4 ms·

While I understand why folks would miss the loss of XUL/XPCOM, stating that it was to ape Chrome feels disingenuous. Removal of those technologies enabled multi

While I understand why folks would miss the loss of XUL/XPCOM, stating that it was to ape Chrome feels disingenuous. Removal of those technologies enabled multiprocess Firefox and bolstered stability. If anything, based on comments I see here and elsewhere, it was one of the things that brought people back to the browser. Adopting/starting WebExtensions made sense, too (IMO), as Chrome already had way too much mindshare introduce anything totally novel.

arp242 3y ago
It was 100% painful and frustrating, but also 100% done for technically valid reasons. I've always preferred Firefox (or Opera Presto, before that) for the simple reason that Chromium's text selection is weird, clunky, and unlike anything else I've ever seen (this is also how you can spot any Electron app). But it really was just ... slow. Moving away from XUL was an important step in fixing that. That said, I do feel that Firefox fell in the "X is more popular, X does Y, therefore, we need to do Y" trap, forgetting that you have your current userbase exactly because you're NOT doing Y but Z. But that's an entirely different matter.
orra 3y ago
100% this. Folks used to lament Firefox for being slow. And for me it would freeze for seconds sometimes, because of synchronous extensions APIs. Removing XPCOM, switching to multiprocess, and switching to asynchronous Web Extensions was absolutely vital technologically. Perf is good now.
pseudalopex 3y ago
No. Firefox 48 supported multiprocess. Mozilla pushed developers to make their XUL extensions multiprocess compatible after WebExtensions development started. Firefox 57 disabled XUL extensions. They removed XUL gradually after.
lproven 3y ago
> While I understand why folks would miss the loss of XUL/XPCOM, stating that it was to ape Chrome feels disingenuous. I wrote the article. I did not state that. I did not even hint it or imply it, and my eyebrows went so far up when I read this comment that my hat lifted. I've never even heard this idea before. Please can you point out where I said that the removal of XUL was in order to ape Chrome?
- arusahni 3y ago
  That was my read of this paragraph: > Mozilla, please stop aping Chrome. Copying is rarely the way to win big. The Australis Chrome-like theme in Firefox 29 annoyed users and was a driving force behind Pale Moon. Firefox Quantum killed XUL addons, and drove The Reg FOSS desk to Waterfox Classic.
  - lproven 3y ago
    Hmmm. I see what you mean, but that was not my intention. The meaning the author -- me -- intended was: "A drove people to B. It didn't drive me to it, but C drove me to D." Expanding that: 1. Stop copying Chrome. 2. Australis copied Chrome and that drove a whole fork because so many people hated it. [Full stop.] 3. Also, comparably, for the author, Quantum killed XUL and that is why the author switched away from Firefox. However, I note that the marketing claim from Mozilla was that Quantum used the same type of extensions that Chrome used ("webextensions") and that this would be good for the Firefox add-on ecosystem. This is demonstrably false, as I laid out in the Waterfox article which this story links to, but it was what Mozilla claimed at the time.

arp242 3y ago

It was 100% painful and frustrating, but also 100% done for technically valid reasons. I've always preferred Firefox (or Opera Presto, before that) for the simple reason that Chromium's text selection is weird, clunky, and unlike anything else I've ever seen (this is also how you can spot any Electron app). But it really was just ... slow. Moving away from XUL was an important step in fixing that. That said, I do feel that Firefox fell in the "X is more popular, X does Y, therefore, we need to do Y" trap, forgetting that you have your current userbase exactly because you're NOT doing Y but Z. But that's an entirely different matter.

orra 3y ago

100% this. Folks used to lament Firefox for being slow. And for me it would freeze for seconds sometimes, because of synchronous extensions APIs. Removing XPCOM, switching to multiprocess, and switching to asynchronous Web Extensions was absolutely vital technologically. Perf is good now.

pseudalopex 3y ago

No. Firefox 48 supported multiprocess. Mozilla pushed developers to make their XUL extensions multiprocess compatible after WebExtensions development started. Firefox 57 disabled XUL extensions. They removed XUL gradually after.

lproven 3y ago

> While I understand why folks would miss the loss of XUL/XPCOM, stating that it was to ape Chrome feels disingenuous. I wrote the article. I did not state that. I did not even hint it or imply it, and my eyebrows went so far up when I read this comment that my hat lifted. I've never even heard this idea before. Please can you point out where I said that the removal of XUL was in order to ape Chrome?

arusahni 3y ago
That was my read of this paragraph: > Mozilla, please stop aping Chrome. Copying is rarely the way to win big. The Australis Chrome-like theme in Firefox 29 annoyed users and was a driving force behind Pale Moon. Firefox Quantum killed XUL addons, and drove The Reg FOSS desk to Waterfox Classic.
- lproven 3y ago
  Hmmm. I see what you mean, but that was not my intention. The meaning the author -- me -- intended was: "A drove people to B. It didn't drive me to it, but C drove me to D." Expanding that: 1. Stop copying Chrome. 2. Australis copied Chrome and that drove a whole fork because so many people hated it. [Full stop.] 3. Also, comparably, for the author, Quantum killed XUL and that is why the author switched away from Firefox. However, I note that the marketing claim from Mozilla was that Quantum used the same type of extensions that Chrome used ("webextensions") and that this would be good for the Firefox add-on ecosystem. This is demonstrably false, as I laid out in the Waterfox article which this story links to, but it was what Mozilla claimed at the time.