I am not personally a fan of it, for a few reasons.

The big one is that it means that all discovery for all uses of a protocol must be uniform across all areas on a single domain. For example, if you have URLs like https://example.com/~alice/ and https://example.com/~bob/ and both of them want to support the foo protocol, then https://example.com/.well-known/foo needs to have some means of distinguishing the two. And the way of doing that can be tricky. Do you have something like https://example.com/.well-known/foo?resource=/~alice/? How do you deal with path normalization (e.g. /~alice vs ~alice/ vs. ~alice/homepage.html)? Do you have to consider things like cross-domain attacks? What if multiple users on a tilde site want to support different sets of protocols, or use different implementations?

Putting things into the query string (the most typical approach for this) also means that you’re going to have to have some sort of dynamic mapping or request routing, which means this can’t work with purely static hosting. It also means that you might have to probe for this protocol support across every single URL being accessed on a domain.

Another concern is that every protocol you want to support requires a separate HTTP request, and these things can add up pretty quickly. For example, whenever I post an article to my website and it goes out on my Mastodon feed, I get many dozens of Mastodon instances each probing an absolute litany of related URLs trying to determine whether this is a Mastodon instance or similar, and sometimes this ends up even overwhelming my server. Even without that, I’m getting a constant flood of requests for things like /.well-known/traffic-advice and /.well-known/wp-login.php and the like.

A much better approach, in my opinion, is to have the discovery baked into the resource itself. In HTML you can use the <link> tag (e.g. <link rel="foo" href="/~alice/foo-support.xml">)¹ and in HTTP in general you can use the Link: header (e.g. Link: <https://example.com/%7Ealice/foo-support.xml>; rel=foo). A single HTTP request can tell a client about all possibly-supported protocols, all at once, and it can be baked in statically and thus supports static hosting. It also means there is no special case for handling multiple resources across a single domain, and if two pages have the exact same link target it can also be inferred that they are the same as far as that protocol is concerned.

One concern that comes up is that of page bloat, when every page needs to include a bunch of <link> tags to express a growing list of supported protocols. This is, to me, a non-issue; for starters, you’re only declaring the protocols that you actually support, and it’s a single point of reference for all clients to discover all of the supported protocols. But also, the amount of bandwidth added to a page for even a few dozen protocols is miniscule compared to the amount of bandwidth taken by other accepted optimizations, such as inline CSS and data: blobs for images, as well as the bandwidth taken up by the incessant .well-known probes that are taking place at this point.

A suite of related protocols could also be offered as a forwarding URL, such as IndieAuth’s current practice of bundling it all together into indieauth-metadata. This does require an extra HTTP request, but it only has to happen once, as that URL is extremely cacheable.

The IndieWeb wiki has more to say about the growing usage of .well-known and why it’s considered an antipattern.

As an aside, I am also not a fan of Webfinger, because not only does it require .well-known to work, but it attempts to flatten a namespace in ways that are difficult to deal with. On this website, you can easily get the update feed for any given section by discovering its rel="alternate" URLs, but there is no such mechanism in Webfinger; you can follow the site as a whole as @beesbuzz.biz@beesbuzz.biz (thanks to Bridgy Fed) but you can’t follow just the /code/ section, for example. There are some hacks such as making @code@beesbuzz.biz, @comics@beesbuzz.biz and so on, but then what about nested subdirectories (e.g. /food/coffee/)? Why not just use the URL itself as the specifier? With <link>-based discovery, you already can, and there’s nothing special about it.

So, anyway: When designing a new protocol, please consider not using .well-known URLs for discovery purposes. Let the URL itself provide its own information.

1. In some cases, such as content feeds, it can be a bit unclear as to whether you should use an existing rel (such as alternate) with an appropriate type to disambiguate, or if you should invent a new rel, but that’s a much smaller problem than those caused by .well-known creep. ↩

   The general rule of thumb seems to be that if you’re designing an active interface-type protocol (e.g. something with a specific RESTful API), you should invent a new rel, and if you’re designing a feed format or other such alternate representation of a page’s content you should use rel="alternate", although that does require some interpolation of the MDN Link Types document and the general common practices across the IndieWeb and Microformats communities.

comments

Several months ago I realized I was feeling like I did back when I was off estrogen entirely because of the embolism and got my levels checked for the first time in a while. And they were dangerously low.

My doctor put me on a higher dose of transdermal estradiol and wanted to wait 3 months to follow up. I finally got that followup last week, and my levels are still dangerously low.

So, tomorrow we’re going to discuss the possibility of injections, which is honestly what I should have probably been on to begin with. But, y'know, doctors love to be super cautious with clotting risks on estradiol, especially with a family history. (I also never really had a reasonable dose when I was doing oral estradiol, either, even before the embolism, because I am bad at advocating for myself.)

