Bruce Lawson's personal site

Opera and WebKit: a personal perspective

I expect by now you’ve heard that Opera (my employer for the last four and half years) has announced that its browsers will, in future, use the WebKit rendering engine. I wrote the announcement, and what follows here is my personal take on it. It’s on my personal blog precisely because it does not reflect the opinion of my employer, wife, kids or hamster.

Opera’s Presto engine was a means to an end; a means for a small, European browser company to challenge the dominance of companies who, at that time, hoped to “win” the web through embracing, extending and extinguishing web standards.

Presto showed that it was possible to make a better browser while supporting standards. Other vendors have followed this path; the world has changed.

These days, web standards aren’t a differentiator between browsers. Excellent standards support is a given in modern browsers. Attempting to compete on standards support is like opening a restaurant and putting a sign in the window saying “All our chefs wash their hands before handling food”.

Rendering engines are now highly interoperable – largely due to the progress commonly known as “HTML5”, begun by Opera in 2004, then joined by Mozilla, in order to protect the web from proprietary platforms, keep it open and promote interoperability.

It seems to me that WebKit simply isn’t the same as the competitors against which we fought, and its level of standards support and pace of development match those that Opera aspires to.

It isn’t run by a single organisation; a report on WebKit this month says “it is also noteworthy how the diversity of the project is increasing, with new players starting to show a significant activity.”

It therefore seems silly to compete against it. Instead, we’ll join and use our experience and resources to improve it further.

Although a small organisation, we’ve always played an active role in developing standards – CSS, Media Queries, HTML5, native video being high-profile examples. This is important to me; I’ve worked in my own small way for 10 years now to help protect and advance the web and want to work for an organisation that does too. So when it was announced internally that we would switch to WebKit, I worried that standards work might stop.

I asked the CEO and Engineering lead at an all-hands meeting if we will continue that work. They replied that we absolutely will continue to work on standards, and we’ll submit changes to advance WebKit. Our CTO, Hakon Wium Lie confirmed it by demonstrating internal WebKit builds that have some interesting new standards support. Today we contributed a small, symbolic patch that can bring all WebKit browsers’ CSS multi-column support to Presto’s level.

One rendering engine will go. Some lament that. Some of those who lament it seemed never to test in it, excluded it from their demos, or actively blocked it.

I’m both English and a man. That means I have no emotional life at all (so consider this carte blanche to be incredibly rude to me in the comments) but even with those two significant handicaps, I’ve found myself with a pang of regret that the Presto rendering engine will disappear. I’ve experienced that feeling before – eighteen months ago when having a final walk around the house that had been my family’s home for a decade, before getting into the car and following the removal van to the new home that we’d dreamed about.

Of course, a browser is much more than a rendering engine. Very few consumers of the web choose a browser because of its rendering engine – they just expect it to work. And if it doesn’t work as well as native apps, they’ll choose native apps.

Opera has 300 million active users —almost a third of a *billion* people— many of whom would otherwise have no access to the web. For many users around the world, a browser is more than a tool to browse the web. Sometimes it’s a school when you can afford none, sometimes it’s the only line to an outside world shut off by an oppressive regime.

The web needs to win. Browsers are highly interoperable, because all vendors know that if they’re not, they risk being overtaken by proprietary platforms. It used to be Flash and Silverlight that threatened the web. Today’s threats are proprietary app platforms and locked-in “eco-systems”. Tomorrow, new threats will rise.

Developers who care about the web will code to the standards, test across browsers and block none. We all want the same: we want the web platform to grow, to remain open, to become ubiquitous by being the no-brainer development platform of choice for all.

(Last Updated on )

Buy "Calling For The Moon", my debut album of songs I wrote while living in Thailand, India, Turkey. (Only £2, on Bandcamp.)

62 Responses to “ Opera and WebKit: a personal perspective ”

Comment by altarius

you should simply give folks a build to play with and most complaints will disappear 😉

but as webkit has it’s own devtools, is there a posibility to keep dragonfly alive and running with the new opera? the simplicity especially with remote debugging would be a great loss

Comment by Steve Fenton

Although I too have written a lament, I am rather hoping that such scribblings will be read and taken into account to ensure those concerns don’t come true. If we still get innovation and we don’t end up with lots of forked WebKits then hooray.

