Markdown Mail

(replying to a complaint about HTML Email)

Personally I think something like basic* Markdown would have worked out better: just enough formatting to be useful, while still keeping it fully responsive to different displays, safer to include in web apps, easy on bandwidth and human-readable in text-only clients.

A lot of that syntax was around as informal conventions in newsgroups and plaintext email, and we would’ve been better off if it had been standardized instead of just plugging in the full web renderer. Though that’s still better than Outlook using Word to render formatted mail.

* Inline formatting, headings, quotes, maaaaybe links. Not any of the extended syntax for tables or anything that introduces layout or breaks readability.

Whoa…the email forwarder I’ve been using for more than 25 years (ok, *that* makes me feel old) is discontinuing one of thei…

Whoa…the email forwarder I’ve been using for more than 25 years (ok, that makes me feel old) is discontinuing one of their non-email features: URL forwarding. You could set up a redirect from your username on their website that would always point to your current real website, no matter where it was. I used it extensively in the late 1990s as I moved from AOL to one set of school servers to another and finally to the domain name I registered after graduation.

I wonder how many links to that old URL still exist somewhere online. (Not counting the Wayback Machine.)

Wayland/NVidia status update on my machine

I’ve had no problems running GNOME and Wayland-native applications on Wayland and my NVidia card for the last…year?

XWayland, however, still has issues on this card. 2D apps are a bit laggy, but running 3D games? Slow, flickery, or both.

Which means I’m still logging into an X session to play Minecraft or anything from my Steam library.

I tried running Minecraft directly under Wayland over the weekend after installing the latest driver update, but it’s still not playable.

With Fedora planning to drop X entirely soon, I’m going to have to start thinking about what to do if the remaining compatibility issues aren’t fixed by the time F39 hits EOL.

  • Stay on an unsupported OS?
  • Switch to another distro?
  • Move all my games back to the Windows partition (just in time for Windows 10 EOL)?
  • Buy an AMD card…and a new motherboard that doesn’t have this one’s incompatibility with Radeon GPUs, and a new processor to go in the new board, and probably new RAM…

On the plus side, if I do that, I should be able to switch the rest of my Windows games over to Linux+Proton (except for bedrock Minecraft) and reclaim that space instead of staying on a soon to be EOL Windows version.

I think if I wasn’t already looking at buying a new laptop I’d be more inclined to upgrade multiple components at once.

I need to dig up my notes on what exactly was incompatible between this mobo and the Radeon GPU I tried to use a few years back, and whether I went with NVIDIA because the board conflicts with AMD GPUs in general, or because I just didn’t want to go through more rounds of compatibility testing. Or just redo the research from scratch. Assuming I can get a search engine to actually show matches for the specifics instead of what it thinks I should’ve asked for.

Argh…the only post I made about it was too vague, only mentioned that “Indications pointed to chipset compatibility problems with the mobo.” And I couldn’t find any notes on my computer about what chipset was likely to be the problem. No bookmarks either. The only thing I found was a reddit thread I’d saved to Pocket, which suggests that the ASMEDIA driver on the board for the SATA controller can interfere with AMD driver installation on Windows (one of the problems I had at the time), but switching it to IDE mode clears up the conflicts. Oddly enough, I ran into another issue with the ASMedia controller a year later, which involved Windows completely losing track of drives attached to that controller, which I solved by moving the cables over to the AMD controller instead.

It’s possible I saved the thread to Pocket because of the disappearing drives, but it’s also possible it’s the same underlying issue, and now that I don’t have anything attached to that controller, an AMD GPU will work properly on here. If so, that’s a much simpler (and cheaper) upgrade than swapping in a new mobo/cpu/ram combo at the same time.

It’s even worse

Replying to a comment on why Ro would have seen someone’s post about The Bad Space.

It’s even worse: The relevant Kolectiva user replied directly to one of Ro’s posts on what had previously been a constructive thread about blocklist sharing in the context of last week’s spam attacks.