Hopefully I’ll get on injections and that will help me start to feel normal again, and maybe I’ll actually get a decent amount of breast growth out of it this time (unlikely).

I feel like I’ve been operating well below my baseline for the last… I dunno. 9 months or so? I remember still doing pretty well as far back as, like, last June, and I don’t really know when my activity levels started to really decline because I haven’t been good at keeping track of that stuff. I do use a daily mood tracker thing and it looks like the last time I marked a day as feeling “good” was last July. So, possibly that’s when this hormone stuff started happening. Looking at my blog I definitely noticed a turn for the worse back in August and also looking at July I was starting to feel a bit off. So July is probably when things started to go haywire.

I’m definitely motivated to do a bunch of stuff right now; I want to build my Nerdy Gurdy kit which I was finally able to obtain after a few years of interest and a few weeks of deliberate intent, and I really want to work on my various albums that I’ve had on the back burner. I did start to work on Misfits a bit but I haven’t managed to sustain the energy so I’ve made very little progress on it. I also have three other albums on the back burner which I’d love to be working on as well. Plus finishing up the music for the game! And a couple of other low-priority commissions I’ve gotten!

And there’s so many projects I need to do around the house, especially a few things in the basement which will make the studio more accessible. And getting my kitchen organized. And getting a bunch of stuff I no longer need/want to Goodwill or whatever. A bunch of it has been sitting in my car’s trunk for months, now, but I also don’t feel comfortable driving down to Burien, both with my fatigue and with my car still needing a brake job.

Most of what little energy I can muster has gone to either going back to choir practice (which I’d missed so much!) or doing my various VR performances. Several of my recent VR gigs have been hour-long sets and that’s just been completely too much, man¹, even though previously I was able to do 90 minutes no problem. Even 30 minutes feels like a stretch nowadays. So there’s definitely room for improvement.

It’s also been supremely frustrating that all of the dispensaries in my town have closed aside from Uncle Ike’s, which only stocks high-THC stuff and what I need is high-CBD (and what few high-CBD products they have still have much higher doses of THC than I’d like). They seem to only care about getting wasted, and not getting pain relief. There are a couple of smoke shops around which claim to carry CBD products so maybe I’ll check some of them out, but smoke shops always feel so shady, compared to the nicely regulated cannabis industry around here.

I also can’t help but wonder if the other dispensaries closed due to pressure from Ike’s, as he has a long history of anticompetitive behavior.

Now that it’s getting warmer I want to be gardening, and I want to be working on music and my coding projects (especially the music-related ones), and I want to be doing so many other things, but even walking to the grocery store ends up wiping out my energy reserves for the rest of the day.

Anyway. Hopefully getting my hormone levels back in check will help.

1. As an aside, I’ve been watching the full series of Bewitched lately, and while I’m glad they haven’t tried remaking the show (poorly-rated Will Ferrell movie aside), I think Kristen Schaal would make an incredible choice for Gladys Kravitz. ↩

comments

My pain and fatigue have been especially bad lately, and that also feeds into my lack of patience and temper, and also is probably due in no small part to the fact that my hormones are, as mentioned, out of whack.

I’m also somewhat looking for work (music ain’t paying the bills) but it’s hard to find something that I’m capable of doing and for the limited hours with which I can do them. If folks know of anyone looking for someone who can do any of the many things I can do, could you maybe point them to my resume? Thanks!

comments

So, Cloudflare wasn’t even really doing anything for me anyway, and certainly wasn’t worth the problems it caused (such as being subject to its many outages and having my site become yet another source of privacy-destroying analytics and so on).

In other news, the crawlers mostly seem to have gotten wise to my tarpit and now it’s only averaging around 20 requests per second, down from a peak of 310. Oh well.

At some point I think I’ll replace the forced login thing with a simpler sentience check since the login thing feels a bit aggressive, although I’m not sure I can easily fit that into Publ without modifying Publ itself due to vagaries of how Flask routing works. It’d probably be nice to just make antibot measures part of Publ anyway, though. Or maybe I can abuse error 429 for this, which would probably be a better choice than error 401 anyway.

comments

Cloudflare was also super aggressive about caching some stuff that I didn’t want to be cached, and of course, there are many, many political and ideological reasons to not want to use Cloudflare. So my plan was always to switch back to not being under Cloudflare, but the longer I waited the harder it seemed like it would be, due to how SSL certificates work. In particular, I use wildcard Let’s Encrypt certificates, which require DNS to be current, and a big thing that Cloudflare does is… take over your DNS.

