Updated: August 19, 2017
Ever since Mozilla embarked on the Chrome-me-up journey a few years ago, my enthusiasm took on a six-weekly decline cadence, with each new release of the Firefox browser bringing in more of what Firefox shouldn't be and less of what made it such a cool program in the hands of its loyal users. But the best is yet to come. The true rite of passage. Only the most righteous will survive. WebExtensions.
While trying to salvage some of what it still has left while actively scuppering its fanbase and killing off its powerful extension mechanism, Mozilla is working on giving its browser a breath of fresh air. More speed, it seems, as though it is the critical factor that made people abandon ship. But assuming it is, does it make a difference? Let's test.
You've all heard and read about multi-processing in Firefox. Per-tab processing, better security, sandboxing, better stability. All noble goals, but they may be an answer to a plea that was never uttered. Regardless, I'm playing ball, so I do wanna see what Firefox can do when its new technological underpinnings are given a chance to spread their wings.
As you may expect, on most of my systems, I am running Firefox sans Australis, without its bold new interface that has been designed with only 45 chromosomes, which means the most supreme CTR extension, which means no e10 multi-processing. Indeed, CTR is one of the many extensions that will die come Firefox 57 or 58 or such.
I want to see what Firefox can do without the extras. In my Kubuntu distro, it has no extensions, save for the default Ubuntu Modifications, which actually interfere with the multi-processing implementation. So I had it removed. And then I added Noscript, Adblock Plus and Downloadhelper, which seem to cooperate nicely and might be in the small set that survives the winter culling. I turned Noscript off, and did most of my tests with only the other two turned on, and sometimes, only the adblocker.
It is important to take this into account and compare to Chrome, also installed side by side with Firefox in this setup, with the Adblock Plus extension in its arsenal. Hence the reason for having it enabled in both cases. Now, it is also possible to do a test without any modifications, but I am not running a browser benchmark per se, and there's no reason to run the checks without these extensions, because I find them necessary.
Well, with the browser in full MT mode, there is a definite performance improvement. It is evident through simple website interaction, but also the overall responsiveness of the GUI to innocent actions. However, the startup still remains awfully slow compared to Chrome. Google's browser launches instantly so to speak. Firefox takes its merry time.
Multi-processing leads to an increase in memory, and Firefox is still a hungry beast CPU wise when you do things, even seemingly innocent things like downloads. In this regard, it is no different from Chrome, but the latter suffers less penalty as a consequence, probably because it has a more efficient multi-threading architecture.
There are other things that affect the perceived responsiveness. For instance, smooth scrolling is one of the features that you should consider turning off right away. Hardware acceleration may be in your favor. Having Flash enabled can also impact the performance, especially during video playback.
roger 6719 2105 ... /usr/lib/firefox/firefox
roger 6778 6719 ... /usr/lib/firefox/firefox -contentproc
-childID 1 -isForBrowser -intPrefs ...
Firefox without CTR modifications looks like any typical modern browser - an abstract representation of stupidity. Tabs on top will go down the history lanes as one of those things that were created for the sake of it, justified post-creation as the next best thing since the Spanish Inquisition, and will remain around despite being a manifestation of all that it wrong in this world, even worse than hunger and Ebola.
So if you can't have a sane browser, what can you? Well, you know you can't really customize Chrome, but it has always been that way, you don't expect to. With Firefox, you can play with half a dozen options and buttons, which go top right, plus icons for your various extensions, just like Chrome. No more status bar. No more tabs on bottom. No more back, forward, reload and stop buttons as they used to be. It feels minimalistic, but not in a good way. Yes, you can get used to it, the same way you can get used to working in a mine quarry.
In an effort to mollify users, Mozilla offers new, lightweight, compact themes. These offer an additional level of abstractization and a bit more squarish geometry, but it's a far cry from a simple, intuitive and ergonomically correct form that used to be. Whatever you decide to do, it will be a compromise.
So what does this mean? Well, the way I see it, Mozilla is probably trying to create a platform that will sort of retain most of its users through inertia and apathy, adding some changes that kind of minimize the damage. But they do not take away.
Improvement? Well, in a way, yes. Multi-processes is a must in the modern era, and it does bring some benefits. The compact themes can almost be excused as something really neat. But that's not the whole truth.
The price to pay is quite high. The reward isn't good enough. Not visually and not actually. The performance isn't as good as you'd hope, so you would be willing to sacrifice ancient extensions for the sake of awesome speed. Nope. Chrome remains ahead in pure user experience responsiveness, and with both browsers now sharing the same pointless design, Firefox losses still more points, because it has neutered all and any advantage over Chrome by moving to WebExtensions.
I don't know if Firefox ever regains its smooth elegance that it had around 2011, but whatever it does, it's still on the defensive. Chrome is ahead, and should remain so, because so far, it has. With one big difference. It remains largely consistent, and the changes, when they happen, are mostly invisible. Firefox goes for a highly disruptive revolution, willingly sacrificing old for new. But why would there be new?
All in all, it seems we will get another happy-go-lucky browser, without any distinct features, maybe, hopefully a few useful extensions, and performance that is not quite as good as its chief rival. And if you tell me you need the latest and greatest CPU to realize the benefits, then you're deluded. Browsers aren't computer games. They should be slim and fast, and if they cannot realize the entire spectrum of computing power available, then it's a lousy implementation. The perceived slowness and startup speed issues remain. And unless solved, there will be nothing to give Firefox the edge.
There is progress. I do not dispute that. When you throw away legacy stuff and let the browser engines run in parallel, you do get a very nice boost of oomph. Some popular extensions still work, and hopefully they will continue after Firefox 57. Compact themes are reasonable. When you just focus on these bits and pieces, it feels that Mozilla is doing something right.
But to throw away a decade of hard work is a foolish way to redeem oneself. It's just not how professional work is done. Yes, in the long run, by all means. What's missing is the seamless transition. That's what Mozilla didn't get right. Seamless. Transition. The magic of doing software the correct way. On top of that, we still don't get better results than the rival, both startup and normal use, and the lightweight themes will not fully appease the veterans. So the gains are offset by big losses. And the biggest is not beating Chrome fair and square. If you look at this whole fiasco in simplistic terms, you sacrifice a lot and still come second. Why bother then?
Anyway, here we are. Firefox 54 - and what it signifies - has some neat points. If I were not so jaded and scarred, I might even be cautiously optimistic. But this progress comes at the expense of years and years of loyalty and passion, and that cannot be translated in pure technological terms. This might be the right thing. And even the needed thing. But it is too little, too late. Not enough to defeat Chrome, more than enough to destroy the fanbase. This is not the winning formula. Might change still, but for now, it's a dire situation. More to come.