4 ms·

I've got some work reversing code out of the 80s and into stuff that works today. Takeaways from porting a 1987 foxpro application: The fact that this still w

I've got some work reversing code out of the 80s and into stuff that works today. Takeaways from porting a 1987 foxpro application: The fact that this still works - If not hanging on by a thread, it's pretty incredible that it still has seen something like 34 years of usage, well more than my years on this planet. The guy who originally wrote it died though, and the CD with the code lost, so I'm stuck reversing old formats. Format encoding - No unicode back in the day. If you read file formats from back then, make sure you get this right and figure out what encoding it was using or you will end up with muntzed data. This was regionally based back in the day. Tooling - I found great use of perl in this mission because it's also good at handling ancient 1980s file formats, it itself being a language from around the same year. Check this out [1] Database typing - Haha guess what format they were using for dates? Well it sure as hell wasn't Y2K compliant. The now-dead original programmer had to go and monkey wrench that in there. [1] https://www.burtonsys.com/download/dbf2csv.php https://www.burtonsys.com/download/dbf2csv.php

inglor_cz 5y ago
Ah, no unicode. Yes, yes. The 1990s in IT in Czechia were a time of several competing encoding standards. Normal ASCII does not cover most Czech accentuated characters, so we had to choose from: a. ISO 8859-2, typical for *nixes and often found on the web b. CP852, an IBM-based DOS standard c. Windows-1250, a Microsoft standard for Windows; therefore, programs localized for pure DOS (with CP852) would have malformed Czech resources when running under Windows 95 or so d. Mac CE, for Apple only e. so-called Kamenický kód, a very clever local encoding (could still be somewhat readable even when switched to ASCII), but with zero institutional support, so doomed to insignificance f. KOI8-CS, a Cold War Czechoslovak standard for a line of computers that could not compete with the Western ones. I've seen Turbo Pascal programs trying to deal with all of them... TGIUnicode.

inglor_cz 5y ago

Ah, no unicode. Yes, yes. The 1990s in IT in Czechia were a time of several competing encoding standards. Normal ASCII does not cover most Czech accentuated characters, so we had to choose from: a. ISO 8859-2, typical for *nixes and often found on the web b. CP852, an IBM-based DOS standard c. Windows-1250, a Microsoft standard for Windows; therefore, programs localized for pure DOS (with CP852) would have malformed Czech resources when running under Windows 95 or so d. Mac CE, for Apple only e. so-called Kamenický kód, a very clever local encoding (could still be somewhat readable even when switched to ASCII), but with zero institutional support, so doomed to insignificance f. KOI8-CS, a Cold War Czechoslovak standard for a line of computers that could not compete with the Western ones. I've seen Turbo Pascal programs trying to deal with all of them... TGIUnicode.