I actually use Opera. I use o prefixes. I demo’d a web app in Opera this week and even put an Opera screen shot of the app on Twitter. I hope to be proved wrong. I really do.

Comment by Charles Roper

This reminds me of when a struggling yet innovative and important computer maker (hint: rhymes with schnapple), only days away from bankruptcy, decided that in order to make best use of its resources it must switch its OS to an open source variant of Unix. That, as it turned out, was a pretty good move. Unix continues to be diverse, relevant and vibrant, as do the competitors. And it allowed that struggling computer maker to focus its capabilities where, had it stuck with it’s proprietary OS, it would have surely been a different story.

If WebKit means Opera can continue to exist and be relevant and innovative, then it’s a good decision.

Comment by Jameson

I am one of those few who chose Opera for the rendering engine. In fact, I chose it because it had small market share. This switch removes the only thing keeping me here. Opera is a great browser, but this change just makes it a closed source version of Chrome/Safari/Midori with a few more bells and whistles.

Please, do what you can to release Presto to the community. It may not go anywhere, but at least it won’t just be thrown away.

Comment by Charlie

“I’m both English and a man. That means I have no emotional life at all..” great post, but this one line is the best.

Comment by Tiago Silva

What about Opera’s features that depended on Presto to function? Are you porting all those to WebKit as well? Will Dave Hyatt be up for it?

Comment by Maurizio

I think the bad thing of leaving Presto is the work which produced it – years of development, experience, skills have been spent; throwing it away, you feel like that work was wasted. I hope you will open the sources, at least the Presto spirit would remain in the Web 🙂

Comment by Phil Wills

@Maurizio if, as Bruce argues, the existence of Presto over the past decade and a half has been a force for keeping open, interoperable standards prevalent on the web, then regardless of the fate of the code itself, that’s a legacy many engineers would love to be able to claim.

Comment by Evan

“I’m both English and a man. That means I have no emotional life at all..”

No comment the English part — but just because _you_ don’t have an emotional life doesn’t mean the rest of us men don’t…

Otherwise thanks for writing this. It clearly was a hard decision and I hope the best for Opera going forward.

Comment by Me

Well well. I will be honest and direct though. When I did switch to Opera some ten years ago, it was indeed not for the rendering engine. It was for, at that time and still today, incredible smart and helpful built-in extension which Opera’s main competition targets of those days not even thought about at any time and even just rising Mozilla would not even provide with third party plugins. Be it CTRL+11, be it a complete per-site configuration for JS, plugins, cookies and whatsoever, be it the badly needed assistment in disabling most of that bugging Javascript eyecandy playfoo so called web designers felt like needing to add to their customers’ sites in order to make themselves memorable and show what crazy (and crazily superfluent) annoyment they were capable of coding, be it seamless page zoom, be it minimum font sizes for people like me never wanting to suffer from eye cancer after browsing these creepy {font-size: 4px; color: #ccc; background-color: #fff} sites or be it custom page styling beyond all limits.

Still I haven’t mentioned all the things which let me endure even the devastating changes introduced in 11.5 which, IMO for the first time then, made a clear yet implicite statement about Opera A/S’ upcoming strategy (if you like to call it so).

I understand clearly that, at the end of the day, even a small and, at least at some long time, obstreperous company has some bills and wages to pay. And obviously the former idea of refinancing the free-of-charge frontend clients with backend infrastructure (i.e. mobile and stuff) did not work out as expected. Then, as one might guess after not hearing much more of it in the past months, the Facebook deal seems to have done just the same. Thus, I suppose that there were not many choices left. But unfortunately this does not help it – I fear, that the Opera era (in a definition like the one I posted above) is over by right now. And that makes me really sad. This is because I find it hard to believe that the core change will be it. In the long run, I am pretty sure that a fundamental design change will follow as most folks yet unfamiliar with Opera but now being target group will wonder why Opera doesn’t support all these cool plugins other webkit browsers do. Sooner or later the question will rise why all the money is put into development staff dealing with the built in goodies I mentioned while there are plenty of (and, that statement will come for sure, “very up-to-date because maintained by so many volounteers!”) plugins doing just the same (“and hey, we could reduce code size and server load with smaller install files!1!”).

