Desprez,
Keep in mind that ResEdit is 21 years old. For software that's an incredilbly long time to still be in use. Hardware and software in general have evolved far beyond the environment that originally spawned ResEdit, so it's not too surprising when it doesn't play well with others. Personally I rarely have ResEdit crashes, but then again I'm running it with OS 8.6 which probably helps a little bit. You've already taken care to save often, increase memory allocation, and watch out for resource fork limitations, all of which are the most common causes of problems. That said, here's some things that I do that you might try out in order to minimize your problem:
• One resource fork limitation that is not easily monitored is the 32767 byte maximum for resource names within a given resource type. All the other limitations can be readily observed (file size, type count, total resources), but nothing tallies up the total character count for the names within each resource type (sounds like a candidate for an MC utility). When I suspect such a problem, I use the "listing only" option of wOOtWare's ConText utility and then strip out everything except the resource names themselves with a text editor. Then I can either look at the resulting file size or use the text editor's character count capability to see if the 32 KB limit has been broached.
• ResEdit does not play well with other applications, so running it as a solo application can help. This includes background tasks as well. Sometimes I get carried away and wind up with EV:N paused while I open a suspicious resource with ResEdit and then open a specification in a text editor, then fire up a spreadsheet with relevant data and then open some more files with any or all of those applicatiions. And sure enough after 20 or 30 minutes of this I get a CPU lockup or crash and have to reboot. Now I don't know that ResEdit is the culprit in those crashes, but I've simpy gotten out of the habit of having multiple applications running simultaneously with ResEdit.
• ResEdit has some flawed stack and heap housekeeping routines, so I always quit ResEdit after each editing session. Even when working with just ResEdit and no other applications I've had problems in the past if I use it for more than 30 or 45 minutes at a go. And ResEdit doesn't take very long to launch, even on my archaic 300 MHz G3, so relaunching is not a real burden.
• Now another thing to remember is that even when running just ResEdit you're actually running multiple editors. Every time you invoke a Novatools resource editor you've launched a program within a program. I'm not saying that Novatools itself causes crashes, but feeding unspecified experimental data into a resource via ResEdit Novatools can sometimes cause a crash, since there are calculations being made within the Novatools resource editors that can be put into divide-by-zero or endless loop situations. If you encounter a repeatable situation that causes crashes, try using just the EV:N TMPL to edit your data and see if the problem goes away. If using the TMPL method clears up a specific repeatable problem, then you're probably asking Novatools to do something illegal.
That's all that comes to mind for now. Keep up the good work that you've been doing. We all really appreciate it.
