wallace - target for mid-Dec)

Website Team

• Overall review of website; make it up-to-date and user-friendly
• Review documentation auto-build system Resolution: now documented at https://github.com/psi4/psi4meta
• Review binary auto-build and distribution system
• Review github feed -> website automatic update system Resolution: now documented at https://github.com/psi4/psi4meta
• Review PsiNet role and tasks
• Review / decommision obsolete websites that may confuse people Googling psi4 (we recently had a query about MPI in Psi4 based on sirius pages) 112
• Set up user forums
• Add user forum link to navigation panel in upper right of psicode.org
• Develop announcements of the release - [x] Create YouTube Channel and link from psicode.net - [x] Create video introducing Psi4 to users - [x] Create video introducing Psi4 to developers - [x] Create tutorial videos

As time allows

• Make PCMsolver an externally-detectable module (Roberto)
• Review coverage of test cases. 108
• Python 3 compatibility priv86, 110, 111 (Daniel Smith)
• Make CI code work same as CC code in terms of property analysis user input (Daniel Smith)
• Minor tweaks to usability (parser improvements, etc.) 114
• Complete updates to psi4public wiki regarding developer information
• Test CUHF and update manual based on findings
• Complete libEFP energies 115
• Resolve whether or not there is a bug in ROHF updates
• Review status of stability analysis code; finish for RHF->RHF, UHF->UHF, ROHF->ROHF, and RHF->UHF if possible
• RHF and UHF QCSCF code completed (Daniel Smith)
• Fix bug in ROHF QCSCF code (Hessian seems off)
• Mac binary distribution 49
• Implement one-step DF-MCSCF (Daniel Smith and David)
• Find bug in FNO-DF-CCSD(T) for large cases on i7-5930K 209

19 Apr 2016

Notes for the meeting

• Everyone work from a branch on a fork of psi4/psi4 (https://github.com/psi4/psi4/wiki/1_Obtaining#forkpsi4public)
• If you're working on docs, make sure your sphinx-build --version is >=1.1 and <1.4 . conda install sphinx=1.3
• If your commit should close a ticket 999, include usual commit message fixes psi4/psi4#999
• Shout if you notice anything in docs that looks like it ought to be clickable but isn't clicking
• Note that I'm marking all of these 1.0.0 milestone for sorting convenience, but some are less pressing and more for variety of skill levels.
• If inspired to rework method pages, model table is http://psicode.org/psi4manual/master/occ.html#id33
• 109 Jerome
• 149
• 179 DGAS
• 213 DGAS
• 216 Evangelista
• 249 Jeff
• 349 Lori
• 350 bumped to 1.1
• 351 Lori
• 352 Lori
• 353 Brandon
• 355 Jet bumped to 1.1
• Unticketed
  • Testing gdma(), pcmsolver(), chemps2() Trent
  • Testing sapt psivars (esp. component sums) Lori
  • main capabilities table Lori (still need crawford response)
  • sos-mp2 seems to have shifted energies
  • dftd3 is refusing to run occasionally env=lenv problems, it seems
  • revamp docs for rearranged driver, impaired cbs
  • cbs_wrapper needs merging by afternoon 4 May, so LAB can build binaries for workshop 5 May (not totally necessary. we can tell them this one line changes in future.)
  • test and update psi4webmo interface for 1.0 (maybe Brandon)
• 356 DGAS
• 357 DGAS
• 238 Ryan
• LAB has a bug report on Lithium with SCF to investigate if you'd rather not do docs
• Get executive decision on 279. LAB favors no change.
• Get executive decision on Others section of http://psicode.org/psi4manual/master/basissets_byelement.html . Are some of these too close to widespread abbreviations of common basis sets or abbreviations that molpro allows? Shall we rename?
• CDS: move your green caution block http://psicode.org/psi4manual/master/tutorial.html#geometry-optimization-and-vibrational-frequency-analysis to thermo.rst?
• DGAS: check hessian example [2], gradient example [1] for numpy correctness
• TDC: does test case cc5 really need 12gb? max memory of test suite otherwise is 1gb
• JG: are there any more SAPT-ct versions needed? => No, we have all of them already.