CSS and design systems people - BlogFlock

Reflections from TPAC 2026 - Hidde's blog

2026-11-06T00:00:00.000Z

TPAC is W3C's yearly work week. A large part of the Working Groups and all sorts of committees get together to collaborate in person (mostly hybrid). This post has some things I got out of it this year.

Some of my week had been slightly overbooked, as is the faith of most TPAC attendees. There is so much going on! Like last year, FOMO was had.

Accessibility

I spent some time with the Accessibility Guidelines Working Group (AGWG), which is primarily focused on developing WCAG 3. I missed the retrospective, but particularly liked the discussions on the impact of accessibility supported (it is not exact science, but has helped move on from “accessible in theory” to “accessible in actual people's browsers/devices/AT”), and how we can write and scope supplemental-to-WCAG guidance for policy makers and regulators.

I also attended some of the ATAG Community Group, which is getting ready to charter as a Working Group. I am particularly excited about that, as I strongly believe we can and should use CMSes and other authoring tools as a way to achieve accessibility at scale. Your CMS is an accessibility assistant), we can find opportunities to fix issues while we create them. ATAG is being modernised, and the next version will address how people work today: sometimes that's content creation with AI (maybe not great for readers?), content improvement with AI (yay fancy spell checkers and tone of voice improvers), sometimes it is doing accessibility better (yay suggested text alternatives or captions).

W3C and Policy

Following on from last year's successful policy lunch, this TPAC Tara organised a session specifically for policy makers. It was held under Chatham House Rule, and closed door. I gave a short talk about how active standards participation benefits (my team at) the Dutch government. Great questions were asked and discussed.

AI and society

One central theme at TPAC 2026 was the intersection of Web and AI. Broad consensus emerged to form a new Agentic Web Working Group and there were lots of breakout sessions on AI.

Me, Tzviya and Christian ran the “AI and Society mini workshop”: 9 short talks + discussions spread over 2 hours. In it, we looked at AI and the web through in the context of the Ethical Web Principles and the W3C Vision. We're working on a proper report about it. I loved hosting this and hearing all these different angles. I was a little sad we didn't get to use airhorner.com, as we had severely threatened all speakers that we would, were they to exceed their allotted 5 minute slot.

An AI policy for W3C

Last year, me and Elena wrote Use of LLMs in standards work, which the Advisory Board (AB) published as its first “position paper”. It wasn't perfect, but it was nice to publicly share a reflection of the AB's thinking at that point in time. We got useful feedback and ideas.

This year, the W3C Team build onto our paper to create a policy on the use of AI tools. We discussed it during a breakout at TPAC. Personally, I was thrilled with Team's work: it nicely removed some of the imperfections of our position paper, and turned our “advice” into something only Team can do: an actual policy for the community. Thoughtful, succinct, and reasonable, I hope it will get adopted before too long.

Furthering human rights

Me and Dan ran a session on standards and human rights (the ones defined in the Universal Declaration, that is very widely adopted). This session was to share and get feedback about our AB position paper Further Aligning W3C's Program of Technical Work with Human Rights. In that paper, we tried to put the OHCHR report on human rights and technical standards into practical W3C-relevant recommendations (see slides).

Since that position paper, a community group was formed (you can join it if you want to help us move forward), and it was great to get feedback at TPAC on how to turn this into something helpful and sensible. We talked about how this work could fit in with existing things like horizontal review, and the Equity CG's in progress equity maturity model, and discussed ideas like a “Societal impacts” section in specifications. It was clear we need to work on how exactly this work complements other things.

More stuff

I did not manage to attend the main Sustainability Web Interest Group (SWING) meeting due to overlaps, but had great quality time with people from that group, including the shared enjoyment of traditional Irish music (shoutout Chris Adams for organising) and somewhat questionable but not terrible late night snacks.

