Hacker News Reader

Comment by developer2 | Hacker News Reader

Mozilla absolutely made the right decision to dump a decade's worth of legacy code to improve performance, and make the future happen now. Let's be real: had they supported legacy extensions, extension developers would ignore porting until Mozilla finally said "that's been long enough, we're removing legacy support". The result is simply that the rough transition happens now instead of a couple of years down the line.

Every popular extension will be ported within a matter of weeks, worst case scenario 2-3 months for those that are unabanonded, but less popular. If you are missing a deal-breaker extension, then don't upgrade yet... how is that a bad thing? "But I want the latest and greatest, super fast Firefox now - with all my extensions!". Talk about wanting to have one's cake and eat it too. The new Firefox wouldn't be new and improved if it was backwards-compatible with the stone age.

View on HN

You forget the big technical issue: many web extension API features that are needed to port old Firefox extensions do not exist yet (e.g. filesystem access) or are conservatively crippled (e.g. adding buttons, menu items, other GUI elements). Your 2-3 months of frantic effort will start not immediately but in a vague future when the Firefox team feels like improving the web extension API. It is fairly obvious from API documentation and bug discussion threads that they only care for feature parity with Chrome web extensions, not with old Firefox extensions.

The transition could have been managed gradually, by piece by piece replacement of the old extension API with the new web extension API (made available to old style extensions). Extension developers would have made the small changes needed to port from a deprecated old API to a new API that does the same thing, up to the final step of reorganizing without significant code changes the old extension, now relying exclusively on web extension APIs, into a web extension. Abandoned extensions would have fallen by the wayside very fast, deficient APIs would have been fixed before actual usage, and Firefox would have moved forward without betraying users.

You're pretty much describing what Firefox has been doing for the past few years.

At some point, however, the rest of the world keeps moving and, no matter how painful for everyone involved, it becomes unfeasible to wait for all add-ons to be ported (or even portable).

What they're describing is hybrid WebExtensions, which were only introduced earlier this year.

not all, just mine

"Betraying users"? 52 ESR isn't going anywhere, and I hope you'll forgive me for finding it a strain upon credulity to imagine that any regularly active HN participant could fail to observe even one of the many front-page articles over the past year which have talked about Firefox 57's breaking changes. If it means that much to you, stay on 52 for a year or two, until the extension API and ecosystem have had some time to catch up and stabilize. In the meantime, slinging heated rhetoric on the subject helps nothing and no one.

> 52 ESR isn't going anywhere

In 4-5 months, ESR will be replaced by Firefox 59, and support for 52 will end shortly after.

Sure, the old version isn't "going anywhere", but running an unsupported browser these days is pretty foolish security-wise.

It isn't a matter of "failing to observe": without the necessary API, extensions cannot be ported. I personally looked into rewriting the extensions I needed a couple of years ago, I figured out it was impossible and I gave up.

Some very important extensions have fortunately been able to pressure Firefox into supporting them, but the typical Firefox user who depends on some niche extensions has no clout.

Such is life. I'd rather have Firefox survive and continue to compete effectively than go chasing after perfect compatibility with those extensions I've also temporarily lost the ability to use until the necessary APIs land.

If Mozilla could have done both, Mozilla would have done both. They could not. I get that that's super frustrating. I'm not happy about it myself, because the change broke my workflow a little as well - until I engineered my way past that, because I'm a grown-ass adult and I solve my problems, or learn to cope, instead of whining about them - and also because I put myself on the hook for a Firemacs reimplementation that can't land for probably another year at best. That's annoying. I get it.

But slinging vitriol on the subject obtains nothing and aids nobody. It makes the people who do it look like jackasses, it makes the people who do the actual work feel like they can't win and may as well not try, and it makes everyone else embarrassed both on the behalf of the whiners and for their own sake in being associated, however loosely, with a community so full of Tumblr-grade drama. It's embarrassing and stupid and pointless and counterproductive and I wish people would stop. There are better ways to spend the same effort - like, for example, contributing patches to the webex implementation. Hard work, I know, and whining is easy. But that doesn't really play in favor of the whining, either.

(Yeah, I get that you're not really a major example of what I'm complaining about. You just happened to be right here when I lost my patience. All the same, though.)

Reply on news.ycombinator.com