[aparsons]

The IE effort was gargantuan - it was not NT in terms of sheer difficulty, but 1) we had built operating systems before NT, but never a web browser, and 2) Microsoft wanted to keep it under wraps as much as they could.

The IE team had a carte blanche to succeed by any means necessary. If they wanted a system call to work differently, you had to get it done or risk getting a meeting with your boss’ angry boss.

[sharikous]

Interesting. I would have thought that the state of the web 25 years ago was such that a browser would be a rather simple endeavor. Especially with a company with products like MS Word.

[pavlov]

Sure, the standards-based web was quite simple 25 years ago. But of course that's not the web that Netscape and Microsoft cared about.

Netscape and Microsoft were engaged in a land rush to extend the web into their proprietary platforms. If they could grab enough marketshare, they'd get to define the web's direction. Initially it was Netscape in the driver's seat, but by 1999 Microsoft had taken control.

It was an opportunity to invent everything from scratch. There was no standard styling, no standard DOM, no standard plug-in API, no standard programming language...

Solutions and answers that seem predetermined in hindsight were anything but. For example, Netscape was pushing their own scripting language initially called LiveScript but then rebranded into JavaScript as part of a Sun partnership. Meanwhile Microsoft's browser supported multiple programming languages that could access a common object model: you could program web pages in VBScript, or Microsoft's own JavaScript look-a-like "JScript", or other languages not yet defined.

Maybe a language-agnostic model would have been better for the web than JavaScript dominance? We might have had something like WebAssembly much sooner. We'll never know. But the answer to these design challenges certainly were not obvious in 1996 to engineers at either browser company.