But tonight I got a hair up my butt and switched back to my own termination, and it wasn’t too hard to do, with just a little bit of DNS and TLS juggling, and I wanted to minimize my website downtime.

The steps I took were roughly:

1. Added some temporary A records on Linode’s DNS, mapping @ and * to one of Cloudflare’s exit servers
2. Added some propagation-debugging TXT records (_moo.beesbuzz.biz and _moo.sockpuppet.band in case you’re curious)
3. Swapped my SOA records back over to Linode
4. Waited for dig TXT _moo.beesbuzz.biz and dig TXT _moo.sockpuppet.band to return the expected value
5. Reenabled certbot on the domains and waited for it to get updated certificates
6. Flipped the TLS certificates in nginx over to Let’s Encrypt
7. Flipped @ and * back over to my actual origin server

Thankfully, I had basically no site downtime, since Cloudflare was accepting my LE certificates on the origin (rather than requiring the use of their self-signed origin certificate). I did have a bit of delay on getting the updated cert on sockpuppet.band due to DNS propagation¹ being slightly slower on that one for some reason and I got impatient, but it only took a few minutes to catch up.

So anyway, yeah, I’m now Cloudflare-free. No more creepy injected analytics spyware, no more contributing to the erosion of the decentralized web, and no more having to worry about Cloudflare suddenly deciding I needed an enterprise account instead of being fine with the free tier. My site is still performing just as well as before, as far as I can tell.

At present I’m not even seeing a significantly higher request rate on my server, although presumably it’ll be some time before the AI crawlers have caught up to the new DNS records, because I’m sure they’re doing their own broken caching like everything else they do is broken, so maybe in a day or so I’ll be regretting this choice. Also I’m sure there’s still plenty of people actually going through Cloudflare because of things like DoH caching, so it’ll probably be a day or so before I get the real performance metrics. But according to Cloudflare’s own dashboard it was only caching about 2.5% of my overall requests.

The domains are still configured on Cloudflare for now, but I’m sure at some point they’re going to notice that the zone isn’t mapped to their DNS anymore and so at some point any stragglers who are still trying to use Cloudflare’s exit server are going to get some sort of error message. And I’m totally fine with that.

One way to tell if your browser is still going through Cloudflare is by visiting my robots.txt, which Cloudflare also adds a bunch of crap to. Or you can view source on this page; if there’s a <script> tag referencing static.cloudflareinsights.com at the very bottom then you’re still going through Cloudflare. For me at this very moment, Firefox is still going through Cloudflare to beesbuzz.biz but not to sockpuppet.band, and Safari and curl are going direct to my own server for both domains with no TLS problems, so I’m satisfied with the current state of things.

1. This is a total misnomer, incidentally; there’s nothing that really “propagates” so much as there’s some caches that need to expire. ↩

comments

As usual I’ve been full of pain. And my brief surge of enthusiasm for working on music got cut kind of short because of it. But I’m hoping that tomorrow feels better and I can go back to the studio.

I’m also not sure where I stand with my fatigue. The last few days I’ve had to run quick errands by car and those felt fine. But I’m not feeling courageous enough to drive further just yet, especially with how any escape from White Center means following curvy roads which are especially triggering to me.

Software engineering

I’ve been applying to jobs again. So far it’s difficult finding jobs that I’d be willing to do, and of those, I either get ghosted or an immediate rejection. My resume is currently a mess though. I should probably have separate resumes for media shit vs. programming shit. Not that I really want to work on programming shit anymore.

And on that note, I’m getting very weary of how much time I spend working on software projects that end up not going anywhere. I feel like the entire state of programming has gotten to be stupidly overcomplex and like I’m shouting into the wind when it comes to simplicity of design and building things based around extensibility and interop rather than coming out of the gate with being an Everything Solution. And then when I try to explain more fully what I’d like to see as next steps, the response I get is, “Oh, well, I hope to see what you make.” When the whole reason I’m trying to collaborate with others is that I only have very limited time/energy/pain budget to be able to work on code. I just can’t do it all by myself, and even when I do do it all by myself, I can’t seem to get anyone else to care about actually using the damn thing.

(So, no, I haven’t gotten much forward momentum on Canimus, for example.)

There’s so many things I’d like to be doing but so little energy with which to do it. I’m tired, y'all.

I definitely see the appeal to AI coding agents. I would not feel good about using one to do the things that I want to do, and I wouldn’t trust them to build things even remotely how I’d want them to be built. But I totally get why others do it.

Side tangent about cars

The shitbox Corolla has been doing what I need it to do, but I keep on running into little bitty problems with it. Like the other day I got stuck behind someone who was just, like, waiting at a four-way stop for several minutes, and that was a great moment to learn that the horn doesn’t work. (And then the person behind me started honking wildly and then got super road-ragey. So that was fun.) So that’s another thing to have to investigate. Because I don’t honk often but when I need to it’s with good reason.

