I'm a stickler for quality but back when some of these documents were created I'm sure they didn't have scanning and OCR in mind much less thought it might even exist.
A large proportion of these documents I have collected over time. In a lot of cases they were scanned and maybe even OCR'd by someone else. Where that has occurred I have tried to retain the credits in the METADATA.
I'm a big fan of PDF done right so in a lot of cases I have re-optimised and re-OCR'd the document. The objective was to improve the quality of the document (yes that is possible even with an old scanned image based or OCR'd PDFs) and reduce its file size. Its a balance.
At this stage I have been using Adobe Acrobat's Clearscan method for OCR. Its really good. Unfortunately Adobe software has become too expensive for me to maintain so I am currently evaluating Nitro PDF to see what results it can produce. Preliminary testing produced results similar to Adobe Acrobat so I'm optomistic.
Most documents that are on this site, or eventually will be, have selectable text. Please keep in mind that while the page may look quite presentable (or even better than the original document) copying the actual text might produce mixed results. OCR, particualry on old documents, is not an exact science.
At this stage the TRS-80 Epoch does not support searching the actual content of the file. The technology that drives The TRS-80 Epoch does support file search but I have chosen to turn it off at this stage. That is another project.
Apart from wanting to properly archive TRS-80 documentation I aslo wanted to create my own repository for search. My current search spiders rely on some heavy hitting of other sites so the plan is to eventually index the file content of The TRS-80 Epoch and not burden other collectors who run TRS-80 web sites on a small budget by consuming all their bandwidth.
The TRS-80 Epoch