Squeeze Play Attempt
Jan. 1st, 2024 07:00 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
krpalmerWhen I’d restored an antique Color Computer 3 word processor not just to minimal usability (running in a more elegant emulator) but to minimal usefulness (with another program to turn that emulator’s “printer output” into printable PDF files), I did ask myself how much “dot matrix aesthetic” I needed these days. Still, making printouts of some of the (not very good in retrospect) stories I’d written more than thirty years ago with Max-10 wasn’t the only quick and simple thing I could now imagine doing with it. Its ads in the late 1980s had played up combining graphics with styled text (as MacWrite’s ads had in the mid-1980s.) To get the graphics was just a little complicated. Instead of using Max-10’s complement CoCoMax III, I resorted to the generation, importation, and conversion of RLE image files. In the end I had a PDF printout... and the impression it looked vertically stretched. I’d had a sense of this happening before. Having needed to adjust the Epson graphics code in a hex editor to make the printout more than one black bar, I’d then felt out how to reduce the blank top and bottom margins and keep each Max-10 page from spilling onto a second PDF page. This, however, was that much more obvious. Zooming in on a PDF page revealed the command-line program I’d managed to compile formed its graphics from tiny circles, and I was starting to wonder if the circles were a little further apart than they could have been.
One thing I was thinking of was taking some printouts home for Christmas to compare them to the antique output I still had saved on a shelf. Not that long ago now, I’d taken another look at a story from high school intended to be a magnum opus. It was supposed to have run for eight hundred-page “books,” but I’d only finished two of them while coming to wonder if I’d only complete the whole thing around the turn of the century. Years before that, I abandoned it only somewhat more than a quarter finished. With that new look many years later, all of a sudden I thought about the quickest conclusion I could write (that didn’t amount to “everyone gets blown up”), and even managed to write it. Now, I intended to go not just from one typeface to another to hint at the passage of time, but from a “Max-10 printout” to a modern printout. As I finished that, though, I did just happen to find an old printout of something else tucked into a binder of mostly hand-printed pages I’d picked up from home a while ago. That was hard proof the type in my new printouts was indeed stretched out.
My thoughts did turn to a different program that had also promised to turn Epson printer codes into PDF files but which I just hadn’t been able to compile. I resorted to the Linux virtual machine I do keep “just in case,” and at last I managed to not get the program running but to remember how to get files into and out of the virtual machine. These PDF files weren’t as stretched, although I did notice the program output separate pages as separate files and, although it formed its graphics from tiny squares, produced output that took up more disk space.
Faced with that I did start looking at the source code of the program I’d used before. As minimal as my ability to program in any sort of modern computer language is, there were just enough comments that I wondered what changing a number on line 419 of the file Interpreter.cpp would do to the spacing of the dots. The program continued to compile, but the first thing I noticed was space was beginning to open up between narrow strips of output. It took more experimenting to make it obvious those strips could be compressed by reducing that one number, but I had to deal with some other part of the program as well.
Taking another look at the source code, I found a comment that made me think line 194 in Interpreter.cpp might also be crucial. My first thought was to reduce the second number in that line by the same ratio I’d reduced the number on line 419. That didn’t work. I then just happened to think about increasing the number by the proportion I’d reduced the first number by, and now things hung together. Eventually, I supposed that changing the number on line 419 to 10 and the second number on line 194 to 216 did the trick. Mentioning that here isn’t the same as setting up on GitHub to “fork” the code myself, but it still just might get into the hands of people who actually know what they’re doing.
With that done I did happen to take one more look at the program I could only run in Linux, and I just happened to find a “fork” of it promised to compile on macOS by taking out references to software components not installed by default there. It’s an alternative, certainly, even if the circle-formed graphics are somehow just a little more interesting. One thing about the other program, though, is that it can use the Epson codes that set plain text to bold or italic. That would seem to mean I could use the Color Computer word processors that didn’t depend on the overhead of “forming everything out of graphics,” but as I said before that remains a bit of “dot matrix aesthetic” I haven’t got around to yet.
One thing I was thinking of was taking some printouts home for Christmas to compare them to the antique output I still had saved on a shelf. Not that long ago now, I’d taken another look at a story from high school intended to be a magnum opus. It was supposed to have run for eight hundred-page “books,” but I’d only finished two of them while coming to wonder if I’d only complete the whole thing around the turn of the century. Years before that, I abandoned it only somewhat more than a quarter finished. With that new look many years later, all of a sudden I thought about the quickest conclusion I could write (that didn’t amount to “everyone gets blown up”), and even managed to write it. Now, I intended to go not just from one typeface to another to hint at the passage of time, but from a “Max-10 printout” to a modern printout. As I finished that, though, I did just happen to find an old printout of something else tucked into a binder of mostly hand-printed pages I’d picked up from home a while ago. That was hard proof the type in my new printouts was indeed stretched out.
My thoughts did turn to a different program that had also promised to turn Epson printer codes into PDF files but which I just hadn’t been able to compile. I resorted to the Linux virtual machine I do keep “just in case,” and at last I managed to not get the program running but to remember how to get files into and out of the virtual machine. These PDF files weren’t as stretched, although I did notice the program output separate pages as separate files and, although it formed its graphics from tiny squares, produced output that took up more disk space.
Faced with that I did start looking at the source code of the program I’d used before. As minimal as my ability to program in any sort of modern computer language is, there were just enough comments that I wondered what changing a number on line 419 of the file Interpreter.cpp would do to the spacing of the dots. The program continued to compile, but the first thing I noticed was space was beginning to open up between narrow strips of output. It took more experimenting to make it obvious those strips could be compressed by reducing that one number, but I had to deal with some other part of the program as well.
Taking another look at the source code, I found a comment that made me think line 194 in Interpreter.cpp might also be crucial. My first thought was to reduce the second number in that line by the same ratio I’d reduced the number on line 419. That didn’t work. I then just happened to think about increasing the number by the proportion I’d reduced the first number by, and now things hung together. Eventually, I supposed that changing the number on line 419 to 10 and the second number on line 194 to 216 did the trick. Mentioning that here isn’t the same as setting up on GitHub to “fork” the code myself, but it still just might get into the hands of people who actually know what they’re doing.
With that done I did happen to take one more look at the program I could only run in Linux, and I just happened to find a “fork” of it promised to compile on macOS by taking out references to software components not installed by default there. It’s an alternative, certainly, even if the circle-formed graphics are somehow just a little more interesting. One thing about the other program, though, is that it can use the Epson codes that set plain text to bold or italic. That would seem to mean I could use the Color Computer word processors that didn’t depend on the overhead of “forming everything out of graphics,” but as I said before that remains a bit of “dot matrix aesthetic” I haven’t got around to yet.