This is a nice reminder of how much better we have it now. Sure, it sucks that Chromium now dominates in the way IE did. But at least all relevant browser engines are now open source, which means quirks like “how can I get the engine to know about my fancy new tag” can be addressed by looking at the source. That’s a big step up from IE.
It’s pretty impressive to me that a combination of Stallman inventing the (L)GPL, the KHTML team choosing it, and then market forces, brought us to the point that all big browser engines are open source. Mozilla would’ve always been, but I have no doubt that Apple and Google would’ve kept it all closed if they could have easily done so. To me, a long time GPL sceptic, this is the strongest example in favour of viral licenses.
Paul Irish is a little quiet lately. He used to be very active on Twitter, but only posts sparingly now. Also his second last post is from 2018[0] with a three year gap between that and his latest post, which was last year. No post for 2022 yet. I always admired that guy, and still do.
He contributed a lot to the webdev scene. But the scene is saturated now with developers of the same Paul Irish caliber pushing out tonnes of brilliant stuff, it makes me wonder if it's hard to get noticed now, with the constant barrage of new webdev material & tutorials.
> makes me wonder if it's hard to get noticed now, with the constant barrage of new webdev material & tutorials.
As someone with 10+ years of blogging about webdev, publishing tutorials, and some indie success with monetizing all this: Yes it’s harder to get noticed now. And also it feels less interesting.
When I started, the web was new and exciting. At least to me. The world was our oyster and we were the frontier of a new wave. That was fun. And there were few of us digging around, playing with tech, and figuring out new strange ways of holding existing tech to push the envelope. Even if just a little.
Then someone discovered content marketing is a good growth tactic. Then they professionalized it. Then they hit scale. Now much of the content you see is either targeted at beginners because there’s a lot of them and the content is easy to produce – but uninteresting to us experienced folk. Or it’s produced by teams of developer advocates financed by deep pockets and supported by all sorts of processes. You can’t compete with that no matter how hard you try. A team of full-timers will outwork and outproduce you every day.
That leaves the more interesting content, the deep truths of software engineering that play behind the framework and tactic du jour. Those are hard to uncover, difficult to explain, and usually require so much backstory that there’s no way to fit them into the attention-grabby environment of casual browsing. And they don’t play well with SEO (beginners don’t search for it), neither does it play well with recommendation/social algorithms (not instantly catchy).
The nail in the coffin for me was a youtube comment saying “Hey can you make NextJS tutorials instead? Nobody uses React anymore” – Next is built on React.
tldr: There’s too much content, mostly for beginners, churning through framework after framework, and all of it has become exhausting instead of exciting. Audiences don’t have time for the interesting stuff.
edit:
Counterintuitively, I think the good stuff is no longer online. It’s in books. Lots of great new books coming out regularly.
I think the true depth of content moved to be behind paywalls. I consider myself an advanced Frontend engineer and I find only Frontend Masters and specialized (and expensive) courses can really help me fill what gaps I do have or learn something with more depth.
I also read a lot more technical books than blogs too, but I’m finding even those harder to find that are relevant
Yeah I hadn't heard from him for many years. Just recently I submitted a request to the chrome devtools team, and Paul responded. I was like "hey I know that name!"
Wasn't this moved into Modernizr? There's a library I haven't thought of in a long time.
Oh, yep. There it is at the bottom:
> April 2011: IEPP v2 comes out. Modernizr and the html5shiv inherit the latest code. Meanwhile developers everywhere continue to use HTML5 elements in a cross-browser fashion without worry.
Paul Irish also started (or helped start) the html5 boilerplate project[1], which I am surprised to see is still actively being developed. I relied on that boilerplate a lot earlier in my career (started full-time in 2011). If you want another blast of nostalgia, check out the earlier releases[2,3] of it. I'm so happy I don't have to use conditional IE statements in my html anymore
I like how they've misspelt "misspelling" there - I wonder if that was deliberate! (I also like how there's even an open pull request to fix the same...)
I sometimes miss the days of XSLT and XML. They were very performant and easy-to-use client-side templating technologies that had a lot of power. I still think there was a lot to be gained by combining XSLT and JavaScript client-side.
Then again, when I look at old projects I do find it extremely verbose and JSX makes me so much happier.
It’s pretty impressive to me that a combination of Stallman inventing the (L)GPL, the KHTML team choosing it, and then market forces, brought us to the point that all big browser engines are open source. Mozilla would’ve always been, but I have no doubt that Apple and Google would’ve kept it all closed if they could have easily done so. To me, a long time GPL sceptic, this is the strongest example in favour of viral licenses.
Thanks, FSF and KHTML team!