Hi Xypron, > please, find in the appendix a description of the functions that I > implemented, which cover most of the whishis mailed in this thread.
Thanks. Two things I would like to avoid are day-light saving time and UTC/local time. That is, the idea is to interpret the calendar time just as the number of seconds elapsed since 1970-01-01 00:00:00 not defining whether it is UTC or local time. In fact, it is always possible to define the timezone offset in the model to translate between UTC and local time, if necessary. > A special function tmfunc is not needed, it is just part of the > conversion of time to string. > Thats exactly what in my implementation functions gmtime() and > localtime() offer. > param year := gmtime(t, "%d"); ... Unclear. As I understand gmtime returns a symbolic value (character string) suitable for printing while tmfunc returns a numeric value suitable for arithmetic calculations. Probably it would be reasonable to add two functions atoi and atod to convert symbolic values to numeric ones. _______________________________________________ Help-glpk mailing list [email protected] http://lists.gnu.org/mailman/listinfo/help-glpk