While all this may seem “without alternative”, as Mrs Thatcher would state, in the end it will not be anything else than what became apparent since a good couple of months: Becoming yet another major browser (“because user count has proven that the others’ concepts turned out better”). And with this, losing any unique feature forever. And, as for me, probably any reason against ending this ten-year-relationship.

I still do hope that I’m wrong, chances are I am. But i figured it better to have someone know it who may bring up these points at the right time and place before it’s too late.

Good luck with that and all the best!

Comment by operamaniac

I hope for you to submit a patch to webkit that allow users to select linked text. Opera seems to be the only web browser allowing that. It is not exactly a killer feature but one of the examples where Presto excelled!

I am looking forward to Opera powered by WebKit to see if it would enable me to switch permanently to Opera. Right now, I run Chrome and Opera together. Opera to read and Chrome for complex web apps. Cheers!

Comment by tom jones

there is a subtle, yet important error in your logic:

“It seems to me that WebKit simply isn’t the same as the competitors against which we fought, and its level of standards support and pace of development match those that Opera aspires to. “

i remember the IE6 days. its level of standards support was pretty good (the best?) at the time it was released, and the pace of innovation leading up to it was so staggering that it was referred to as “the browser wars”.

“They replied that we absolutely will continue to work on standards, and we’ll submit changes to advance WebKit.”

and if apple doesn’t like those changes, they will not adopt them in iOS webkit, and you still wont be able to ship iOS Opera with those features, like say Chrome/iOS can’t use V8 (and has even slower JS than the default because no Nitro), nor can it include other new web app features like WebGL, IndexedDB, WebRTC, etc, etc..

and how many developers will adopt those new features and build sites that use/depend on them if they are not available in iOS browsers? i’m guessing even less than the number of developers testing/optimizing for Opera today.

so what exactly are you gaining with this move again? ah yes, ability to ship Opera on iOS. but how does that benefit the open web?

“The web needs to win. Browsers are highly interoperable, because all vendors know that if they’re not, they risk being overtaken by proprietary platforms.”

except when the ratio of (major) vendors that also have an interest in (their) proprietary platform winning over all the others (including the open web) grew from 2/4 (apple, ms/opera, mozilla) to 2/3 (apple, ms/mozilla).

again, as explained above, you can’t really expect for Opera to have influence on webkit, or say webkit is not controlled by apple. if they feel it’s opposed to their business interests, they don’t ship it, and developers don’t use it, it’s as simple as that.

and you can’t say that’s a good turn of events..

Comment by Spudley

I’m surprised and yet… not surprised. I guess more accurately, you could say this caught me off-guard.

Opera is in a no-win situation with the Presto engine. The amount of dev work being poured into Webkit by Apple, Google and others is phenomenal. This is great news, but clearly Presto was never going to be able to keep up with that level of change, given that it’s closed source and only has the dev resources of a single, relatively small company. Gecko is only keeping up because it’s open source, and Trident… well, MS have basically committed to spending as much as it takes to keep it relevant.

No matter how good the Presto engine is (and I have a lot of respect for it), all of the above means it’s in a position where it can only slip further and further behind.

Therefore, I fully understand the move. It does leave a tinge of sadness, but it’s a pragmatic and necessary step.

It feels like it would be good move for the engine to go open source. It always feels harsh when a really good quality codebase that people have put so much time into is ditched.

If it had gone open source some time ago, I feel it might have had a fighting chance of keeping up with Webkit and Gecko. As things stand, it probably doesn’t any more (particularly as Opera themselves won’t be pouring resources into it any more), but as an open source project it would make a great resource for developers wanting to tinker.

