Just couldn’t stop reading this whole series <Reck...
# linking-togethers
Stefan
08/18/2024, 9:51 PMJust couldn’t stop reading this whole series Reckoning about modern web development. I haven’t done any web development in two decades and am totally checked out of what is state-of-the-art over there. Is it really such a dumpster fire? Is this a credible source (his credentials certainly seem like it)?
Since at least 2016 I have the frontend community to step back from the brink of JavaScript excess, acknowledge mobile-first reality, and prioritise users at the margin.
Instead, frontend became a responsiblity-free zone, externalising costs onto users and businesses with 'full-stack' fairy tales and used-car sales tactics.This as culminated in heartbreaking hurdles to access crucial public services thanks to JavaScript.Don’t skip over the footnotes, the best stuff is hidden in there. Also has tons of links to really good other material that I hadn’t come across, like good advice for making tech decisions and “innovation tokens”, and why they’re not helpful.
💯 3
🔥 2
👀 1
ivan 1
❤️ 2
📣 1
i
Ivan Reese
08/19/2024, 4:12 AM
Oh yeah, this series has been absolutely fantastic! Glad you're on it too. I want to shout it from the rooftops.
Stop taking dependencies!! If you need something, build it yourself. If you can't do that —look, I get it, material circumstances— then take the smallest, simplest, most delete-friendly dependency you possibly can. And never, ever, ever use someone else's build tooling.
But I'm over here optimizing every SVG I upload by hand, so I might be too extreme in the other direction.
💯 1
Ivan Reese
08/19/2024, 4:29 AM
I can imagine an ethnography covering the effect of tech influencers on the proliferation of devtools:
Let's say someone does a tech podcast. They need to come up with a fresh topic every week. It's easier to cover a library or framework than to cover, say, the value of using plain data structures 99.9% of the time, and how to identify the 0.1%. It's easy to talk news, very hard to dredge up wisdom week after week. And if you cover the new hotness, you get to participate in hype waves, which are great for audience growth. This in turn encourages devs to make more of the sorts of devtools that get coverage. A race to the bottom ensues, where devs trade on empty promise, and audiences are never exposed to true exemplars of craft.
I should try to dig up this episode of… some JS podcast I listened to a decade back. They had some devs on to talk about their new library. It was blowing up on twitter. The devs had, thus far, only put together a static landing page and a logo. No code, no demos, barely any description of goals. It was pure vacuous hype. And on the podcast, they were already hyping the next thing.
c
Christopher Shank
08/20/2024, 4:37 AM> Is this a credible source (his credentials certainly seem like it)?
I think Alex's is credible on these topics. I’ve been following work on browsers and web standards for a long time and his challenges on the defacto approaches for making website is usually something nod along to.
Christopher Shank
08/20/2024, 7:37 AMPart of it feels like JS tooling has gotten so “good” and frameworks are abstracted far enough from the browser that we have no idea the final output of the websites we make. No idea its size, load time, how it’s impacting marginal user’s.
d
Don Abrams
08/20/2024, 1:15 PMThis isn't ringing widely true to me since most people still just use wordpress or something similar that's mostly HTML
His app examples are public projects made by cities and govt agencies (who are generally underpaid/trained) or contractors (incentivised to keep it complicated, pump up hours, and stay needed)
In 2016 when our 6 person FE+app team moved to react+react native (from jquery/ObjC/Dalvik), we definitely optimized for time-to-display (using server side rendering and rehydration) because it mattered for SEO and real UX for our users (mostly women over 50)
Why did we go react? Greatly increased debuggability and faster time to market, plus 60% code reuse between app and mobile (kept app because the per and experience were better)
We wanted the better DX and made sure we didn't sacrifice our UX
Then again, last week I had a contracting call where I did have to discuss mobile first with a desktop-focused CEO (and then pivoted from selling him FE services to a small UX study)
So maybe this is just a reminder to stay close to the users and not solely focus on DX or whatever google/facebook does