And of course I still need to get the brakes done.

Even though I barely drive (usually 2-3 miles a week on average) I’m still always inclined to look at the current state of EVs. At present this is the entire list of cars I’d consider buying new, if I had to:

• 2026 Nissan LEAF

If it comes to recommending a car for someone else to consider I’d also add the 2026 Toyta bZ and the 2026 Subaru Solterra (both have been fully refreshed and are much more interesting vehicles now, and they’ve diverged enough that there’s relative pluses and minuses to both of them), but I’d rather have something more car-like, and the LEAF, despite being marketed as an SUV, is still basically a hatchback/fastback, and is actually even more compact than the old one.

Telo and Slate are also super interesting but neither is even close to coming to market, and Telo’s expected price is pretty high for what you get, and Slate looks super cool as a work truck but that’s not the sort of vehicle I would find useful.

But of course the whole reason I went shitbox is because I don’t drive enough to justify the added expenses of owning an EV and pretty much everything I actually use a car for is better served by an ebike anyway, aside from the occasional thing where I need to haul cargo and kitties, and in the case of cargo it’s cheaper overall to just order delivery, and in the case of kitties, well, the shitbox suffices (and so would Lyft in a pinch).

Now that the weather is getting nicer I can start thinking about riding my ebike again, at least.

Cannabis blues

The only cannabis dispensary within walking distance of me anymore is Uncle Ike’s, which I do not like frequenting; ethically-speaking, they have a long history of predatory, anti-competitive behavior, and anyway I need CBD-dominant products and they only seem to carry high-THC get-fucking-wasted shit.

There’s some dispensaries all the way across town but getting to them is a bit trickier, because the local transit is awful, driving there would likely be an anxiety or fatigue trigger, and biking doesn’t feel safe since that part of town is very stroady. And even then their CBD products also seem lacking; one of them has their menu online and the only CBD stuff they carry is 1:1 (which is still “get fucked up” territory for me) and the other one’s online menu is completely broken (and their website looks like it’s been hacked by spammers) so who even knows what their deal is.

Aspirations

I’d really like to feel safe driving again, just so that I can expand my range out. And if that does happen maybe I’d feel justified in getting a better car.

I’d also like to feel good about making music again. Right now it just all feels so empty to me and I’m having trouble finding the motivation to do stuff.

I’d like to feel like I’m not just taking up space.

comments

The entire world of “tech” has been broken for ages, and now it’s falling apart. The narrative about Amazon laying off so many people because their jobs can be replaced by AI is just a smokescreen; in reality, the software industry has been super high on its own farts for way too long and the end result of unsustainable infinite growth is just finally coming to pass.

Personally I’m also super tired. I keep looking to see what sorts of jobs I can do and I just can’t do any of them. I can barely take care of myself throughout the day, and I’ve had little to no energy for making music aside from doing 2-3 shows in VRChat a month, and that kinda wipes me out. I have a few musical projects I’m kinda-sorta working on but not really able to put as much attention and care into them that I’d like.

I also feel like my actual housing/living situation isn’t where I need to be. I miss living in a city and where things are walkable and transit-accessible, but I need my recording studio and sonic isolation. I wish I had the funds to just like. Build a house that’s designed the way I want it to be. I’ve been making do with this house but there’s so many problems with it.

I’ve been having a very short fuse with a lot of things lately. Especially my cats. Tyler’s developed some bad behavioral problems, and Fiona is refusing to eat dry food anymore but is constantly begging to be fed. And technology keeps on failing around me and I get so very frustrated by stuff just not working right.

I also hate how isolated I am. Fortunately, choir is starting back up this week and that at least gives me a reason to actually see people in person for a change and have people to talk to. VRChat doesn’t really do it for me anymore, and in daily life I only ever even go outside to go to the grocery store and everything’s self-service these days and it’s like there’s just dozens of people who are alone together but never actually communicating. Pretty much the only time I ever have to even talk to anyone is to pick up a prescription and that’s not exactly a basis for conversation.

I’ve been kind of spending a lot of time playing Animal Crossing again, which feels like a symptom. Also spending way too much of my day “watching” YouTube, although at least I’m getting better about listening to music instead, which at least feels better.

I’m also still feeling the loss of a couple of previously-important online communities due to personality conflicts that just didn’t feel worth resolving. I don’t think those communities were particularly good for me but they were at least something routine and comfortable in the moment. At least I’m still in touch with some individuals from those communities, but I still feel like I’m missing so much.