Comment by Charlie

Obviously a difficult decision. I share with others a sense of scepticism that might lead to a Webkit-monoculture. Maybe we’ll get a duopoly as I fully expect IE 10 to be the last browser from Microsoft based on their own engine. So it would be great to see Presto released as open source to allow people to experiment with different implementations for specific features – maintaining a rendering engine is a huge project.

But let’s hope, as you say, that there will continue to be sufficient creative competition amongst Webkit members to drive innovation and not recreate the stop energy of the W3C in the 2000s.

Comment by Me again

@all/”moving forward”
Of course it is obvious that the web is a moving progress. But blaming the W3C as a stopping instance is just ridiculous. If it weren’t for them, we were still at doing nasty browser hacks and everyone trying to set new standards. What annoys me is that most guys are way too (pseudo) feature-driven. Not every new idea is worth extending existing standards, au contraire most of them are not worth the bytes being posted just a few weeks later (not to call it mindshit). Standards is much about consistency and reliability and understanding, that far not anyone is willing (or, at all, able) to “update his browser”.

So my major concern is still the same as it was when switching to Opera long time ago (and what Opera dropped more and more already): I *want* far longer standards cycles. OK, CSS3 with over ten years is also no good thing. Agreed. But really: No one in the whole damn world does need new extensions every six months and more. And most of those who claim to do are, IMVHO, just not capable of presenting what they want to with the means given at a time.

However, complaint is almost useless, as we are already facing another era with much to do for web developers – not too bad for the dev inside me, as this means safe cash flow for many years to come. But as a user, I feel already today just left behind. On the other hand: I always did, despite being blessed with Opera. So nothing too new about it.

Finally I want to add to #28: Please, please. Make presto open source and give it a chance to live on and drop any selfish thoughts like “but all our current 300 Mio users would go away then”, because that they will probably anyway.

Thanks!

Comment by Thorn

Still don’t see any logic in that move to WebKit. The one possible reason is “spaghetti-crutch” engine, which cannot be improved w/o rewriting all the code (and [low] quality of that engine proof that).
Despite any reason, if Opera change gears, it becomes just a toupee above WebKit. With completely different protocol (internal API) and hell knows which features – definitely not all of ’em can be easy ported to the new engine.
Your main problem is not web standards, but management – far from IT and blind for future. People who leaved your boat felt that.

Comment by Davyd McColl

Whilst I 100% get the concept of LEAN development (which would support using WebKit) as well as contributions back to WebKit (which it direly needs — the amount of high-impact orphaned bugs is astounding), I’ll be sad to see the Opera rendering engine go, primarily because, unless Opera plan to distribute a patched version of Webkit to Android users, I’ll be back to all the buggy rendering that the default Webkit-enabled browser on Android gives me. As a user, I don’t see the benefit in that at all :/

Comment by Bill

I wonder what will be the reason to use Opera now. If on apple, it uses apple webkit, and on others it uses Chrome, why not just use Chromium?

Although I am a fan of Opera’s mail feature built into the browser, other browsers have already adopted tabs and speed dial, for me the only other “addictive” features besides mail.

Comment by NNM

Good riddance if you ask me (I know you didn’t ask me).

I never even bothered testing things I make on Opera. Not even once.
Seriously, 300 million? I think that number is fake.
No one uses it voluntarily. Most Opera users had it pre-installed and never knew they could change it. That’s most Opera users grabbed by bloatware & crapware practices…
No serious business (massive corporations, industry,..) uses it as their browser.
The stats on my corporate webs confirm <0,1% Opera.
Those few hits were people testing it.

So, in summary of my rant:
Good riddance, I hope Opera fades away now, and that I don't have to uninstall it from bloatware loaded computers.

Comment by Opa

Comments like the one before just prove one thing: Opera has always been an elite browser for people wanting to gain a control level over content most others not even missed at all.

Unfortunately this property was self-destructive. The masses never felt a need to change and, thus, never realized the advantages Opera offered, and the few lads recognizing just that would never ever make some kind of critical mass for whatever business model.

