The huge archive reconfig project continues.
The TRS-80 Software Archive has been growing since its inception in 1994. When it became apparent that screenshots would be helpful, I had to run each program, save a screenshot, and link it to a file. At first, it was filenames, but those change over time as other versions or copies of programs came out, and so it needed to be linked to something more permanent … a file description and author.
That was all well and good for a long time until I realized that many of the “Author Unknown” programs I had probably were the same copies of programs that I did know the author, and that someone along the line just deleted those lines from the basic program.
At that point it became clear that 20 different copies of “Biorythm” (or Blackjack or Star Trek or Roulette or Poker …) by “Author Unknown”, were not 20 copies of the same program, but 20 different programs … all linking to the same screenshot.
So the project emerged. First, I needed to go through every file and adjust the filename to 8 characters. If I had fewer, then an attempt to find and get a file could bring in more than the one I needed.
Then, I needed to go through every single file that was “Author Unknown” (around 6,000 files), run them, and change the zip comment to say something unique about the program. It couldn’t just be the title screen, or “Enter your name”, because those were too easily removed or modified (or simply the same) by the same people who edited out the author name to begin with. So I needed to run each program somewhat in earnest to get to some differentiation point.
But running each program isn’t so easy. Some require a certain DOS, some require a certain TRS-80 model, some can only run in Level II BASIC, and some don’t seem to work no matter what I do. But they all have to be run and their zip comment expanded. At least this stage also allows me to see if there was an author name further in, or if a program is public domain, so I adjust accordingly.
Once that was done, I needed to audit the database finding every cross-reference I had to any file with an “Author Unknown” and get rid of it, because if any existed, a file could easily default back to the “Blackjack” by “Author Unknown” even it had a bracket. So the database needed to be scrubbed.
After that, I needed to go through AGAIN (yes, very wasteful, but the programs I use to maintain the archive need to be updated with the proper zip comment to move to the next step), running each program to get a meaningful screen shot. Sometimes this step, which had me go deeper than just trying to get a few words to differentiate, would show that the file had errors, needed a dependency I didn’t have packaged with it, or just locked up or produced no screen output. For these, I needed to see the error could be corrected, if I could find the dependency somewhere, or determine why the lock up.
For this stage, I had to go to George Phillips to once again implement some new features into TRS80GP that probably only I would be using. One, was to allow me some way to turn on TRON and TROFF, so I could see if a program which wasn’t doing anything was locked or was doing something but the not showing it. Then I needed a way to re-enable a disabled break key, largely for the same purposes (people LOVE to disable their BREAK keys). Then I needed an easier way to get the time and date into the emulator because many programs required a valid date/time or they would error out. Then I needed a way to actually get printer screenshots, because many programs had printer output only … which also doubled to help me figure out if a program had locked or wasn’t displaying anything … when it was simply printing.
This is the stage I am in now. I have about 800 more files to go to finalize getting all the “Author Unknown” files, now all having descriptors, to have screen shots which cross reference to those expanded filenames.
Once that stage is done, I need to go through the other 22,000 files to put in descriptors on files where the author is known. After all, it is all well and good to have screenshots of each of 20 different “Biorythm” programs, and to note which ask for Birthdate and which ask for Birthday, but until I also know that for “Author Known” files, I won’t be able to match them up.
Once they all have descriptions, I then need to try to see what I can do to combine files. At that point, the project will be done, because going forward I know to put descriptors on all new files added to the archive.
(Dusty).jpg)
(CINTUG)[1979-09].jpg)
(CINTUG)[1979-11].jpg)
(CINTUG)[1979-12].jpg)
(Dusty).jpg)
(VR-Data).jpg)