Developers › A Journal from Behind the Trenches

Poll: What do you think about Opera switching to WebKit?

February 13, 2013 · by Haavard

I have added a poll to the right sidebar of this blog: What do you think about the announcement to switch from Presto to WebKit?

Feel free to expand on the reasoning for your vote here.

Readers' comments 169

Left under the original article, as archived.

inDigazzZWednesday, February 13, 2013 11:33:55 AM
If you change only what is below the address bar and above the status bar, then "as you want" ...
PierreWednesday, February 13, 2013 11:36:34 AM
For
```
489069232="In memory of Geir Ivarsøy."
```
and for Jon and Yngve : Sad

For the future : don't know where you go...

For now : excited to see your first webkit things
SushubhWednesday, February 13, 2013 11:37:35 AM
Good for WebKit based browser users.
Bad for folks like me who loved Presto for what it offered.
Probably not very good for the open web. Especially on the mobile front which is now largely webkit other than Windows Phone devices.
Anyways, hard to say until I see how it changes the behavior of the browser while I am using it till you guys release the first build!
FransWednesday, February 13, 2013 11:53:08 AM
I voted "not sure" because for now there simply isn't enough information. I wouldn't like to see things change that I've depended on since Opera 5, 6, and 7, and I've also grown rather fond of newer additions such as user styles, UserJS, and DragonFly. (And Turbo, but I feel confident about its safety. I'd add Unite, but that's already a goner.)
mrdWednesday, February 13, 2013 11:53:51 AM
As long as the browser remains from a UX perspective what I know, then from a user perspective it's fine. Probably a good thing in the short term (all things being equal).

From a wider perspective of how this will impact the web... it's a bad thing, a sad thing. Opera was a hold out and a bulwark against the monoculture that we used to have with IE and now have with Webkit.

With this Opera now becomes Internet Explorer circa 2000. That doesn't make me happy or fill me with joy for where we're going to be in a few years.
ouzowtfWednesday, February 13, 2013 11:55:26 AM
I voted for "not sure" too, because I dont know how much "Opera" will be left after the change.
Martin KadlecWednesday, February 13, 2013 11:56:42 AM
What is good:
- Better compatibility with google services
- Hopefully better support for WebGL
- Possibly more developers working on other things
- Support for text-selection: none and other web features longly ignored by opera
What is bad:
- No more Dragonfly (Inspector is crap)
- Chrome-like extensions (every time I have to work with chrome manifest files I have headache)
- Harder to make new Opera-only features. E.g. if you want to have Speed Dial extension you will have to push it to webkit (so everyone will use it as well) or patch it every time you use new webkit version which would make it very hard to maintain.
- Many long-time Opera devs left Opera (Anne, Patrik, Rijk, ...) and don't tell me it has nothing to do with it.
- Less engines to test when making websites -> harder to figure out what is right and what is webkit bug
- Huge increase in Opera size as webkit is several times bigger than Presto
- Big increase in memory consumption. I wonder how many tabs will webkit-based Opera handle.
FransWednesday, February 13, 2013 12:00:17 PM
Originally posted by BS-Harou:

- Chrome-like extensions (every time I have to work with chrome manifest files I have headache)

Could you expand on that? I don't think XML is very user-friendly either.
Originally posted by BS-Harou:

- Less engines to test when making websites -> harder to figure out what is right and what is webkit bug

Yup. My rule of thumb is that if Presto and Gecko do one thing and Webkit another, Webkit is probably in the wrong. Of course I'll check the spec if I want to be truly sure.
s/ashWednesday, February 13, 2013 12:02:27 PM
I'm not shure.. I couldn't find a single WebKit-based browser that doesn't act weird, the whole thing looks like a window that comes from another OS. At least on Windows where I tried other browsers, as on a Linux distro the first thing I do after installing GPU driver is to Install Opera.

I'm pretty skeptical in terms of Opera's smoothness while on top of WebKit, that being the main reason why I'm using it. :| The second being its flexibility that any WebKit-browser lacks.
FransWednesday, February 13, 2013 12:06:02 PM
Originally posted by s/ash:

I'm not shure.. I couldn't find a single WebKit-based browser that doesn't act weird, the whole thing looks like a window that comes from another OS. At least on Windows where I tried other browsers, as on a Linux distro the first thing I do after installing GPU driver is to Install Opera.

Actually I think Chromium on Linux does a fairly reasonable job at looking native these days. Besides, Epiphany and rekonq are also Webkit-based and you can hardly accuse them of not integrating.
Originally posted by s/ash:

The second being its flexibility that any WebKit-browser lacks.

Yep. I really hope Opera won't let us down there.
wardiesWednesday, February 13, 2013 12:06:53 PM
Originally posted by BS-Harou:

Many long-time Opera devs left Opera (Anne, Patrik, Rijk, ...) and don't tell me it has nothing to do with it.

