* * * * * ♫I can't dance, I can't relate, only thing about me is the way I confabulate!♫ Mark [1] and I were talking about Roko's dancing basilisk [2] and he suggested I feed it my 6809 ANS Forth implementation [3], on the assumption that no one has fed an assembly-based project through it. Before feeding it that one though, I decided to try a simpler program, my 6809 disassembler written in 6809 assembly code [4] and … well … I'm not sure if anything is terribly wrong with it because it's just the source code in prose [5], and repetative prose at that. But it's a single 1,200 line file—way smaller than mod_blog [6] and a09 [7]. Makes sense that it's probably okay, if boringly repetitious. I then ran it over my ANS Forth implementation [8] and … wow … it can't count [9] to save its life. None of the line counts are accurate in that table, and the line count for the source file is way out of line. No, the source file [10] isn't 30,000+ lines of code! Yes, it's sizable, at 12,000 lines, but it's not 30,000+. But it does lead me to believe that any project that is close to, or over, 10,000 lines of code will have errors in the generated “documentation.” The other sizable error I found before falling asleep is the main runtime routine isn't DOCOL (which is a name used by a lot of 8-bit Forth implementations) but forth_core_colon.runtime. If there are other errors, I don't know. I couldn't make myself to through more of its “documentation.” Ugh. [1] http://gladesoft.com/ [2] gopher://gopher.conman.org/0Phlog:2025/12/02.1 [3] https://github.com/spc476/ANS-Forth [4] https://github.com/spc476/6809-DISASM [5] https://deepwiki.com/spc476/6809-DISASM [6] https://github.com/spc476 [7] https://github.com/spc476/a09 [8] https://github.com/spc476/ANS-Forth [9] https://deepwiki.com/spc476/ANS-Forth#repository-structure [10] https://github.com/spc476/ANS-Forth/blob/master/forth.asm Email author at sean@conman.org .