day's thoughts: getting on the fediverse, css wrangling, bootleg compilation complication
Three things going on today:
I want to get back on the fediverse, and I'm a lot more interested in self-hosting this time, but I need to weigh my options yet.
I got this blog looking how I wanted, but the CSS is a mess of spaghetti !important tags.
Noticed today that the highest-bitrate extant version of my favorite Eiafuawn track has a big nasty audio glitch right in the middle, which makes me sad.
More of each after the break.
rclone, pt. 2
Figured out where the 100GB of storage in my GDrive came from! Turns out I'm on my ex' Google Families, and she pays for 100GB which ends up shared across all of us (although none of them really touch it).
I figured out that after the API rate limit kicks in, I can upload to it using rclone from my Windows machine at about 18GB/hr (which works out to 40Mbps, maybe 1/3 to 1/2 the speed of my home internet). Those are pretty reasonable speeds for unattended copying, but anything interactive that requires lots of reads (like scanning metadata with Navidrome or Foobar2k) is hell, and writing media tags is even worse (GDrive doesn't support partial file updates--it needs to download it in full, then reupload in full!). I'm having difficulties getting Navidrome to even scan it fully... I think it hits an API rate limit on individual read accesses, which slows it down, and then it gives up. I set it to scan every 24 hours, in the hope that it'll try again after the rate limits reset and finish scanning.
rclone actually rocks, I take it back
I just got rclone configured (with my own GDrive API key, a local 4G writes cache, and all the other fixins), and it works better than I thought it would. I've got my Navidrome media folder pointed at it now, with a couple albums in FLAC uploaded to it for testing, and even with transcoding off it seems to work without hiccups.
The thing that tipped me in favor of trying it was realizing that my second Google acct has 100GB of its own that's completely unused, and which I don't think is tied to any temporary bonus storage offers (like the extra storage on my other Google acct, which I think is from a 1-year storage offer with my old Chromebook).
Something else I realized: when I do eventually get that 1TB of storage on the host's internal network, I can use GDrive as a second destination for backups (the provider has their own backup service for like 50c/mo, which I haven't signed up for yet but probably will).
Now I need to get rclone set up on my (Windows) laptop, so I can copy all my FLAC to GDrive without using this box itself as an intermediary, which is thrashing my disk and network unnecessarily. It's not slow, just slightly disconcerting to be pushing bits through so many layers.
First post: progress so far, rationale and intent, and the human side of this box
On multiple occasions, Jason Scott, proprietor of textfiles.com (among other projects), has spoken at length on the importance of understanding the human side of tech history. This shows quite clearly in his documentary works, BBS and Get Lamp, which, while being documentaries about the BBS and the text adventure, are more importantly collations of personal and community memory. The technical and implementation details, in his eyes, are secondary to the socio/anthropology of the spaces. I share that passion and carry it to most of my interactions with the digital realm. This box is somewhat more than a domain name and some symlinks in /var/www/whatever: I hope for it to be a reflection of myself and my own digital heritage, in addition to some neat info and services.
(Lord, maybe I should've been an English major.)