Original thread start:
https://sauropods.win/@futurebird/111969325179943537

Ro replied in the discussion with a link to an article he’d written at Nivenly about the topic. The person in question replied saying he wouldn’t trust Nivenly because the Bad Space was a Nivenly project. Ro replied that it wasn’t true, pointing out that TBS was independent and he’d used it as an example in the article because he knew the project, having built it…

…and then it just went off the rails completely with the guy repeatedly trashing TBS, Nivenly, anyone involved in the project, and just doubling down every time someone posted contradictory information.

And then he started a new thread summing up the same claims he’d made and completely ignoring anything anyone else had said.

Ro’s posts in the original thread appear to have been deleted, but his first post was a simple “I wrote something on this exact topic you’ve brought up, here’s the link.”

blathering about combining metadata formats on a web page

This article on the minimal markup needed for link previews has got me thinking about consolidating metadata again.

I could go pure #IndieWeb and throw out everything that isn’t visible on the page and marked up with microformats2, but I there’s a lot of stuff out there that doesn’t read microformats2.

I’ve already consolidated most of the <meta> tag-based labels like OpenGraph and older HTML conventions. What I may get rid of:

  • OpenGraph category/tag/etc. details, unless something out there actually uses it.
  • The redundant chunk of JSON-LD for Schema.org

I assume anything that uses Schema.org will fall back to OpenGraph or plain HTML for anything they have in common, but I don’t know whether they’ll still fall back if I keep a JSON-LD chunk with the Schema-specific fields like more detailed article types, what media/event/place is being reviewed, etc.

I wonder if I can add them as microdata where I already have the microformats2 info….

Re: Ephemeral

(In response to Tantek’s post on The Ephemeral Web, which he roughly equates with JS;DR.)

I think there’s a case to be made for deliberately (on the part of the author) ephemeral vs accidentally ephemeral as well. I look through my old social media posts and while some of them are worth preserving, a lot of them aren’t. And of course some people don’t *want* that permanent record. This is where things like Mastodon’s auto-delete feature, and the way you can easily set exceptions to keep some posts online, are useful.

On Wandering.shop

This is also why I still take a PESOS approach a lot of the time: my posts on Mastodon or GoToSocial end up being rough drafts for a more polished article on my website.

>The latest storm was the third “thousand-year” event — one with a 0.1% likelihood of occurring in any given year — to hit So…

The latest storm was the third “thousand-year” event — one with a 0.1% likelihood of occurring in any given year — to hit Southern California this winter.

— LA Times in an article on trying to balance flood control and stormwater capture.

The current flood control system managed to handle 60% of average LA’s annual rainfall hitting in the space of three days. The current stormwater capture system can’t: 80% of that water is heading straight to the ocean. Projects to improving it are in the works, but increasing that capture is only going to get more important.

Note: combine with locally-sourced water and link to water-management

Contemplating adding the etymology…

Contemplating adding the etymology for a bunch of streets that are clearly named after rich American industrialists of the 19th and early 20th centuries (Carnegie, Vanderbilt, etc.), but remembering the computer lab in college where we allegedly named the servers after four Renaissance artists. Perhaps it’s best to confirm that there isn’t another pattern at work.

Apparently a recent-ish issue of The Flash made a reference to the “Flush Man” name that the comic was published under in Arg…

Apparently a recent-ish issue of The Flash made a reference to the “Flush Man” name that the comic was published under in Argentina back in the 1980s

https://screenrant.com/flash-nickname-flushman-argentina-dc-comics/

I should look for the collection of cover scans that an Argentine Flash fan sent me ages ago!

OK, at this point it’s probably safe to say that the contractors working on the neighbors’ balcony aren’t coming today…

OK, at this point it’s probably safe to say that the contractors working on the neighbors’ balcony aren’t coming today. They’re not done – there are boards primed but not painted, and others not even painted, and there’s a small stack of unused boards by the wall. But it’s also raining today.

I guess it makes sense that only the biggest cloud providers would offer preemptible/spot instances, since they’re going to h…

I guess it makes sense that only the biggest cloud providers would offer preemptible/spot instances, since they’re going to have more spare capacity.

Annoying, though.

Azure
AWS
Google
Oracle

#hosting

Also looking at ARM-based cloud offerings. Looks like Oracle, Google, Amazon, Azure

Again, the big ones. Plus Hetzner, but they only have ARM servers in their European datacenters right now.