Or alternatively titled: "A (pointless) exercise in shell scripting".
Recently Twitter turned off non-ssl use of the API. This affected a lot more TTYtter users than I thought it would; one of the key benefits of TTYtter is the ability to set it up to track tags and mentions such as
#ttytter and this way an invisible support network is formed (even to those unfortunate TTYtter uses not tracking those terms). I had assumed, since using TTYtter requires at least a base level of technical know how, that most users would figure out what was going on. Perhaps because it had not been updated since the end of 2012 most just assumed the issue was with TTYtter?
Anyway, this episode prompted me to think seriously about delving into to the inner workings of TTYtter, especially after stumbling across this comment from the author who has abandoned ship to App.net (with the TTYtter inspired Tex.app). I thought a good starting point for understanding the almost 8000 lines of Perl code would be figuring out defaulting to ssl. And then perhaps looking at the fairly recent " -- rate limit rate failure: using 180 second fallback" issues of which the temporary fix was commenting out a couple of lines.
First of all I wanted to build up a full Git history of changes; Floodgap do not provide a public version control system, just directory listings of previous versions. I thought having a full history would give me a better chance of understanding the code. I figured this would be a quick shell script to save doing it all manually. Ha. It took me 4 days. And loads of tweaks for things I hadn't thought about. I could have done it manually.
So in the end I got the full history built up only to then discover that the two issues I sought to fix had actually been fixed, it's just that the version number had not been incremented.