Auto-Making a Gemini Index
A few days ago on Mastodon, I came across a very useful toot by Fixato. He had provided a comprehensive shell command for updating you TLS certificates in light of the recent update to the agate server. He also kindly helped me troubleshoot additional issues I had.
Fixato followed me back and came across a recent post I made on automatically updating my gemlog index page via a shell script. I’m new to shell scripting, so my solution had lots of room for improvement. To my amazement, Fixato took up that challenge and wrote a wonderful response to my post:
I know it’s a cliché to people already using Gemini, but the ‘socializing’ potential of Gemini, the smolnet, and the floss community in general never ceases to astound me. I was so happy that somebody else, a total stranger, could happen upon my attempt to solve a technical problem and offer a far more elegant and effective solution. Of course, this is how open source and free software has always developed, it’s just nice to be reminded of that. Like many others, since discovering Gemini, I’ve ditched things like twitter for the most part, and I don’t miss them at all. In fact, I am angry at myself for allowing applications like that to take of so much of my time in the past without delivering anything tangible in return.
Needless to say, I will be incorporating Fixato’s script into my workflow. He points out in his post that his solution may not fit my needs exactly, since I was manually inputing a date separate from the file. That was due to my own inexperience working with archives/data. I can see now that it is indeed much more efficient to include the date somewhere in the file. As Fixato mentions, this is especially that case when considering posting the material elsewhere. For now, I’ve just changed the filenames of my gemlogs to include a date, but in the future I might work more on structuring the files themselves more rigorously, with dates and tags.
In the same vein, I also like how Fixato’s script takes the ’title’ for the index listing from the actual file heading. Before, I was inputing the titles manually, and that tended to lead to discrepancies in the index title/file title. I far prefer this, more consistent, method (and it involves less typing!).
Moving forward, I’d like to adapt Fixato’s script to also generate a ’tag’ index page. This is really the best thing about responses like Fixato’s - they are opportunities to learn and grow. His background, I presume, is in tech, so I can learn so much from his solution to my problem. I’ll have to learn more about how functions work in bash, but using his script as a template I think I can get a lot closer to that.