I mostly solved our IE6 problem by moving over our web product to a single page architecture based on ExtJS in 2007. ExtJS had a browser abstraction layer, and as long as you built your screens out of its components you rarely had issues. The major pain points that remained were memory leaks (IE6 was a digital sieve, and SPA made it worse), and performance.

It is clear to me that we are in a holding pattern for web components to become the norm and for component frameworks to become decoupled from application frameworks, so I have no idea what I will be using in 5 years.

So, we'll arrive to something similar to where we are with JavaScript and jQuery, where the latter is less relevant now that JavaScript is appreciably better? Meaning, that with web components the norm, we might arrive to less need for frameworks like React or Vue?

(I might be misunderstanding due to the sleep deprivation, so apologies if what I write comes across as gibberish.)

We've experimented with the combination react + stencil to build in this way, but there were just too many problems relating to workarounds to get things to work and performance which just wasn't good enough.

That blown got blown away completely when Chrome and Firefox started doing updates. People finally got the message that instead of for coding to browsers you code to standards.

The loser was IE because 1) they had been really loose with standards in feeble attempts to lock people in 2) they had stagnated for years while confident their market position could never be overturned 3) by stoically carrying the line that IE was actually part of the operating system and couldn’t be updated independently they locked themselves into a place where IE only got major updates along with major OS releases 4) they had a mess of their double diamond titanium partner program members that would howl every time they tried to fix anything in IE because they had internal applications that depended on the broken behavior in IE and 5) BYOD happened and IT depts no longer had authoritarian control over browsers and their versions.

It worked out for the better I think. Today IE is only used to download Chrome or Firefox though I get a taste of the IE 6 experience trying to find a security profile for IE that actually allows me to hit a URL on the public internet, but it’s all good once you get one of the installers downloaded. Set the default browser to something else and never have to deal with IE again.

Edge could become, hands-down, the best browser in the world. It could automatically put free cake, pizza, and beer into my mouth. And i still would not use it because of the pain instilled on me by Microsoft by IE 6.

I spent over a decade getting smacked upside the head by IE 6. Until Gates and Ballmer personally drop by with a cheque for ten years of wages, I will never give any Microsoft product the benefit of the doubt, and I question the sanity and integrity of anyone who does.

Decisions should have consequences, and MS deserves a fate worse than Yahoo.

I worked at a web agency during the dying days of IE and we spent so much time fixing IE problems that we would instantly add on a third to the quote if the client wanted IE7 support, double for IE6. When faced with that, clients who previously insisted support was vital suddenly agreed that 11 was fine after all.

Microsoft may be a completely different organisation now, but I’ll never forgive or forget the amount of time they made me waste.

I understand. For me, it wasn’t the time but the frustration they caused me. So I won’t use their browser on any of my devices in order to keep their percentage market share one tiny tiny fraction lower.

Then the standards improved and the entire IE team was either caught flat footed or possibly even disbanded to work on other projects. IE7 was almost half a decade after IE6

Those of us who remember web development in the 90s and early 00s will vouch for how the pain wasn’t just supporting different rendering engines, but even different versions of the same rendering engine.

What happened is that standards finally got their act together, and FF was eating IE's lunch anyway. Chrome just sped up its demise. But, monoculture is never a good thing. Especially as the leading company is actually an advertising agency. Do you think we can trust them not to abuse their position?

I remember that I was working in an app back in 2010 that was heavily using the showModalDialog https://www.tjvantoll.com/2012/05/02/showmodaldialog-what-it...

This API was not supported by any other browser but was fundamental for our work because you could display a web page in a modal window that would return a value to the parent page when the modal was closed!

It wouldn't surprise me if people are still running IE6 in some virtual environment to be able to access such apps.

Another big thing was IE6 was supported by pretty much every online bank, particularly in South Korea, which is not a huge market, but SK was one of the last, most advanced economies to switch off IE6, primarily due to banking platforms.

I've worked on a project ~2y ago to rewrite the software powering the whole core business of a huge company (multi-hundreds-of-millions-a-year revenue I'd guess), the existing software being done somewhat like this, although it was a heavy client rather than web.

A guy developped software to automate his job 25y+ ago with some microsoft tech (FoxPro), somehow the whole company ended up using it and the software kept growing until it was fundamental to the business. For some reason it only ran on something like Windows 95, so they got Windows95 servers that you had to connect to remotely just to use the software. It was incredibly slow of course, and you couldn't copy-paste through the virtualisation layer, which drove everybody crazy for years.

The guy had developped that in his free time apparently (wasn't employed as a software developer), pretty good deal for the company who exploited that for decades and became a major player in the industry.

Ok, looks like business as usual.