As is often the case lately, I feel like I’m just taking up space, and if I were to die in my sleep it would be ages before anyone would find out. Maybe people would notice my absence from my various regular meetings but I don’t know if anyone would know whether to send help, and where to send it to. I have at least been using Wellness Ping as a daily checkin thing so that certain people close to me will be notified if I need to be checked on (and I’m glad this service exists now, as it’s something I’d been wanting to build for a while but never had the spoons to) but it would still be days before anyone gets notified, and I don’t know if those people would know who else to notify in case something happens that needs notification.

Right now the Internet feels awful but the Internet is the only place I can really exist. Maybe I should spend more time just like, riding the bus into town and trying to find things to do, but then issues of practicality just get in the way. Like I can’t just spend all day hanging out in coffee shops and still being isolated-in-a-crowd.

I miss all the community I used to have, and I just feel so much like everything went to shit in 2020 and that something’s broken and can never be fixed.

I keep feeling like the best situation for me would be to live in some sort of small village where everyone knows each other and is willing to help each other out and ask each other for help. Maybe that’s why I’m so attached to Animal Crossing these days.

I know from experience that I need my own dwelling space, like, roomates are absolutely not the answer. But having an actual community would be nice.

I don’t have a good through-line for this ramble. Just some stuff that’s weighing heavily on me.

comments

The previous denial came with a whole bunch of transphobia and also a claim that I am able to work, doing jobs which simply aren’t available and not reflective of the reality of my situation.

I haven’t received the formal denial yet, just the notification that my appeal was denied, so I suppose in a couple weeks I’ll be getting another giant packet from SSA to fume over.

In the previous denial they cited two things, namely:

1. that I successfully underwent HRT for transgender transition (as if that has anything to do with disability)
2. that I am good at managing my money and therefore don’t really need it

Point 1 is just plain asinine and I don’t think it needs any further explanation.

Point 2 is also asinine, because like, yes my finances can hold out for many years (especially if I continue to live minimally/frugally), assuming the economy doesn’t shit itself and wipe out my savings (which is super likely), but what I’m seeking isn’t the money so much as the ability to access Washington State social services that are all gated behind a disability determination, which is then linked to receiving Social Security.

If I were to pay for those same services out-of-pocket (which I have been doing on occasion but am trying to avoid needing more of), I certainly wouldn’t be financially solvent for very long. Because hiring help is expensive.

I have great difficulty working more than an hour or two a day, on average, and nobody would want to hire me part-time if that means only working at best 10 hours a week, with extremely unpredictable availability.

I’m making basically nothing on my music right now, nobody wants to hire me for the kinds of software I do (at least, not at the rate at which I’m able to do it), and I’m in pain — often excruciating — most of the time. When I’m not in pain I’m too fatigued to do anything. I try to get exercise when I can but if I overdo it just even a little tiny bit then the next few days I can barely function at the most basic level.

I’m also too fatigued to feel safe driving. Like, I’ll go to the grocery store a mile away and that’s about it. I’ve driven my car all of 12 miles over the past month.

I’m also reaching the point where even if I were to be declared disabled I’m no longer eligible for social security income anyway; the requirement is that you’ve worked 5 of the past 10 years, and freelancing doesn’t count. So that’s:

• 11 months of HBO (February 2016–January 2017)
• 10 months of UW (December 2018–October 2019)
• 30 months of shitty tech jobs (January 2020–June 2022)

So, 51 months over the past 10 years, when the minimum is 60.

In theory since I filed for disability sometime in 2023 it should be backdated to 2013 (and if I were to be granted disability I’d be owed backpay back to when I first filed for it), but my understanding is that it doesn’t actually work that way.

So basically, I’m pretty much screwed on the social security front. And if I make it to 65, I probably won’t qualify for it then, either, because my work credits will have been long gone.

And yes, I have been working with an attorney on this. I have to mention that because every single fucking time I talk about a social security denial people tell me I should be working with an attorney. The attorney is also baffled as to why I keep getting denied.

But even more frustratingly, it’s difficult to see a future for myself just because I’m in so much pain and always so tired and having difficulty even taking care of the basics of my life, much less all the things I want to be doing.

Whatever happens I’m going to run it all into the ground and keep going as long as I can, but holy crap is it hard to see a future right now.

comments

I did finally take the Corolla to the mechanic and got it inspected and some overdue maintenance. They said that the brakes have been badly-neglected and it’ll need a $650 brake job. This doesn’t surprise me at all, and I was planning on needing to put a bunch of money into it to get it safely drivable. The purchase price was significantly lower than the KBB value and I figured there was some sort of “round tuit” thing that the previous owner was putting off that was his last straw for upgrading vehicles. Totally fair.

