Archiving my Twitter, Facebook and other social network activity
Crazy idea that I can’t believe I’m considering: Building a #gemini front end — well, not a front end so much as the “static” views of profiles and public posts – for @gotosocial
(I really need to finish some existing side projects before starting new ones.)
(Admittedly, a full gemini-based #GoToSocial front-end that enables posting and stuff would be cool too. Though it looks like Bubble may be a better fit for that sort of thing: gemini://geminispace.org/u/skyjake/1 )
In other news, #DreamHost migrated my main VPS with no hiccups as near as I can tell. I had to manually launch Agate for my #gemini capsule, but I suspect that’s only because they brought up the new system before moving the users over to it.
Here’s hoping it’s more stable than the Nginx one. I think I may just not have allocated enough RAM to the Nginx box, since the new plans don’t let me choose in-between sizes like 1.2 GB, 1.7GB etc. like the old plans did. The VPSes both ran rock solid for years, but DreamHost’s nginx support has always seemed like an afterthought.
I’m probably going to drop the Nginx VPS at some point. I just need to decide what to do with the last remaining site on it: Consolidate on the Apache VPS (possibly with more RAM), or set up a dedicated Nanode and manage it myself (and find out what resources it actually needs on a lighter-weight system)
#Gemini question: about a year ago I settled on LaGrange and Amfora as my main Gemini clients after trying out a bunch of what was around at the time. And then I kind of stopped paying attention.
Now I’m curious: have any particularly interesting client applications come out of Geminispace since then?
For some of my posts I’ve manually converted the files so I can fine-tune it. But I have one site I’m auto-generating, and I’ve done a mix of site- and page-specific code like @kensanata (ex. for web links that have a corresponding #gemini location, or removing images by class), adjusting the source to convert more nicely where appropriate, and just letting some of it sit there looking weird for now.
Here goes: I’ve taken the troubleshooting posts that I already put on #gemini and set up a web mini-site with them. I figure on adding more as I go along, redirecting some blog posts and linking others.
https://hyperborea.org/tech-tips/
Digging deeper into #IndieWeb and #Eleventy/#11ty, trying to make it both light and responsive, and building a reusable template setup.
Once I’ve refined it a bit more I plan to generalize it & post a git repo of the structure & styles
More rethinking my website in the #IndieWeb context. The “garden and stream” metaphor brings up a good point: Not everything we put online is sequential.
I have a lot of stuff on my blog that doesn’t belong in the middle of a long stream of time-based posts as well as it would in a topic-based collection. Heck, I already reorganized the tech troubleshooting when I mirrored stuff to #gemini
https://indieweb.org/garden_and_stream
gemini://hyperborea.org/howto/
Replying to @ajroach42 asking how people manage their gemlogs:
I write my main gemlog by hand since the code is so simple, and I wrote a perl script to manage the tags, categories, lists and next/previous links.
For another blog that’s in 11ty I wrote a shell script to send each page to md2gemini, which is a really useful markdown-to-gemtext converter: GitHub: md2gemini
My super-awkward but fully self-contained gemlog management script, still full of too much hard-coded stuff specific to my own site that I intend to eventually generalize: GitLab: gemloginplace
Response to a question about telling Firefox to open Lagrange for Gemini links
I don’t remember exactly how I set that up, but I’ll see if I can configure a fresh profile and write it down this time!
What I remember: I either pasted a gemini: url into the Firefox URL bar or clicked on an actual hyperlink to a gemini: url and it asked what application to open it with.
Firefox also has a per-website security setting on whether a website is allowed to open links in other protocols.
Yeah, that’s basically it. Paste the URL into Firefox, choose the application (either from the list or from the file picker), and check the “always…” box. Then gemini: links in web pages will open the confirmation box and you can click always-allow per site.
(It won’t work on Mastodon, unfortunately, because Mastodon doesn’t turn gemini: URLs into links.)
I’ll write this up on the #gemini capsule too!
—
Here it is: opening #gemini links in your preferred client from Firefox.
When I write something new on my gemlog or import an older blog entry, sometimes I’ll also import related posts so I can link within the #gemini capsule instead of out to the web.
I was importing a post on how functional memorabilia is both irreplaceable and easily replaceable depending on how you look at it, and it mentioned my #Mozilla #Coffee mug – which I still have.
Gemlog: (Ir)replaceable?
Blog: (Ir)replaceable?
But yeah, Mozilla Coffee was briefly a thing, back in 2003-2004. It wasn’t a run by Mozilla, it was someone who had a mail-order coffee business who branded some of it and donated half the profits to the Mozilla Foundation.
Gemlog: Remembering Mozilla Coffee
Blog posts tagged Mozilla Coffee
If you’re not familiar with Gemini, it’s sort of a stripped-down web or souped-up Gopher, very bare-bones and text-oriented. Perfect for blogging.
A bit more in this post, including links to info about Gemini and related software. (You do need a different browser to access Gemini links – I’ve been using Lagrange on the desktop & Ariane on Android .)
Still a bit rough around the edges, but my Gemini conversion for my Les Misérables commentary is up! The blog is now available both on the web and via Gemini.
It features commentary from two full read-throughs (in different translations) plus reviews of movie, stage, radio and comic adaptations.
Gemini: Re-Reading Les Misérables
Web: Re-Reading Les Misérables
It’s still very experimental, but I’m messing around with #gemini at my new capsule, gemini://hyperborea.org
So far I’ve copied a few pages from my website and started a #gemlog. Mostly I’m trying to get a good sense of what works best in the ultra-minimalist environment. I’ll probably still be moving things around for a while.
Pages copied over so far include tech troubleshooting and some old MIDI files. The log’s mostly a subset of my blog so far (life, tech & entertainment catgegories), plus some recent posts on my early impressions of Gemini itself.
Question for #gemini users – is there a consensus on the best way to handle photo galleries? Browsing around I’ve mostly seen just index pages with links, which I suppose works, but it feels like the old days of Twitter when you had to click on a link to twitpic or yfrog or wherever and there was no preview to give you any hint unless the author posted a good description in the tweet.
I do like the way Lagrange handles that case, though, loading the linked image inline when you click on the link. It helps to avoid losing the context of where you are.