kevin@kosman.UUCP (Kevin O'Gorman) (07/19/89)
I'm looking for any comments and suggestions on the following events: I ported Crystal Cave from UNIX to the Mac on LSC 2.15, and it seemed to work. It was my very first Mac program, so I was pleased. I ported it to THINK C 3.0, and it still seemed okay, so I sent it off to comp.binaries.mac. It's been a while and I haven't seen it here, but it seems to be out on the net somewhere, because I'm starting to get bug reports. And they puzzle me. People told me that "after a while" the stuff they typed no longer would show up on the screen. They could save the game and restore it, and all would be well for another while. I tried this under the THINK C debugger, and could not get it to happen, but it does indeed happen when running standalone. Weird. One savvy user had suggested I was getting a full TE record, so I used TMON to check that. While I goofed and did release a version that does not truncate the record, the thing was pretty well frozen with 12.5 K bytes in the record, which should be pretty comfortable. I stepped through the input loop, and there's a perfectly normal call to TEKey. It returns okay, but the screen is not updated, and hText has no more characters in it. Well, I was fiddling around anyway, so I started selecting text and seeing if I was updating the menus right. I was, but after a Copy, sometimes the selection range highlight had become permanent. This seemed to go away if I changed the monitor to 2-colors. I went back to 256 and it was still okay for a while, but then with repeated attempts, I could get a permanent highlight again. The oddest thing was that at this point, the original problem had gone away, and I could continue typing to the program. I hardly know where to start looking, though it does seem to be a good idea to put in a truncation algorithm on the TE record. Help?