Even if that were the case (and I've no idea if it was or wasn't) that doesn't mean progress is a bad thing. Perhaps they said to themselves, "my work here is done" and, having helped to grow Opera from a small seedling to what it is now, felt they needed to move on and find new challenges.
Martin KadlecWednesday, February 13, 2013 12:07:21 PM
Originally posted by Frenzie:

Could you expand on that? I don't think XML is very user-friendly either.

JSON is a great things when it comes to size and easiness to read the data with javascript. On the other hand it is harder to read by human (mostly because nobody formats it with tabs and all).

XML is much more descriptive and the size of it does not really matter because extensions are stored locally.

But it is not only that, there is also lots of restriction you have to "undone" before you can do anything in Chrome. I once tried to convert my Opera extension to Chrome and I spent two hours just figuring out how to set everything right in manifest file to make it work.
Michael DitumWednesday, February 13, 2013 12:07:26 PM
I voted for "not sure".

As a general web developer it makes my testing a lot simpler.

As an STB Developer who does a lot of development on Opera Mobile it's a little worrying. Whilst all our apps try to be as standards compliant as possible we have to use a number of opera propriety extensions (like opera.lockImage to get around the browser discarding images from memory) in order to give the user a smooth experience. I assume that features like these will be removed which would affect our applications.

As a developer who's still having to put up with the remains of IE6 and IE7 (damn you enterprise customers!) I'm not keen on one engine getting too much of a monopoly. The fact that it's an open source engine makes it a little more agreeable as it can always be forked but we're already starting to see websites being written just for webkit and it's a slippery slope back to IE 6 era web.

Overall I'd say I'm on the "Bad Idea" side of "Not Sure", but we'll see.
FransWednesday, February 13, 2013 12:17:40 PM
Originally posted by BS-Harou:

JSON is a great things when it comes to size and easiness to read the data with javascript. On the other hand it is harder to read by human (mostly because nobody formats it with tabs and all).

But surely you would format it nicely when writing an extension? I'm afraid I don't follow your objection.
Originally posted by BS-Harou:

XML is much more descriptive and the size of it does not really matter because extensions are stored locally.

I'm not talking about the size; I'm talking about human readability. JSON is hardly optimal, but surely it's better than XML? But really they should use something more akin to operaprefs.ini (read: like just about everything in /etc).
Originally posted by BS-Harou:

But it is not only that, there is also lots of restriction you have to "undone" before you can do anything in Chrome. I once tried to convert my Opera extension to Chrome and I spent two hours just figuring out how to set everything right in manifest file to make it work.

I'm fortunately wholly ignorant on that matter. No doubt Chromium extensions are awful to make, but I just can't believe JSON's to blame.
pseudo555-tafWednesday, February 13, 2013 12:22:56 PM
Not sure as well :

Presto was not perfect, but i really hope that this change won't change all ui features present in opera since years (like Arrange -> cascade feature)
СергейWednesday, February 13, 2013 12:29:11 PM
Not sure. What about smooth scrolling?
}MoTeNaY{Wednesday, February 13, 2013 12:41:33 PM
Here, not sure too. Will I be able to use over 300tabs?
Erik HauboldWednesday, February 13, 2013 12:46:19 PM
as a developer i have mixed feelings about this.

presto is a good engine, in some parts slightly behind some others but nevertheless it's fast, lightweight and always pushes us devs to keep up with the standards (e.g. removing prefixes ...). it's quite sad to see it go.

also there is the dragonfly part. i really like the simplicity of debugging remote with it, be it desktop, tablet, smartphone. i've found no other tools which could do this without much work or extra apps. apart from that i just got used to work with it. chrome dev tools are complex and when trying to debug something i just spend most time in searching stuff.

so after all: if you preserve what makes opera unique, like M2, dragonfly, irc, low ressource consumption (...), i don't have a problem with it in first place.
FransWednesday, February 13, 2013 12:49:56 PM
Unfortunately BS-Harou informed me DragonFly is a goner. ;( https://twitter.com/runeh/status/301616059729969152

But yeah, Web Inspector is almost impossible to figure out. And the remote debugging on your cellphone is the best.
sleepyheadWednesday, February 13, 2013 1:03:00 PM
Originally posted by motenay:

Will I be able to use over 300tabs?

Yeah, that's one of my main concerns too. I usually have tons of tabs open and not having the newest computer with lots of Ram, Opera still handles it very well.
IlgazWednesday, February 13, 2013 1:04:07 PM
Actually, no comment nor even a vote. I will likely use& report issues on Firefox mobile. An engine kind of steered by Apple& google (main contributors) can be too much for me.
I understand the reasoning but Opera compatibility is ages ahead compared to what we had in 3.6 days. I guess that is why many old school users will feel sad.
Timothy WilsonWednesday, February 13, 2013 1:11:42 PM
So many "web-devs" here say that its not good because they have less reference to check Webkit bugs.Thats clearly an abusing intension and is Bad from u guys.
Patkós CsabaWednesday, February 13, 2013 1:26:40 PM
I am not sure this is a wise step. I see certain advantages but I also fear that some features may disappear in the near feature. I am already imagining another message saying that the mail client will disappear, than tab grouping, than speed-dial, etc.