This is why I, too, hope we will see the Opera idea live on in an open source fork. Really from leet for leet. I mean it.

Comment by Flans

Seems as if it lasted 37 post for a troll to appear in the 38th….

I’m an Opera user since 1997 and am on the fence on how to judge the decision to use Webkit instead of Presto in the future. I hope Opera still can be for me what it was for 16 years: The most comfortable and secure browser I could get at any time in this period.

Comment by Matt

Hi Bruce,

I think transition to webkit is a good move. I’m concerned however that the features that I use everyday are going to vanish. M2 (Mail client), Dragonfly, tab stability, feedreader, UI customization, keyboards shortcuts. etc these are reasons I use Opera. Any comments of what will become of the of feature set?

-Matt

Comment by NNM

Mr 37 is back.
As Opera moves to webkit, there’s just no reason left not to use Firefox instead. It’ll just do whatever Opera does, better.
Mr 40, I’m not a troll, I’m realistic. You say you’ve used it since 1997, that makes you a biased fanboy. The main bulk of Opera users come from devices that have it pre installed. It’s not there by user choice. It’s there by default, and then tech people like me have to help them get rid of the poison.

Parallel: just like the gMail icon on phones and tablets… How many times have I tried to explain to them that gMail is not eMail, so, of course, you can’t read your other inbox there. They register a gmail account because they are told it’s the only way after clicking the gmail icon; and the evil empire recruited another ad target.

“Ask toolbar”. How many users do you think they have now? Does that make them successful and trustworthy? Is their number of “users” something to brag about or be ashamed of? Same technique as Opera has used.

Comment by Bruce

@NNM

“I never even bothered testing things I make on Opera. Not even once.”

good job you’re not a web developer, then!

“Seriously, 300 million? I think that number is fake.”

right.

“No one uses it voluntarily.”

Now you’re just being silly.

“No serious business (massive corporations, industry,..) uses it as their browser. The stats on my corporate webs confirm <0,1% Opera."

And your corporate website = the whole world, right?

"Those few hits were people testing it."

fascinating. Can you share how you know what people's intents were when they surfed? (You could make a fortune from the profiling/ ad industry with that kind of analytics)

"Good riddance, I hope Opera fades away now"

why? Because you don't like it and don't use it, so nobody else should?

Comment by Bruce

Matt

I can’t comment on features of the new products (because they’re still being worked on and also because our PR department are evil, taser-wielding psychos who’ve kicked me before about it).

But the aim, after integrating the new engine (which,as you’d imagine, isn’t trivial after 18 years of development of Presto), it to greatly improve UI and features.

Comment by ledahulevogyre

Authors should code to the standards. But some don’t. They only care about the browsers their visitors use most. It’s sad but understandable.

Browser makers should support standards. But some don’t. They include proprietary stuffs to make attractive products (-webkit properties). It’s sad but understandable.

Users should use standard compliant browsers. But some don’t. They use browsers that work with their favorite websites. It’s sad but understandable.

As David Storey writes, “we all failed in some capacity”.
Why ? Because of short term interests ?
David continues with “We all need to do better, starting today”. But, come on, that’s what a lot of people do all the time. (and what Opera did until today, imho)

Has this capitalist idea of competition considered as a good thing reached its limits ?
Apparently, it’s very good for innovation, but not so good for interoperability. I personally don’t care about new shinny features that are not working on every browsers.

If we are about to promote Webkit as standard, how do we ensure it’s not controlled by a single corporate ?
And how do we prevent stagnation (on bugs and features) ?

Comment by Matt

@Bruce

Thank you for the response. It’s an exciting time for Opera and I’m looking forward to Opera’s contributions to webkit.

Hopefully features like M2, and Dragonfly will survive the transition. 🙂

-Matt

Comment by Rudolf

It is a lie that Opera just changes rendering engine from presto to webkit. Opera will use Chromium. In fact it will be Chromium itself, not Opera. All key features of Opera will be killed. Such as:
1. Deep customization
2. View cached images only
3. Fit to width
4. Horizontal scroll by mouse wheel
5. Fast back in history
And more.
We don’t need yet another Chromium browser with changed icon!

