I don't remember. I learned microsoft basic on a trs80 Coco and if has the same limitation. I bet Getting Started With Color BASIC makes it clear but I'd have to look it up.
I think you can have overlap between type sigils like start$ and stink% are different. But A1 and A10 are the same variable. Interestingly on old Forths I think it's like 3 letters are significant but they also distinguish by the length of the identifier. So hello and hello2 are ok because they are different lengths. And forth lets you use the full character set so that makes collisions less likely. -- John. On Wed, Dec 24, 2025, 2:56 AM Alex <[email protected]> wrote: > I didn't know only the first two characters of variable names were used > either. > > Is that documented in any of the manuals or just the kind of dirty secret > you figure out the hard way? > > > On December 24, 2025 7:32:49 AM UTC, "John R. Hogerhuis" <[email protected]> > wrote: > >> So I was inspired to play with Gemini Pro and generate a little game for >> the Model 100. I get Gemini Pro usage for free as a student since I take >> woodworking classes at a local community college. >> >> The first two levels kind of work. It took a lot of iterating. It cannot >> actually do its own testing in gemini, so it generates the code blind with >> no testing and throws it over the wall to SQA/me (like a real programmer!). >> Somehow it was generating Model 100 BASIC code almost off the bat without >> directly feeding it any manuals. I eventually fed it the quick reference >> guide. >> >> One bad problem it had through several iterations was that it didn't know >> only the first two characters of variable names were significant. So it was >> controlling different things with the same variable. Teaching that fixed a >> lot of frustrating jank and if I realized it it probably would have take >> far less time to get it to this point. >> >> The idea is that you pick up gems and a key. When you get the key the >> portal opens and you can go to the next level. >> >> Kind of fun. I didn't have to write any code at all. I only had to look >> at the code to see why it seemed unable to fix reported bugs, and that was >> a failure to understand our version of BASIC. >> >> -- John. >> >
