Splitting up the work

[davidlinsley]

I've created an issue for each source file (14) under https://github.com/davidlinsley/DragonBasic/issues and tagged them with OCR.

If you plan to convert the whole file, I'd suggest assigning the issue to yourself and remove the Helped Wanted label, otherwise leave it unassigned but comment on the issue saying you are taking scans 153-161 for example. We can then take chunks of work without duplication. There are more files than there are us though, so you can assign and then unassign later when more help is needed.

@xtramural do you want to take ECBOEM.BAS and ECBO64.BAS for old times' sake?

[davidlinsley]

See #2 for suggested repo layout for the OCR. Please comment on that thread with any suggestions on changing that.

[davidlinsley]

I've committed the first fixed up page so you can see the column alignments: https://github.com/davidlinsley/DragonBasic/blob/main/ocr/ECBRAM.SA/CCI10222022_0001.txt

[davidlinsley]

I've completed all the files that appear to be "core" - Microsoft copyrights but no METTOY macros. I'll pick up ECBMOD next.

[davidlinsley]

I'm done with all the Microsoft files (that begin with Copyright Microsoft comments). Duncan has volunteered to pick up his two files, which are all that remain now. We are through 311 of 340 pages!

[xtramural]

Hi David, Huge thanks for all you’ve done to get this far. On that note... I know I’ve volunteered by my GitHub skills need refreshing since it’s so long since I’ve contributed to any projects. It may be a couple of days before I get my head around this so I’m worried the project will stall unnecessarily waiting for me to get my act together. Once all 340 pages are done what are the next steps? I guess trying to assemble from the source files? Cheers, Duncan

…

On 24 Jan 2023, at 15:35, David Linsley ***@***.***> wrote: I'm done with all the Microsoft files (that begin with Copyright Microsoft comments). Duncan has volunteered to pick up his two files, which are all that remain now. We are through 311 of 340 pages! — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you were mentioned.

[davidlinsley]

Don't worry about time, I've no expectations. I had no idea I'd rattle through the rest so fast and thought it would be a much longer slog. The quality of the OCR service, and then the workflow I settled into made it quite fast. If you need help, or want me to describe my workflow, I'm happy to help.

Next steps for me will be trying to get authorization from Microsoft to either just open the repo to public (they grant me permission to release their source), and/or have them take a copy their source for publishing. It was suggested to me that we redo the time stamps on the source files, so they show as 1982/83 :) Apparently users got a kick out of seeing that in the GW-BASIC sources.

Getting this to assemble would be a great project, and I see two forks there: assembling on a PC with lwtools/asm09/asm6809 (using a script to transform source would be preferable, to easily pick up any fixes for mistakes we find, but maybe Cairan Anscomb would be up for making asm6809 support the format as-is) and also using rasm09 in the Exorset simulator. The WDA files list multiple Exorsets as potential collateral against loans, but I also read (on The Register article from 2012 I think) that a PDP-11 was used. Do you recall? I'm in touch with Frederic Le Duigou who wrote the Exorset simulator on trying to get rasm09 working.

However, I'm not so sure it was rasm09, or maybe there is a newer version than ~1979 that I found the .dsk image for. The documentation (see the doc folder in this repo) doesn't mention the ability to include files that I found, but the listing starts each source file with ECBCOM.SA on drive 0 before moving to the real name (like ECBRAM.SA) on drive 1. Given the listing output is clearly using the rasm09 format we're missing some information, as it looks like ECBCOM.SA defined some macros (like SKIP2) and then included each source file. Maybe it just needs more reading and experimentation in the simulator to figure it out.

Speaking of macros, the listing glosses over what is being output for the likes of SKIP2! We can see what they are in the ROM though.