Even with the cost of this brake job and the other upgrades I’ve put into the car, I’ve still spent less on it than it would have been to get the Niro operational again, and I also got a bunch of cash and a nice ebike out of the deal. I do still resent having to own a car at all though. Maybe in a few years there will be cheap third-gen LEAFs available on the used market and I can go back to an EV.

I’ve been super fatigued lately and having trouble with actually doing things, but today I at least spent some time rehearsing for my weekend shows, and also finally figured out a better way to do soundchecks which will make future things a lot easier.

I also keep finding that 60-minute sets are way too long. 30-minute sets are a bit too short, though. It’d be cool if VR shows would split the difference with 45-minute sets which would be Just Right. Maybe that’s something I’ll mention to the folks who book me most often since they’re also the ones running most of these shows.

I should also re-document my live show setup for the blogpuppet since it’s changed significantly since the last time. UPDATE: done

Anyway. I’m still feeling super fatigued and I would really like to figure out what’s going on and, more importantly, how to fix it. It’d be super cool if things like fibromyalgia and ME/CFS got proper study or even had things like objective tests that weren’t based purely on Vibes and personal experience, especially since those very things make it super difficult to get help.

comments

Then Google got into it and made Google Reader, which was a really good feed reader, but they couldn’t figure out how to make it profitable and wanted to shove everyone over to Google+, so they shut Google Reader down, and then the tech press breathlessly claimed that RSS was now dead as a result, and because RSS was (in their eyes) dead, it became a self-fulfilling prophecy. Nowadays, RSS is seen as an obsolete thing only used by diehards, and corporate social media has become the norm, with a relatively tiny faction of people settling on ActivityPub instead, despite it being a poor match for the kinds of things that people used RSS for.

And now the same goddamn thing is happening with Virtual Reality. It was a thriving enthusiast space taken up by people who had a lot of fun seeing what they could build, and making amazing games and social environments and so on in a very DIY-friendly way. And then Facebook decided they needed to get in on that and acquired so many of the manufacturers and game producers involved in the space, pivoted hard to trying to make corporate metaverse bullshit a thing even though literally nobody wants it (Meta Horizons is basically Zoom with extra steps and even less whimsy), and shoveled so much money at it.

Now they’re realizing they have no idea how to make it profitable and are shutting down all of the companies that were making VR great, including studios that were actually profitable but just not profitable enough to offset their misplaced ambitions.

And the tech press is now declaring VR dead, meaning that those of us who are continuing to thrive in virtual spaces using enthusiast hardware and so on are likely to be left in the lurch at some point.

Thankfully, Valve and ByteDance are keeping the hardware side of things alive in ways that are actually along the lines of what people want, but I still worry that “VR is now dead” narratives will cause people at those companies to re-evaluate that as well.

VR has been super important and helpful to me, both as a multiply-disabled independent musician and as someone who cares deeply about the trans community. Hopefully VRChat survives, and hopefully there continue to be hardware manufacturers who care about VR without caring about it being hyper-profitable, but I’m already very worried about the lasting damage that will happen from this. (And I was worried when Meta decided to get into VR to begin with, because I very much saw this situation coming!)

As always, I am frustrated and horrified at how modern capitalism doesn’t allow something to just be good and sustainably profitable, it has to get All The Money or else it’s seen as a failure. This happens over and over and over again and I’m so sick of it.

comments

• Schedule a sleep study
• File my quarterly business taxes
• Edit background video for Saturday talk
• Set up projection mapping for Sunday show
• Start working on specification for game dynamic soundtrack engine
• Take care of the major clutter piles in the office
• Empty out the car at the nearest Goodwill
• Take the car to the mechanic for servicing

And some bonus things to take care of:

• Edit and post the video from Gilly’s Cave show
• Edit and post the video from this week’s Moonlit performances

comments

First of all, to be clear, I am not in favor of LLM-based programming. My personal opinion is that LLMs are catastrophic to the environment and to the Internet as a whole (especially due to the added burden of every server being overwhelmed by the relentless AI-focused crawlers), and to the intelligence of the humans who wield them.

However, in the current landscape, I have also found myself (begrudgingly) using them on occasion to a limited degree, and in the mind of some folks, this probably irrevocably taints my code. So I’d like to at least discuss my perspective on them.

There are two specific places where I’ve substantially benefited from the use of LLMs in my work: building FFmpeg visualizers for my music YouTube channel, and building a bulk updater/scheduler tool for the same. Neither are uses I’m happy about, but I’d like to discuss them a bit further.

FFmpeg visualizers

FFmpeg’s filtergraph functionality is pretty comprehensive and powerful. It’s also poorly-documented, with a bunch of useless/trivial examples, very little actual syntax documentation, and no good explanations of how filter chains actually connect together.