We also had two AB meetings to ensure we were aligned on what sort of feedback to collect (start of week) and to reflect (end of the week), and used the opportunity to meet the TAG. And we managed to deliberate in person with the current ongoing W3C Council—again, great to have that opportunity to chat in person.

Summary

I'm sure every single attendee had a different week, even if we intersected all the time. I really enjoyed mine. It was another excellently organised event, thanks to everyone involved from W3C Team, and all the friendly people that were in attendance.

Originally posted as Reflections from TPAC 2026 on Hidde's blog.

Reply via email

Bye Deno, Hello Cloudflare - Adam Argyle

2026-11-05T16:30:00.000Z

So Deno, I still really align with it. I had a blast customizing the crap out of an artisanally rolled server rendered blog I built on it but…

They were acquired by Cloudflare and Deno Deploy was set to shut down.

I took this as a good time to get off the SSR + cache system I'd created on Deno + KV, and get closer to what my site actually was, a mostly static site.

I picked Astro + Cloudflare. It's done, and you're reading this on it.

I've always liked Cloudflare, and I knew exactly which pieces I would use to support my site migration:

I'm sure Deno was basically spinning up equivalent services on Google Cloud, so now I'm saving money per month and even closer to the individual tools I needed to adopt.

The plan #

JAM, CDNs, SQLite, Cron

Make the site static.
Every page is built ahead of time, no server rendering.
Keep the 3 dynamic bits alive in a tiny Worker.
The anonymous like button, live view and like counts, and the social cron jobs.
Put everything on Cloudflare.
Cron, edge endpoints, D1 for likes and counts, R2 bucket for static assets.

Fresh → Astro #

The posts on my site, the core content, were markdown already.

Since Astro is rad, my Deno Fresh / Preact components were kept and are rendered inside Astro at build time. Kinda nice I didnt need to migrate my component model.

Most of the migration was spent porting Deno and Deno Deploy features to Cloudflare.

Stuff I juggled #

Deno KV migration: Likes and counts lived in KV. They moved to D1. I exported production's KV (about 1,900 entries) and imported it, so no counts reset.

911 MB of media: Way too much to ship with every deploy. It all moved to R2, and the Worker serves it from the same /media/* URLs, so no old link or embed breaks.

Crons: Deno.cron became Worker cron triggers. The social stats job got smarter on the way: fresh posts get checked often, old ones rarely, so it stays inside the platform's limits.

Analytics: Pirsch hits now fire from the Worker on the server side, so adblockers can't eat them.

A moving target: I kept writing posts and merging PRs mid-migration. Instead of freezing everything, a sync script replayed new commits onto the Astro tree, right up to the end.

Secrets: Some lived in Deno's dashboard and can't be read back out. I had to conjure freshies and set them on the Worker. Not hard, just annoying.

The cutover #

DNS switch went smooth. Just had to add the zone to Cloudflare, copy records, swap nameservers. I flipped them this morning, attached the custom domain to the Worker, and boop done.

I'll keep the Deno site up for a few days as a rollback. Plus, a couple old DNS caches will still hit it for a bit, so I'll reconcile any likes that landed there in the meantime. Maybe.

The results #

Thanks Deno, it was a good run.

Same site, pixel for pixel
Faster
Cheaper

🦕 → ☁️

The forcing function worked out pretty good. Sometimes you need a good kick in the ass.

WWW Ep257 Local Engineer Runs Out Of Tokens - Adam Argyle

2026-11-03T01:38:44.000Z

Ep #257
Local Engineer Runs Out of Tokens, Discovers "Programming"

Robbie and I sip a rye and talk token budgets, Copilot getting yanked mid-workflow, SDD with worktrees, whether frameworks still matter, slop as a choice, Jeff the cliff-diving lawnmower, and Amp as the a Bowflex.

