Opera: The Past, The Present & The Future
I must admit, after trying the first public preview of Opera 15, I was pretty underwhelmed. It felt like a Google Chrome clone with Opera logo on top of it, but then it got me thinking… Is there more to it?
With the release of Opera 15, Norwegian browser maker has decided to completely reboot the project. Some could argue that its years too late but slow progress is better than no progress, right?
So why do it at all? Let’s face it, Opera was (almost) always bad at rendering web pages. You can blame user agent sniffing, developers or your mom but that won’t fix the issue. We’ve been playing the blame game for how many years now? And that’s excluding awful scrolling performance, buggy WebGL implementation, broken out of process plugins and so on.
The foundation is old, rendering engine is a trainwreck and the whole Opera architecture currently looks like this:
Opera 12
Do you think we can stack more features on top of it?
So what is Opera 15? Opera 15 is management’s way of saying that they have had enough. Untangling the mess that Opera 12 is would require so many resources and so much time that’s it’s easier just to start from scratch and that’s exactly what they did. What we got is a solid foundation that looks like this:
Opera 15
Thanks to Blink / WebKit rendering engine integration, Opera has just eliminated its Achilles’ heel: web page compatibility. With Google and Apple behind our back, there is no way that we will ever experience same issues that Opera 7-12 users had.
But wait, there is more. Since Opera developers no longer have to spend so much time patching individual web sites and fixing broken architecture, resources can be shifted to adding new features and overall polishing. This also means faster release cycles and quicker implementations of new web technologies.
Basically, Opera has decided to go with a plan that inflicts pain in the short term yet provides great benefits in the long term.
Sure, it lacks tons of features with questionable new implementations (such as getting rid of bookmarks and using 3rd party extensions for RSS feeds) but we hope that Opera will listen to its community and add features that made us love Opera in the first place. Hint: it’s not a giant Speed Dial search bar that can’t be removed.
As I wrote in a previous post, “Project reboots are hard and require tons of resources (just look at the Windows Phone 8, which still lacks basics that were introduced 10 years ago in Windows Mobile like VPN support, separate volume controls, etc.)“ so keep that in mind. There is a saying that goes like this, “Patience is bitter, but its fruit is sweet” and that’s exactly what we need, a little bit of patience. On the other hand, there is another saying, “Abused patience turns to fury”, and this is something that Opera should keep in mind, you don’t want to piss your most dedicated fans off, especially when your overall market share is so low.
TL;DR: For Opera 15, Opera has built an entirely new and expansive foundation. They then proceeded to build a small house on top of it. Over time, Opera intends to extend your house and eventually replace it with a skyscraper, which the older foundation couldn’t have supported.
So hop on for a ride, it’s going to be a good one.
About (Author Profile)
Vygantas is a former web designer whose projects are used by companies such as AMD, NVIDIA and departed Westood Studios. Being passionate about software, Vygantas began his journalism career back in 2007 when he founded FavBrowser.com. Having said that, he is also an adrenaline junkie who enjoys good books, fitness activities and Forex trading.
Subscribe
If you enjoyed this article, subscribe to receive more just like it.
Good.
Though as it stand now. Opera lost its only avail on ibm pc’s: low resource consumption. And with looks different their users will shop else where.
Fortunatly Opera want to keep Opera 12 around :)
With Chromium & Opera devs working together, I’d be quite happy for Opera’s resource optimization team beat Chromiums bloated butt down to a beautiful and manageable size considering how well they’ve done over 1-2 decades. I’m HAPPY Opera has Blink underneath it right now, and I’m quite hopeful and earnst to see how beautiful Opera devs working in Blink/V8 can achieve together to put our beloved Opera paint/body job on the new beast engine which powers it. I’m also in for this ride even if its for the fun but my guess is that it’ll be more than fun to say the least!
Give them time, no one judges a baby child’s life success based on when it’s newly born, it needs to grow and age a little before it can build it’s life and career to change the world.
More like Chrome engineers ditacting to Opera. Have you seen the chippy back and forth that simmered for years between Apple and Google
engineers in the Webkit developers threads leading to the Blink fork?
They
say it is open source, but they maintain iron control on what goes in
and out, probably more so now b/c they forked it and don’t have the
Apple engineers who were probably the only people who had the technical and political clout to stand up to Google.
This all becomes obvious bullshit once you realize old code doesn’t rust, it gets better as you fix bugs. Rewritting will only bring more bugs.