There are some online tools such as FFmpeg Explorer which can help, but it’s also a bit cumbersome to use, requires a lot of trial-and-error, often results in incomprehensible errors, and also needs you to know what you’re looking for in the first place while dealing with FFmpeg’s rather inconsistent naming scheme for things.

However, in one of my searches on DuckDuckGo (which I’ve returned to using despite their hard pivot towards AI, because the alternatives are worse), I hadn’t yet disabled the duck.ai search helper, and I happened to look at its response, and it was actually… somewhat helpful. But not helpful enough.

So I asked ChatGPT.

ChatGPT gave me somewhat better explanations of things. It was sometimes incorrect, and often completely made up functionality that FFmpeg doesn’t have. It was completely sycophantic (if not downright patronizing) when describing my thought processes back to me, and I wanted to stab it with a fork. But it got me to where I was actually pretty good at building complex filter graphs on my own. Now I do not need to consult it at all.

I was absolutely not “vibe coding” with it, but it gave me a much better understanding of the system, and worked better than the official documentation or anything I could find online.

In this case, AI wasn’t good, so much as the human-written documentation was bad.

yt-updater

At one time, the YouTube data API used to be very well-documented, with pretty good example code provided in several languages. But the documentation quality has gone way downhill, and their search function has been replaced with an “ask AI” chatbot powered by Gemini.

I could not make heads or tails of how to write even the most basic code to log in to the API and perform some basic tasks. But the Gemini-powered manual did a decent job of spewing out some example code that I could work with and adapt.

I never copy-pasted the code directly, but I did read it and comprehended it, and a process which had been causing me to pull my hair out¹ turned into one where I could at least get on my way and get shit done.

Even Gemini, purpose-trained on the docs, still sent me in the wrong direction on a few things, and didn’t write super great code. Some of its examples were incorrect, but the code I tried ended up burning through my API quota very quickly and I had to wait some time before I could try more things. It did an awful job of warning me about things to watch out for which I only learned in retrospect. It also hallucinated functionality, based on old docs for long-deprecated APIs, and when I asked about that, it apologized profusely while subtly shifting the blame to me for not already knowing better, when, my dude, you’re the one who recommended it to me.

But still, I have working software which helps me achieve my actual task: bulk-updating video playlists with templatized descriptions and a posting schedule.

Again, I do not think this is a win for LLMs, so much as a loss for good documentation.

Other things

Sometimes I need to look up a Python function that I’ve forgotten the name or containing library for. I could try doing a billion search queries and wade through poorly-written docs which are scattered across many different repositories. Or I can look at the duck.ai summary which usually tells me the exact thing I’m looking for.

Sometimes I need to evaluate libraries for certain purposes, and search results are difficult to wade through. The AI summary usually presents a couple of options that I can investigate more thoroughly and make a decision on. (But sometimes it completely fails to interpret the question right and gives me completely useless answers as well.)

I also have non-programmer friends who have been making use of Claude and similar to write code that helps them get their things done, without having to find a skilled developer who is passionate enough to help who they can also afford the services of. These friends do usually end up realizing that the AI code is limited and only gets them part of the way there, or that the code isn’t the best way of doing things… but it gets the job done for them, and as the saying goes, perfect is the enemy of done.

The big question

So my question is, since I have benefited from the use of AI in some of my projects, does that mean my projects are irrevocably tainted? Should I be put on the open slopware list as a result²? EDIT: It looks like the “open slopware” repository has been deleted.

In my career I always rejected pretty much all forms of code autocompletion. I absolutely derided things like IntelliSense and XCode’s API autocompletion stuff because I felt like it made it too easy to just write code without having to think about it or understand it, because it would just offer a simple solution and then people would fill in the madlibs on it. I felt like if you hadn’t internalized an API then you didn’t really know it, and that you should try to remember the context of what you’re doing.

I also find the autocomplete things to be super distracting and that they usually present the wrong thing. Copilot was even worse about that, and I’d find it would actively take me out of my flow state, which is super disruptive to an ADHD brain like mine. I couldn’t understand how people would even want to work with those things.

Since then I’ve relaxed my stance a little bit, in that I use Sublime Text’s context-aware keyword autosuggest thing. It isn’t AI in any sense, it’s just something like “oh, you’re starting to type the name of a symbol that I have a few matches for, here’s some suggestions” and I find that actually useful. I also don’t mind its automatic punctuation insertion most of the time, since it helps me with things like parenthesis/bracket balancing (although sometimes it can go a bit overboard, especially for quotes and such, which it often gets wrong).