⤷ whiskey.fm · youtube · spotify · apple

Presented by Mailtrap

Scroll Axis Design System Rule - Adam Argyle

2026-10-28T23:49:04.000Z

Design system rule:

Y scroll = scroll trigger
X scroll = scroll driven

Designing Whiskey Glasses with Denver Cramer - Adam Argyle

2026-10-21T02:48:36.000Z

Robbie and I chin wag and sip on a tasty dram with Denver Cramer; industrial designer, distiller, photographer, and creator of Denver and Liely glassware.

Dude, Denver is coo.

Watch the full episode on YouTube

whiskey.fm · spotify · apple

The dram #

We enjoyed a sweet n' peaty Starward single malt finished in a Lagavulin cask 😍

Red wine barrels, Melbourne's four-seasons-in-a-day climate, and grassiness fading into sweetness.

Plus we dig tons into other whiskeys, like why Octomore 8.3 is the ultimate peat monster.

The glassware #

We dig into designing a bourbon glass for corn spirits versus a whisky glass for barley, smelling at the top versus the bottom of the glass, and eight years of R&D.

Plus the turbine wine glass Ryan Reynolds loved:

Rad.

The whatnot #

Of course we gotta bring the whatnot!

raising kids before and after AI
teaching them to code with retro video games
planting 400 Pinot Noir vines to learn winemaking
luthiers versus engineers
and why the banjo is part drum, part guitar, part lead singer

Presented by Mailtrap

WWW Ep255 Meta Harnesses Medical Midjourney And Specialists - Adam Argyle

2026-10-11T11:33:36.000Z

Ep #255
Meta-Harnesses, Medical Midjourney, and Why Specialists Are Cooked

Robbie and I are back together, talking token limits, OpenCode burning budgets, work trees for parallel agents, Impeccable's art direction grilling, CodeRabbit versus Copilot for PR reviews, self-healing loops that never heal, a robot lawnmower named Jeff, and why the meta harness layer is probably going to die soon.

⤷ whiskey.fm · youtube · spotify · apple

Presented by Mailtrap

My top picks for Interop 2027 - Sara Soueidan – Blog

2026-10-11T00:00:00.000Z

I love the Web platform, and it’s a provilege that we, the web community, get a chance to influence what browser makers work on every year.

For the second year in a row, Firefox has created a page that allows us to rank the Interop 2027 features that we care about the most. Hopefully this is becoming a yearly tradition.

Ranking our favorite features gives the Firefox team data they can use when reviewing which proposals should be taken on in Interop 2027: a yearly project with the goal to increase the interoperability of key features of the web platform across browsers—a dream come true for people who build the web.

I’ve been seeing some of my web friends blog about their rankings. (I’ve shared links to the blog posts I’ve seen and read at the end of this post.) So, instead of just posting about my top choices on social media, I decided to join in with a short blog post—a chance to breathe some life into this blog again, with more blog posts scheduled to come soon.

My choices are primarily based on a feature’s intended use cases/outcome; and my favorite features are the ones that allow us to improve both our websites’ user experience (UX and accessibility) and our developer experience (DX) at the same time.

Here are the features I am looking forward to the most and that I hope make it to Interop 2027. Only the first three are ranked. I don’t necessarily have a ranking preference for the remaining ones; they all got a vote from me and would sure be great to see them supported across browsers next year:

My picks

The HTML focusgroup attribute.
The ariaNotify() API.

Both of the above features—when used correctly and responsibly—enable us to improve the accessibility of our user interfaces and reduce the amount of work we need to do manually ourselves to achieve that.

The HTML headingoffset and headingreset attributes and the CSS :heading() pseudo-class (a whole package, yay!)
Interest Invokers. Another native HTML feature that will allow us to hopefully create more accessible common UI patterns.
CSS border-shape property. I can’t wait to ditch the old hacks when creating simple borders shapes for a simpler and more standard approach.
CSS interpolate-size and calc-size(). We’ve been waiting for these for wayyyy too long.
Allow CSS transitioning/animating display and content-visibility—for all those nifty UX-improving animations on common UI patterns and components like dialogs.
Customizable/stylable