3 ms·
It's interesting to note that this is essentially a domain-specific structural editor; each keypress is a full lexical token that happens to be valid in the TI-
by zozbot234 7y ago
It's interesting to note that this is essentially a domain-specific structural editor; each keypress is a full lexical token that happens to be valid in the TI-BASIC programming language, and a different input mode is triggered for full alpha/numeric input. That's how one can cope with an input method that's as low in bandwidth as a fiddly calculator keyboard.
- avhon1 7y agoEh, not really. The calculator won't stop you from creating syntactically-incorrect programs. You can enter any nonsense you like. Errors are detected at runtime. You get one of 8 (I think?) error codes, and can Goto the line that the error is on. Figuring out what is actually wrong is up to the user, and often requires consulting the manual. While I was in high school, some people figured out that they could write notes in their calculator as programs, and covertly reference them during tests. The calculators didn't care at all about having unquoted strings and nonsense math stored in the programs, because the students never ran them. The teachers caught on to this and started checking that calculators were wiped before tests.
- int_19h 7y agoThis is a fairly typical approach for old-school BASIC. Back in 80s, most BASIC microcomputers used a similar approach to entering tokens, despite having proper alphanumeric keyboards. For example, look at the keyboard of this ZX Spectrum: https://en.wikipedia.org/wiki/ZX_Spectrum#/media/File:ZXSpectrum48k.jpg https://en.wikipedia.org/wiki/ZX_Spectrum#/media/File:ZXSpec... It's not just input, either: many of those systems stored programs in pre-tokenized binary form, as well. It makes a lot of sense when your RAM and storage space is that limited.