Hi George, > > We made the stategic mistake of first moving the code base to GWASS. > > Unfortunately the differences to QMAC continued to cost time and > > bugs. Still today it is not possible to _exactly_ reproduce the QL > > binary of Minerva, although the code seems to work. So if Minerva > > development should continue, I think this is what would be needed: > > I'm sorry that GWASS causes trouble. Why is that?
It can not be said that GWASS causes trouble, it is just not QMAC, and we can not expect it to behave exactly the same. ANY change of toolchain on long and demanding assembler code would not have been easy. GWASS did not produce the same code as QMAC from Minerva source, and required various sourcecode changes, which now no longer allow to use QMAC. We did not hunt down all the subtle differences yet. I'm sure this could all be solved by more debugging and discussing the issues with you, and you would be more than willing to help. But this process would require a continuos period of time to concentrate on the topic, which neither Richard nor me found. All the best Peter _______________________________________________ QL-Users Mailing List