I assume you use a bug free version of DOS then?
I agree with Tiago. You should read this: http://www.joelonsoftware.com/articles/fog0000000069.html
“The idea that new code is better than old is patently absurd. Old code has been used. It has been tested. Lots of bugs have been found, and they’ve been fixed. There’s nothing wrong with it. It doesn’t acquire bugs just by sitting around on your hard drive.”
I’d also add: since code was written long time ago, there might be certain features that are hard to implement or require some workarounds. For example: when writing a code 10 years ago, it would be hard to imagine that you’ll see WebGL, etc. in the future builds. So when you continue adding one workaround after another, it becomes a clusterfuk and it breaks things.
Another example is Windows shift to NT kernel. They didn’t do that “just because”.
Opera devs themselves said that fixing scrolling requires some deep code changes that are very hard to do. Why? Who knows, probably because when they coded Opera 8 years ago, no one thought that we will have sites with hundrds of pics, heavy js and css? I imagine same is true for webgl, etc (it’s been 2 years in the making and latest opera build still is unstable) for the reasons mentioned above.
“probably because when they coded Opera 8 years ago, no one thought that we will have sites with hundreds of pics, heavy js and css?”
Try out an Opera 9.00 and watch. There are no position:fixed bug. They messed up anywhere 10.5x-11.0x version.
You mean, when they added support for something new which Opera 9 didn’t support? Should they just remove that to fix scrolling, then? Of course not. They would have to do it the hard way.
Let me get this straight. Because “there might be certain features that are hard to implement or require some workarounds”, you “work around” the WHOLE code to implement…
wait for it…
…LESS FEATURES?
Bah, I know what you mean, but I’m just not sure you fully grasp the situation here. Rewriting from scratch is a bad move for Opera as we know them. If they want to be something different (including, possibly, a failure), then you can never say rewriting is a bad move. But that’s not what we’re arguing here. Is Opera going to be something completely different from now on? YES. Is it a bad move for the company? I think so.
So, what I mean is, rewriting is a bad move for the company.
And yes, it’s only gonna bring more bugs to fix.
You have failed to show how this is a bad move for Opera. Do you have any numbers yet? Whether this is a bad move or not will only be apparent months from now, when the people who rage-quit Opera no longer affect the stats, and when it’s possible to tell whether new users are sticking with Opera.
Why will having a smaller code base with fewer features bring more bugs to fix?
That article is irrelevant to Opera 15. The quality of the code might be excellent, but sites still didn’t work. And all those features were creating a maintenance nightmare.
It doesn’t acquire bugs just by sitting around on your hard drive? No, but Opera’s code isn’t just sitting around. It’s being developed which means that new regressions appear all the time (just like all actively developed software).
You should not link to an article you don’t understand.
That somehow working the whole thing from scratch is going to produce less regressions and take less time and resources than working on certain aspects of a code base is completely incomprehensible to me. As a developer!
To be fair, though, the whole thing boils down to Opera just freaking giving up and going “fuck this sheet, it’s too hard, let’s just be another Chrome clone!”
I can relate to that. And, from what concerns me, Opera won’t be missed, now, or ever.
Opera isn’t starting the engine from scratch, so they will have fewer engine regressions and it will work with far more sites. They’re starting the UI from scratch like Firefox and Chrome did. When you have fewer features and a smaller code base that is obviously going to take less resources to maintain.
You still don’t understand the article, and you don’t understand coding.
Opera isn’t giving up. What they are doing is the opposite: They are being realistic and doing what they need to do to survive.
They aren’t making a Chrome clone either. The UI is already different, as is the feature set, and this is just the start.
Opera 15 is WAY more stable than Opera 12 ever was. I have been using it for days and it’s not hiccuped once. Every page has rendered perfectly, and I have working hardware acceleration, I have it working aross all my CPU cores. Opera 12 never really managed this.
I now have a fantastic base from which the developers that made Opera so good (given the dodgy base), can move forward with.
I too am expecting great things from the new Opera Webkit base.
Opera 12.15 makes better use from Gpu than opera 15. I used afterburner to see gpu usage during benchmarks. and compared.
For example, on Peacekeeper, one of the final test (crafty), Opera 12 was giving between 50-55fps, and opera 15 between 46-50 fps.
If you have a crossfire setup you will see that the gpu usage will vary between the two cards.
Gpu usage when doing normal browsing depends.