But seeing the kinds of code being written by supposedly senior engineers who have bitten the apple of LLM-based vibe coding, I have reaffirmed my desire to never write code with an LLM directly.

But I also don’t mind so much using LLMs as a supplement for documentation, when the documentation is awful.

But I really wish documentation weren’t so awful.

Where I’m really going with this: Good documentation practices

Back when I primarily wrote C++, my typical approach to building an API would be to start by writing the header files first, complete with documentation (via Doxygen) and sample code. This would help me to hammer out issues with how it would be used and would inform the underlying implementation. To me, this made the separation of .h and .cpp files a strength of C++, not a weakness.

Like, to me, the header file is documentation, and makes a nice interface between the machine-readable parts and the human-readable parts.

As a policy I would always insist on not just having good test coverage, but good documentation coverage. Any code which didn’t have at least a high-level description of what it does and a full explanation of what each function parameter means would fail my criteria.

Some people see this as “gatekeeping,” but I see this as “basic hygiene.”

As I’ve left the software industry and have mostly been writing software for my own ends, I’ve relaxed on this somewhat, but then whenever I have to consult my own code at a later date I end up being confused. So I at least try to do a decent job of documenting my (mostly Python) code with its relevant markup; thankfully, pydoc is built-in to the language itself and it’s more or less standardized how to actually document function parameters, between using docstrings to describe the function and type annotations to describe the function arguments. (Regrettably, the docstrings are typically formatted using RST which I don’t particularly care for the syntax of, but that’s what Sphinx³, and therefore readthedocs, uses.)

One of the more baffling things I ran into in my career was engineers insisting that good docs don’t matter, good code matters; if you can read the code then you can understand the code, and if you write separate documentation, it’ll fall out of sync. I find this mindset to be quite troubling.

To me, the documentation is what people are using, and the code is just how you instruct the compiler how to fill in what the documentation is saying. If the documentation and code differ, then they need to be brought into sync. And if the documentation doesn’t live alongside the code and makes keeping them in sync a lot easier⁴, that is the problem.

LLMs are, to me, an unfortunately-useful tool in filling in a documentation gap. But that’s not a win for LLMs. It’s a loss for documentation, and everyone who needs it: the sentient beings trying to get shit done.

And I hope this doesn’t mean that my code is forever tainted.

1. In a very real sense, this is an accessibility concern for me; I only have a limited pain budget for doing work due to my chronic pain disability, and things which cause frustration eat into that budget much more quickly than other things. When I try working from the docs myself, I’m in excruciating pain and literally want to kill myself. When I have helpful guidance in the form of good documentation — even generated by LLM — I can complete my task and take a nap and feel better. ↩

2. Incidentally, I kind of love this self-reporting commit just out of the sheer levels of malicious compliance on display. Or at least I hope that’s what it is; from context it appears to be. ↩

3. Update: After I posted this article the lovely Piper Thunstrom pointed me to MyST, which allows the use of Markdown with Sphinx. I’ll definitely be looking into that! ↩

4. Incidentally, this is one of my biggest pain points with developing, and even using, Publ, which I kind of chalk up to Publ being my first major piece of Python code and I was still learning how Python works and what the ecosystem is like. Someday it needs a major refactor if not rewrite, with an emphasis on testing and documentation, but the state of the code at present doesn’t lend itself to incremental progress along those lines. ↩

comments

So, Publ finally implements multiprocess-safe token storage, which is something I’d been meaning to add since approximately forever ago (although it seems I never opened an actual issue to track it, oops), and this should make user login reliable again.

Thanks to Spud who finally lit a fire under my butt to fix this annoying issue.

I am, incidentally, doing okay, but frustrated at the state of the world, especially because of how shitty the Internet’s gotten and how annoying it is to maintain software projects in light of the utter brokenness of a world of vibe-coded nonsense and the non-stop onslaught of AI in seemingly every facet of my life these days.

I’m not strictly anti-AI (and I even see some, albeit limited, value in AI-assisted programming these days) but holy crap do certain things make it hard for me to want to approve of it. Pluribus only scratches the surface by analogy.

But at least I got to spend some time with some choir folks tonight and that felt good, and I’m making some progress on some projects and that feels good too.

I think I want to finally learn Rust, and I also want to finally build a player for Canimus, and maybe those two things can go together. The bulk of it is just going to be some database ETL stuff and I don’t see Rust being particularly well-suited for that, and I was looking at media metadata parsing libraries for the private collection server but none of them are even remotely as robust as mutagen either, so I think I’m gonna stick with Python for that part as well.

Alllllthough, Faircamp is written in Rust, and I was planning on contributing Canimus publishing support to Faircamp……… okay, I have my excuse now. Neat.

comments