Thanks guys -- John.
On Mon, Oct 19, 2026, 8:39 PM Gary Weber <[email protected]> wrote: > Kurt's scan is in three parts: > > https://www.web8201.net/Files/LIBRARY_web8201/nec8201_techref_pt1.pdf, > https://www.web8201.net/Files/LIBRARY_web8201/nec8201_techref_pt2.pdf, > https://www.web8201.net/Files/LIBRARY_web8201/nec8201_techref_pt3.pdf > > And yes these are getting uploaded to archive.org, along with all of the > other PDFs I have fro the various other manuals. > > Gary > > On Mon, Oct 19, 2026 at 7:44 PM B9 <[email protected]> wrote: > >> I suggest uploading to archive.org as a start. >> >> --b9 >> >> >> On October 19, 2026 6:54:45 PM PDT, Kurt McCullum <[email protected]> >> wrote: >> >>> I've just finished scanning the manual but it's just over 36 megs in >>> size. I need to figure out where to upload it, but when I do I will send >>> the links. >>> >>> Kurt >>> >>> On Sun, Oct 18, 2026, at 9:41 PM, John R. Hogerhuis wrote: >>> >>> I think it was Kurt McCullum that had a clean copy... >>> >>> I need the following pages from the NEC tech ref: >>> >>> Pg 77, 78, 79, 97, 99, 105, 177, 119, 191, 198, 210, 211, 228 >>> >>> What I am working on is a very accurate OCR with good formatting of the >>> ASCII art and accurate content for all assembly listings, etc. I'm using >>> gemma4:31b on a local rig to do the conversions to ASCII and ultimately >>> .md. My current purpose is informing the AI as to low level access to file >>> system etc for it to generate VirtualT socket tests for Gary's port of >>> SuperROM. >>> >>> The AI is doing a much better ocr of the content than the PDF has. It >>> is doing an excellent job. Even at ASCII art. Sometimes hallucinating >>> portions of diagrams that aren't there, which is bizarre. And it keeps >>> stopping and needing to be kicked to continue working. Even so it will save >>> me a lot of time and we will all have a more tractable version of the >>> content, ultimately. >>> >>> Thanks, >>> >>> -- John. >>> >>>