Normally hardware acceleration and webgl comes disabled by default in opera 12.
If you turn the settings to 1, it will occasionaly use gpu, but if you turn to 2 it will always use gpu,
And if you are watching videos on youtube with setting in 2; and if you open another tab and load a website you will notice that the youtube page will freeze momently because, gpu usage will be very high… I have 2 ati hd6950, and gpu usage when browsing, stays between 30-99% ( 99% happens more when you load a site, then when scrolling gpu usage stays between 30-70%).
Changing Renderer (preferred renderer) in opera 12 will influence browser experience.
I have found that in my system, using Opengl is smoother than directx, but with directx loads pages faster.
Opera 12 is stable on my Pc, the only problem i have found is related to facebook, scrolling is not smoother as Firefox Aurora, and opening some websites that have social plugins make it crash ocasionally , or increase load times, and that is solved with extensions.. Otherwise I can´t see diference in load times, that can make me change browser.
When it comes to “re-writing” only modernising the UI and various interface is being re-hooked up to the old logic which can also catch bugs they never caught before by making sure implemented functionality is as bug-free as possible from this but of course can cause new bugs too. UI code is not that complex in terms of logic its just layout lol. If anything their preventing an excess of bugs by merging 2 major engineering feats instead of having 4 separate projects (rendering engine and possibly javascript engine) into a joint venture allowing more great minds to drill the bugs out of anything already existing since the beginning of KHTML which by the way started in 1998… vs our recent Opera+Presto with Caraken from 2009. So the Old code thing doesn’t hold since KHTML(Blink’s) roots are now older than our modern Opera we used up till now, which is weird in my opinion.
Old code might not rust, but more and more legacy features adds to the complexity and makes maintenance harder and harder.
Let’s assume Opera is dead, and enjoy the rest of our life!
http://www.youtube.com/watch?v=Tldf1rT0Rn0 Is a video of some KHTML history also, if you think about it… Opera is now a ~95% open source browser instead of closed source!!! :D It’s almost Identical in situation as Chrome is to Chromium so if you bash Opera now you must also hate the “Closed Source” Chrome parts vs the completely open source Chromium software package.
I’m actually extremely happy with Opera Next so far. It’s super fast. It’s true, they removed M2, but mail stopped working right for me a long time ago mostly because the mail providers I used were blocking external access of their boxes through anything other than a browser window. I know some peopel are upset over the loss are RSS feeds, but again, another feature I didn’t use. Link will come with time. Right now, Opera Next is looking a HELL of a lot better than Opera for Android is (have we taken a peak at those 1 stars recently?) It’s probably because Presto on desktop was the cruddiest of the engines, and Presto on Mobile was one of the bestest of the engines.
Presto on desktop and mobile was the same engine, dear child.
Alas, you miss my point. What I’m saying is — when you put Presto on Desktop next to other desktop browsers, the engine is not too good. But when you put Presto on Mobile up against other mobile browsers, the engine looks freakn’ rad in what it did and how smoothly and well it did it.
That is why I’m using Opera Next right now and feel like this browser has made a quantum leap in smoothness, but I’m using Opera for Android on my Samsung Galaxy S II and feel like I’ve gone back in time to some browser on one of my old Blackberries.
Opera committed suicide. They are trying to copy Google Chrome all the way down to the design (ugly blue background default). You mean to tell me implementing the Blink comes automatically in its code of Google Chrome menus? clear browsing data? ect. Im assuming going chromium it limits the choices of freedom in code. They know exactly what their doing. Going after that Google Chrome fanbase. Maxthon uses webkit core based and it has its own identity. This is what Opera should of used Instead of chromium webkit. And having the ability to download chrome extensions bridges the clone gap even more. This new so called Opera has no backbone and will forever be in the shadows of Google because of this.
Why did Opera commit suicide? You don’t even know if the new browser has succeeded yet.
“broken out of process plugins”
Have you ever seen plugin container in Firefox? Obviously the PPAPI better with a PPAPI optimized YouTube than oursourcing NPAPI-based plugins into separate process.
“we hope that Opera will listen to its community and add features that made us love Opera in the first place.”
I’m old enough. I don’t believe in Santa Claus or the Opera developers who “listen to its community”.
When Daniel Aleksandersen responded to a user’s comment “I want all my features of opera 12 back” with: “I can say for sure that that is not going to happen. Have you seen some of the new stuff? The downloads experience should be much better now, for instance. We’ve focused on the core experience of web browsing.”
