In this column we keep you informed about bugs that were reported to us by users and how they have been corrected since. Note that these corrections are incorporated into our current version of EMME/2 and will automatically be included into the next software release that is sent to your organization.
Module 1.21: Two problems were corrected. Firstly, the module used to hang up in an infinite loop under certain conditions - this should no longer happen. The second problem was related with the user criteria - it is now no longer necessary to prepend dots (eg "..HS") when specifying user initials, nor will the dots be visible on the report.
Module 1.22: A bug was fixed in module 1.22 that prevented initial transit boardings and final transit alightings at node to be copied along with the remaining information of the scenario. This bug would lead to erroneous output in modules 6.21 and 6.22 concerning initial boardings / final alightings in scenarios that were copied after the transit assignment had been performed.
Modules 2.21 and 2.22: A bug was reported and corrected that destroyed the very last transit segment in a scenario, when - within the same module call - more segments were deleted than added. A different bug was also detected and fixed, which dephased the entire segment list when the first n lines in the transit line table (and only these) were deleted.
Module 2.31: When generating a report or a punch of intersections, the module now outputs correctly those nodes that are included in the intersection definitions, but have no penalized turns defined.
Module 3.13: In module 3.13, the graphic command "N" now works as expected. Only the node and centroid boxes that are interesting are plotted. A problem with the hatch width module parameter has been solved. The hatch width parameter is now also applicable to vertical centroid bars.
Module 3.21: Calculations where column-wise matrices (i.e. full matrices stored column by column instead of row by row, they are identified by the "/c" suffix) are considered lead to incorrect results. The best way to circumvent this problem is to copy the column-wise matrices into row-wise matrices and use these new matrices in the calculations. This bug should only affect very few users, since the column-wise matrices are a "well-hidden" feature of EMME/2 and are used only very rarely.
Module 5.21: "R" and "Q" on the "Select: List device" work now properly and give reasonable messages on report prompts. It is now possible to send an interrupt signal when the auto assignment is in process. This will terminate the assignment after the next iteration. After a quit or interrupt, the auto assignment can be resumed by simply calling 5.21 again. Otherwise, entering 5.21 again will result in a "not ready" message. The active stopping criterion is now listed at the end of the assignment.