Software developer, formerly at Microsoft, now co-developer of the AccessKit open-source project (https://accesskit.dev/) and cofounder of Pneuma Solutions ( […]
🌉 bridged from ⁂ https://toot.cafe/@matt, follow @ap.brid.gy to interact

@matt.toot.cafe.ap.brid.gy
Software developer, formerly at Microsoft, now co-developer of the AccessKit open-source project (https://accesskit.dev/) and cofounder of Pneuma Solutions ( […]
🌉 bridged from ⁂ https://toot.cafe/@matt, follow @ap.brid.gy to interact
@WeirdWriter I gather it's much harder for sites that are running WordPress or some other dynamic-language CMS rather than a pre-built static site.
@miki I would think that disabled children in particular are especially prone to bullying everywhere.
@miki On the other hand, that could lead to a child getting stuck with the same bully all the way through school, right?
More context on this: The BrailleSense U2 is a portable PDA with Braille display and keyboard, a note-taker as we call that class of device. It was released in 2014. It has a 1 GHz Cortex-A8 processor, 256 MB of RAM, and 512 MB of NAND flash storage plus SD card. The original software is Windows […]
Like ignoring the news, that may be a luxury I can only afford due to privilege. It seems callous to just ignore bad things that are going on, but then, why weigh myself down with things I can't do anything about, when there are problems I *can* solve by developing more software? Then again, it […]
An obvious application of this statement, particularly in the context of the original quote (see start of thread), is that to have a more positive attitude, which ought to lead to getting more work done, I should avoid reading negative stuff, about AI in particular.
"You are what you read." It's short; it's catchy. I wonder how true it actually is.
Thinking about this comment about an article on current developer attitudes: https://lobste.rs/c/sgvybl responding to the article's statement "I suspect that everybody’s suffering. Some are just better at pretending."
> You can only believe this is true if you spend all your time in communities […]
Question for blind people who routinely read Braille, especially books and other long documents, which I do not: Are there benefits to reading Braille that was manually transcribed by a skilled human transcriber rather than machine-translated? Or is good Braille translation truly mechanical […]
@simon Re: your latest post where you tried having Claude compose music, you should hear what some people are doing with singing speech synthesizers and Claude. https://mastodon.stickbear.me/@Orinks/117386812749307173 I'm not sure how much human audio production was still required for this one.
@glyph Because, as I learned years later, the rule is that focus can be taken from the process that owns the current focused window by a process that was *directly* started by that process. So yeah, the app's executable should just run the app directly, in the original process.
@glyph I had a kind of related problem on Windows back in 2002. The first real desktop app that I shipped, primarily developed by a team that was there before me, was in Java. To give the app its own executable, icon, etc., I wrote a little launcher in C that, instead of embedding the JVM, just […]
@glyph I think I don't have the context to understand what you're getting at here. Yes, how we treat each other is more important than pumping out more software for its own sake. And I appreciate that in the current moment, it's easy to feel that there's already too much bad software, and it […]
So you know how I said I was gonna do a Linux port to the Braille Sense U2?
[ 0.000000] Booting Linux on physical CPU 0x0
[ 0.000000] Linux version 6.12.111-cip32-rt21-u2-storage1 (freya@carbon) (arm-linux-gnueabi-gcc (Debian 14.2.0-19) 14.2.0, GNU ld (GNU Binutils for Debian) 2.44) […]
Pseudo-boosting a post I just read on Bluesky: https://bsky.app/profile/sunshowers.io/post/3mx3fcicrbs2v
Quoting it in full because it's short:
> a lot of people on both the pro and anti-ai-for-coding sides have low standards. no, if you're not paying attention it's probably slop; no, we […]
@glyph Now, if it matters, the UI of FastSMRW was partly derived from a style of social media client UI that was pioneered by a much earlier Twitter client in Python, starting in 2010 (or maybe 2009, I forget now). A handful of apps have been iterating on that genre of social media client since […]
@glyph Most of the commits by far are from one developer, Mason, who, as I understand it, had hobbyist programming experience, largely in Python, before the rise of coding agents. A friend who is in frequent contact with Mason told me that Mason does not know C++. There have been a handful of […]
@glyph OK, I have those usage numbers for FastSMRW.
I sampled a total of 6000 posts, 2000 each from three instances (CaneAndAble.social, DragonsCave.space, Tweesecake.social). 1007 of those posts were from FastSMRW, and those 1007 posts came from 56 posters out of 245 total. FastSMRW is the #2 […]
@glyph The most successful single vibe-coded app that I know of is FastSMRW (https://github.com/masonasons/FastSMRW), a Mastodon and Bluesky client for blind people. C++ core, platform-specific front-ends. Yes, a vibe-coded app in C++ concerns me. I've never used it. "RW" because it was […]
@glyph You're right about the starting assumption of a scalar called "code quality".
As for how many people are using vibe-coded apps, I'm hearing a lot about people *making* vibe-coded apps, especially blind people, but I couldn't give you numbers on how many are *using* them.