Hacker News Reader

Comment by chrisco255 | Hacker News Reader

This interview is as much about the history of early web development and the serendipitous process by which JSON won out. After reading, it's apparent that JSON could have easily lost and been looked over entirely. Crockford's persistence in evangelizing it even after the company he worked at when it was created went bankrupt led to its eventual adoption.

View on HN

> Crockford's persistence in evangelizing it even after the company he worked at when it was created went bankrupt led to its eventual adoption.

I don't think this is true at all. The reason that JSON won is stated quite clearly in the interview: you do not need a software stack to handle JSON. This feature sells itself for any software development team who needs to have frontends running in a browser consuming data from a backend. You do not need an evangelist to convince people to use a feature that is ubiquitous, free, extensively tested, and more performance than anything in the market.

There's plenty of historical examples of sub-optimal standards developing enough traction and network effects that the whole industry is stuck with them for decades. Initially you did need to pull in a library to parse JSON, it wasn't part of the JS language until Crockford pushed for it (via his later position of authority at Yahoo working on YUI).

Particularly in the early web, the need for async "liveness" of pages didn't exist, so pages didn't consume backends directly. Everything was handled on refresh or navigation and server rendered as HTML. It was only after AJAX-style web apps took off (post-Gmail, really), that JSON started to get real traction. There were no JSON parsers in C#, Java, Python, etc initially. XML was far more common and SOAP was generally used for cross-endpoint communication.

Reply on news.ycombinator.com