Hacker News Reader

Comment by ecaron | Hacker News Reader

When teams halt development on projects, I really appreciate when they say "People should go use project X" instead. I know it is difficult to full their full-weight behind a single endorsement, but the team is obviously picking an alternative and since their followers trusted their original code they should trust the successor.

It would be great if the YUI team stood up and said "We're moving to something, and think you should too."

View on HN

Right now the scenario here at Yahoo is mixed. We have teams using Ember and React, but no enforcement or recommendation to migrate any product based on YUI to any other technology.

I guess Yahoo official support for any other library is very unlikely to happen in the near future.

All of the UI teams in Ads & Data are mandated to use EmberJS. Having said that, I do know of a team exploring AngularJS and know that ReactJS is ramping up in other departments.

A few of us in Homepage & Verticals are using ampersand.js

Given how many things YUI was composed of, I doubt there's one single "move to this" library out there. People aren't making these sorts of monolithic do-everything libraries anymore.

I am. http://qbix.com/EGreg/Q

just sayin' :)

Message on that page: "http://qbix.com/EGreg/Q doesn't point to anything."

There's really no basis for putting a stake in the ground with web front-end frameworks right now. Based on recent history, the leading frameworks today will be YUI in 2-5 years. The best you can do is chose something that has a large community TODAY, get the most mileage you can out of it, and know that in 5 years or so you'll be scrapping most of it.

Scrapping it may not be an option. If you have built hundreds of thousands of lines on top of a framework, the cost of a rewrite will dwarf the cost of maintaining the framework yourself. The trick is to choose a framework which you can maintain yourself (sources to every aspect of it which you understand well enough to maintain).

"If you have built hundreds of thousands of lines on top of a framework,"

Perhaps that wasn't the smartest architecture decision made. Now, don't compound it by rewriting it in the same deeply coupled way for the next shiny framework that sashay's by.

Reply on news.ycombinator.com