I can’t believe in Opera anymore…
The experience bit is awesome. Next time I go shopping, I’ll go get a BJ instead, the experience is much better!
thefreeman55: “i want ***all*** features of opera 12 back”
daniel: “I can say for sure that that is not going to happen”
Obviously not all will come back, only the useful ones.
Original: http://www.favbrowser.com/download-opera-15-next/#comment-911991797
The features that one person finds useful may not be the ones that someone else values.
Along with the ability to customise the browser, one of the things I valued about Opera was that it had a lot of unique little features. Not necessarily things that everyone used, or even things that I used every day, but features that could be a nice time saver under certain circumstances.
For example, being able to tile two tabs side by side, and even make one of them a “follower tab”, so that it opened any link clicked in the other tab. Probably not something that most users even realised they could do, but something that I’ve found really useful. That’s the kind of unusual feature, unique to Opera, that I can see being lost in this transition.
Listening to the community doesn’t mean that everyone who has an opinion gets his way. It means taking the overall feedback into account when making decisions.
I liked your article. I am also optimistic about opera. The attached picture clearly showed the point. Thanks for this.
To the author, you’re in dream land.
“The foundation is old, rendering engine is a trainwreck and the whole Opera architecture currently looks like this”
You have no basis for saying this, unless you have been a core developer in Opera for a long time and had some code ownership.
“Since Opera developers no longer have to spend so much time patching individual web sites and fixing broken architecture, resources can be shifted to adding new features and overall polishing. This also means faster release cycles and quicker implementations of new web technologies.”
Right, the release cycles are needed because they now follow Chronium which has a rapid release cycle. Implementation of new web technologies? You mean getting them served on a silver platter form the Chronium guys and gals.
Regarding resources being shifted, sure, some were probably shifted, a shit-load of core developers got fired during this transition, remember? The whole company is turning into an advertising company, just have a look of their latest acquisitions. For people already trusting Google with their mail and android accounts, there is really no need to give away more of their browsing data to another data-mining ad company just for browsing using a different skin and a few other features here and there.
“a shit-load of core developers got fired during this transition, remember”
Wrong. 90 people left in total. Less than half were engineers. Opera has several hundred engineers, and the core department was supposedly the biggest. That means that in a worst case scenario, a tiny part of the core department left.
I agree!
Starting from scratch is almost always a failure in application development. It took Netscape/Mozilla something like 10 years to deliver a half-decent browser (Firefox 3.5+) after they moved to Gecko. Have fun waiting.
Well, Firefox 1.5 and specially 2.0 was miles ahead of IE and Safari. Can’t speak for 1.0.
Firefox started from scratch the same way Opera does now, building a new UI on top of an existing engine. (Firefox did not take 10 years before it became a success. Firefox started because of the Netscape/Mozilla problems. Like Opera 15 is because of the Presto problems.)
Chrome started from scratch the same way Opera does now, building a new UI on top of an existing engine.
You were saying again?
Nice message and I agree with it, liked the new Opera. Waiting to see old features getting ported to new Opera
couldn’t they just change the rendering engine and leave the UI and features the way they were (i dunno, port them)? this was actually what people have been talking about that “Opera is just going to change the web renderer engine, not the whole thing.” but now looking at Oper 15, i dunno why they chose to do this and give up on every single feature they’ve developed for the past 15 years! everything that made Opera what it is, is gone.
actually most of the people didn’t give a shit about “presto”; the advantage of Opera was having lots of integrated features, advanced options to control every single setting in the browser, and a fully customizable UI like this: http://i.imgur.com/MVfgoat.jpg
If the UI layer depends on the engine it probably takes as long to port it to a new engine as it does to just do it from scratch (or maybe even longer because of all the bugs that would inevitably appear because of the old Opera’s crazy complexity of settings and features).
You’re complaining about Opera 15 as if it’s the end of the line but in reality it’s just the start. Complain when they’ve had more time to do stuff.
“awful scrolling performance”
Disable smooth scrolling and it has the best scrolling in any browser ever and use the least cpu power at that. This is one of the few fear I have with Blinkpera and it is indeed proven in the next build, I just hope it improves.
Sounds good, past, present and future. I think it would be the last step.
Noobs like you see Opera 12 like that, idiots see chrome like in the second pic and they see chrome everywhere.