Editing G09 (section)

== Revision Issues ==

* '''Revision A1 and A2'''; 
** do NOT perform the Dispersion calculation properly. You need the revision B1 or onwards.
** implementation of the QZVP basis set is incorrect, it has not the polarization functions. You need the revision B1 or onwards.

* '''Revision C1'''; 
** problems with the modredundant section. To freeze a distance you can not use "B atom1 atom2 value F", as this does not do anything. Instead, you will have to move the geometry to the value and use "B atom1 atom2 F".
** for some geometries the calculation stops after reading the initial structure.

* '''Revision D1'''; 
** several issues from previous versions have been sorted, and some new interesting features added, such as DFT-D3 for most functionals, ...
** It still does not behave properly with the modredundant section: you can not use "B atom1 atom2 value F" (see above)
** A severe error concerning minima optimizations and frozen geometries of specific systems has been observed. The error consist on a call as if the chk file was incorrect/corrupted "Operation on file out of range...dumping /fiocom/..a large list of numbers...Error termination in NtrErr: NtrErr Called from FileIO." We don't know yet exactly what is special in the systems that crash, although we suspect that it might have something to do with having to connected aromatic/planar rings. The error does not occur in TS searches. There are several solutions to overcome this depending on the project needs: i) to use the version g09.a02 if the the method of choice is available on that version. ii) to use a guess=always, although it is expensive. iii) the use of opt=cartesian also helps out to diminish the probability to get the error although it is  does not always work and it is also expensive.