Either the date
of the 1994 outburst is wrong, or the 2005
outburst proves every model to be wrong!
If we use the date of the 2005 outburst to fix the calculations, instead of that of 1994, we find that the 1994 outburst would have happened in early July and have been finished by the time OJ287 reappeared in the morning sky.