Major bug fixes

Things are moving right along. We got a test cartridge and also have some handheld test units arriving.

After further testing on real hardware, I determined that there was three bugs. The first is that CLS would corrupt the screen on real hardware. Printing negatives would do the same corruption. This was all due to me forgetting to turn the screen off and back on in a few places. No big deal.

The other bug was a typo in the negation handling during the postfix evaluation. Also no big deal, but was a pain to track down.

We will continue testing to ensure that mass production will be worthwhile. Also, I noticed some screens cut off some text at the first line, so I may take measures to scrunch the interface in the initial version 1 of the product. The calendar would also be scrunched.