Future of the engine
-
You know, I've been speculating about the future of the engine Vivaldi uses, Blink/Chromium. We know that Chromium/Blink is written with sloppy code, and it's also resource hogging. It's also a bit tough to customize like you would with Gecko, but I would imagine that these problems could all be rectified. I forgot which article I read it in(and I tried digging for it deeply before writing this), but it's said that Old Opera was always about writing their own code for everything, hence writing the Presto engine. From what I see, Vivaldi took the Chromium/Blink engine and based it off of that. They didn't build their own because that would take a very long time and Chromium/Blink is so widely compatible and supported, and had the biggest market share that it was just easier to base it off of that. But Blink/Chromium are in dire need of improving. I see in the future it could improve a lot, and Vivaldi will be the leanest, fastest, most customizable, best browser out there. Until then, what do you guys think the future holds for the engine Vivaldi uses? A few things could end up happening . . . 1. They end up staying with Blink/Chromium and wait until it improves, or better yet, help improve it themselves once they've established their browser more with the Beta and the final release. 2. They might fork the Blink/Chromium engine, and still enjoy the benefits of the engine, but with their version, under their control. 3. They create their own engine again. 4. Figure out a way to get the license for Presto and improve and build off of that. Anyway, this speculation has been on my mind, and so I felt like speculating with the forum. I personally think 1 is probably the most likely to happen, but 2 is not impossible either. Lets speculate!
-
1. Yes. Optimizations, largely driven by the contributions of NewOpera developers as it happens, are already in the works, and some have been implemented - mostly aimed at the memory footprint.
2. No. Bad Idea for so many reasons…
3. No. See 2 above
4. No. See 2 and 3 above.
-
To elaborate: The death of Presto was site compatibility. This has not changed and never will. Options 2 and 3 suffer from the same weakness at minimum. Option 4 has this problem, PLUS the fact that Opera still owns Presto, is still developing it, and is still shipping it in both embedded devices, and in Opera Mini for mobile. No small, smallish or medium-sized company in this day and age can afford the resources to develop, maintain and continually modify their own engine, to keep up with shifting defacto web standards as sites tailor themselves exclusively to Chromium, Gecko, the MS engine (be it Trident or whatever the hell Edge is on now - they call it EdgeHTML - a fork of Trident) or Webkit derivatives like Safari et al. Opera, with a couple hundred developers, could not do it with Presto. That's a lesson well-learned.
-
The Opera team are also contributing to Blink development, and I am sure that the Vivaldi team will improve it too. Lazy Tab loading should speed things up for those who open many tabs. Blink is the only viable option. All other efforts are best spent on developing the User Interface and features like email, synchronising, turbo, and file sharing like Opera Unite or TOR.
-
That's a good point, Ayespy. I knew you would answer my question pretty quickly :).
So I can see that the New Opera is developing and making the Chromium engine better. Do you think Vivaldi will do work on it too? They seem very talented. I would love their work to go into making the engine light and very easily configurable.
And Pesala, I saw the link you showed me. New Opera is developing on the Chromium engine to improve it. That's great, but on one of their efforts is to hibernate tabs, which is fine, but I hope they have other ways to make the footprint much lighter than just that. Other browsers like Opera Presto seemed to run lightly without needing that.
What're the chances of Chromium becoming as light as Presto or something, or just as configurable? I would love to see it be something that can fully replace Presto, Gecko, etc.
-
I'm sure there are any number of ways in which Chromium/Blink can be improved, but I don't think there is any way to make a browser which uses an additional process for each tab and every extension, run in the same resource-thrifty manner a single-process browser can.
Vivaldi can keep the footprint light by building in as many options and functions to their part of the browser as possible, thereby making the need for extensions a rarity.
Once Vivaldi is a larger (and profitable) concern, whereby they can afford a couple of extra guys, I would not be surprised to find them contributing to making the Chromium code more streamlined and more powerful.
-
I'm sure there are any number of ways in which Chromium/Blink can be improved, but I don't think there is any way to make a browser which uses an additional process for each tab and every extension, run in the same resource-thrifty manner a single-process browser can.
Wow, that's what Chromium is, huh? What's the benefit of that, and can they change it to single-process?
Once Vivaldi is a larger (and profitable) concern, whereby they can afford a couple of extra guys, I would not be surprised to find them contributing to making the Chromium code more streamlined and more powerful.
I'd be happy if they could do that! This team is seriously talented, and Jon understands the power of optimized, great code like Presto was, and has compared it to Chromium(I still can't find the article that talked about that. Sorry). I'd love to see what they could do with the Chromium engine. I'm sure there are tones of Opera devs that have worked on Presto before and are working on Chromium/Blink.
Then there's the Gecko engine, which apparently has an easier way of customizing it over Chromium. I'm sure, and would be happy for those that want it, that Vivaldi may configure the Chromium/Blink engine for their needs and make it easier to configure.
Mozilla and Samsung are creating the Servo engine together from what I hear. I wonder if that'll be even better than the engines out there today so far. And having Samsung back this up is a force to reckon with. Some competition will be good.
I'm sure Google's devs are good, as it's Google, but it does make me question their talent compared to Opera/Vivaldi's. That would be interesting to find out.
-
Jon and crew decided not to go with Gecko. It's declining in popularity compared to Chromium, and is just as resource hungry under several scenarios. Plus, Chromium code is much easier to adapt to mobile, given the Chrome/Android connection. 80% of all mobile devices are running Android.
When Google first introduced multi-process Chrome, the selling point was that a single tab could crash and it wouldn't bring down the entire browser, due to the multi-process architecture.
Chromium/Blink is a derivative of the Webkit engine, which is single-process BUT, Webkit has some serious limitations and running as a browser engine on its own, has a teeny-tiny market share, which means that no one writes for it, and no one tests in it - the very things which doomed Presto.
So I expect this team to stick with Blink and to, at some point, begin contributing to its code.
-
The sad fact is that the whole flap about "web standards" was really just a cover for being anti-MIcrosoft. Now that the web is dominated by a non-MS monopoly with a crappy browser , the same people who complained before are now quet as mice and happily conforming to a single browser engine. They never really cared about competition and diversity of browsers, they just hated Microsoft.
So for now we're stuck with Google's ugly child. But the market changes all the time. We can live in hope that the future will eventually be different.