Comment by Questo

If #51 is true (which I will thouroughly test before “updating” my production installs), then 12.14 will be the last Opera I have used until the original project actually goes open source. Otherwise I might not bother forking my very own chromium engine and building my very own browser and, while that will take time, sticking to old Opera installs as long as possible.

Comment by sigh

This “NNM” character doesn’t seem to get it:

Seriously, 300 million? I think that number is fake. No one uses it voluntarily. Most Opera users had it pre-installed and never knew they could change it. That’s most Opera users grabbed by bloatware & crapware practices…

No, the vast majority of those 300 million users are people who installed it themselves. Opera Mini, Opera Mobile or Opera on a PC that they downloaded and installed manually.

No serious business (massive corporations, industry,..) uses it as their browser.

Except mobile operators across the globe? And of course Sony, Nintendo, etc.

Comment by sigh

I actually think Opera is the only desktop browser that hasn’t done any bundling. Chrome is bundled with Flash and other apps, and Firefox had some bundling deal a few years ago as well. IE is bundled with Windows.

Apparently Chrome, Firefox and IE are all bloatware and crapware. Says NNM.

Comment by sigh

@Rudolf

It is a lie that Opera just changes rendering engine from presto to webkit. Opera will use Chromium. In fact it will be Chromium itself, not Opera. We don’t need yet another Chromium browser with changed icon!

Your evidence that Opera will just used Chromium and not have its own UI is where exactly?

You do realize that just because two browsers use the same engine and framework doesn’t make them identical, right? Different WebKit browsers don’t even have an identical WebKit.

Comment by Sad

I know this if off topic but I’m looking into starting my own blog and was curious what all is required to get setup? I’m assuming having a blog like yours would cost a pretty penny? I’m not very internet savvy so I’m not 100% certain. Any tips or advice would be greatly appreciated. Many thanks

Comment by Simon

As business consultant, Opera movement is the last and worst decision, the same than Nokia when Windows were chosen, both caused by a bad management for years and a desperate need to save someone’s position by cutting costs and leave the responsibility to others.
As a developer, there are only a few real Open Source projects, neither Webkit nor Chromium are included, they follow their main supporters agenda (Those who pays for full time developers as was seen with OpenOffice or MySQL), indeed, Opera has nothing to do against Google and Apple, specially when Google has been its main source of money (At least for Opera desktop).
Why using Opera if Chromium is made by Google?
Why keep using Opera when it lost all the nice features that it use to have (As it happened with Eudora)?
Why working with Opera if Google and Apple have a more reliable management and brand?
Norwegians once were good to make Internet technology but very bad to make business with it. Google won against AllTheWeb, now is the same with Opera. In other words US(2) – NO(0)

Comment by NN

@Simon

Comparing this to Nokia/Windows is a mistake. Nokia decided to drop their own market-leading platform (Symbian) for a failed platform which was being rejected by the market (Windows Phone).

Opera, on the other hand, dropped a platform that never was a market leader (Presto) for a platform that is (WebKit). The opposite of what Nokia did.

Why would cutting costs be a desperate move? Opera is doing well both financially and user growth wise. The move to WebKit was announced at the same time as the announcement that Opera had grown to 300 million active users.

The rest of your comment reveals an equally poor understanding of things.

As a business consultant, you evidently have a lot to learn. Please tell me the name of your company so that I am sure to avoid it in the future, as your understanding of the market seems poor indeed.

Comment by Sergio

3 lb (525g) Ripe Bananas 6oz (175g) Brown Sugar 40z (100g)
Sultanas 2 Eggs 8oz (225g) Self Raising Flour. Most professional wedding
photographers love what they do and truly want to provide you
the best images of your big day. Sweet tasty cake, all yours for the taking; just be sure you continue on
trying completely new methods of cake decorating to help you keep improving your cake making.

Leave a Reply

HTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong> . To display code, manually escape it.