Did anyone actually tried to use Chromium and install a lot of plugins so that it can do most of what Opera does built-in? Well I did, it turned out to be so bloated and slow that there is no way someone could use that.

So, as long as you keep all the functionality Opera has today and change only the engine I say it is a good move. As soon as features will be impacted and Opera already ditched the irc client and torrent client and widgets... I fear it may be on a road to become just another Chromium.
Pharaoh Atem (ニール・ゴンパ)Wednesday, February 13, 2013 1:59:05 PM
I applaud Opera for making the switch, though I wince at the loss of a major layout engine.

I wish you guys were using WebKit 2 rather than WebKit 1, though.

However, I would like to see Opera's Presto code open sourced so that there's a chance of it being kept alive by someone else. Not everyone wants just a WebKit world, after all. And given how much work it would take to move to Gecko, I'm not surprised that Opera chose not to go that route.

Will you guys continue to use GStreamer for the
kangarrouWednesday, February 13, 2013 2:01:53 PM
How much will it take to see the first webkit snapshot?
Anton DiazWednesday, February 13, 2013 2:16:16 PM
Time will tell
prestonwiiWednesday, February 13, 2013 2:44:10 PM
I've been using Opera for over 10 years and for the most part I've been happy with it. The most annoying thing about it is when I find site that it isn't compatible with or that causing it to crash (Google Docs crashed it until recently). I've come close to moving away from it due to lack of certain features (like extensions), but they seem to add them before I do. It is difficult to tell what kind of impact this change will have on Opera as a whole. I'm hoping this will be a change that I don't even notice. I don't really want big changes to come up of this to my Opera experience. I want my extensions to continue working, my speed dial to keep working, the email client still working, the bittorrent client still working, etc.

If this is simply a change that goes unnoticed except for greater compatibility with web sites and even a smaller footprint on mobile devices (because they can use the built in webkit engine) then I'll be quite happy. If it changes the whole user experience and I lose a lot of features, it might force me to jump to another browser.
myownfriendWednesday, February 13, 2013 3:07:52 PM
Horrible idea!

Getting rid of Presto gets rid of one of the bug things that made Opera Opera. It was a reason to open Opera instead of IE or Chrome

Presto was lean, mean, and has been my favorite engines for testing web pages on. If my webpage was displaying incorrectly in Opera, 9.9 times out of 10, its because I screwed up not it. And generally, if it worked in Opera, it worked in other browsers.

Also dragonfly has always been waaay better than inspector.

I understand that switching to webkit would free up resources going towards the rendering engine, but I think the better option would be to open source Presto.

Yes, other browsers might switch to it, but Opera would always implement Presto the best.

I'll be honest. I've told a bunch of people about Opera over the years and its just gonna be another shell for Webkit, then I honestly would see no reason to use it myself.
adioniWednesday, February 13, 2013 3:10:03 PM
Hrrm. Not sure how I feel about this. I suppose if all the major desktop Opera features (single keybaord stroke page and tab navigation, opera link, highly customizable toolbars, etc.) remain---which I *assume* they will---then I guess it is ok.

That said if Opera wanted to transition to a pre-existing, widely support layout engine to free up time for innovation in other areas, I don't understand why it did not go with the Gecko engine.

Monoculture is bad.
mubaidrWednesday, February 13, 2013 3:13:54 PM
As far Opera will have all the features it has now, this will prove good decision

Please keep:
-M2
-real Smooth-Scrolling
-customization
-about:config
-about-plugins
-plugins on-demand
-Interface
-Notes
-Gestures
-and many more

Opera should be exactly same as it is now, only and only engines should be replaced. If you are about to remove any of these features, then keep in mind that there are already many web-kit browsers out there.
Samed AyanWednesday, February 13, 2013 3:14:48 PM
Tenno SeremelWednesday, February 13, 2013 3:28:21 PM
Originally posted by Frenzie:

JSON is hardly optimal, but surely it's better than XML?

It's worse, not better at best. Spaghetti markup :}

As for Opera switching toWebkit… well, I was suspicious of people at Opera side fleeing it. I was right. Oh well. Personally, I think it's bad idea, but I'll give it a try, I guess. One try max, though.

Maybe someone will finally make UI fonts NOT GRAY, ARGHHHhhhh %) Well, I doubt it as it was somewhat recent development, which means… yeah.
minhoWednesday, February 13, 2013 3:31:29 PM
webkit cant open many tabs on the same time like presto.
Terrible decision...
webtaxWednesday, February 13, 2013 3:35:47 PM
voted "not sure"

I'm interested to know what features and behaviour Opera may lose.
morbitWednesday, February 13, 2013 3:44:20 PM
I'm afraid there is more room for things go for worse than the other way around. I'm wondering about distribution model too. Plain tars for FreeBSD were really neat and (comparatively) light on dependencies. Not sure if it's even possible with WebKit...

Chromium is a dependency mess compared to Opera.
FransWednesday, February 13, 2013 3:53:02 PM
Originally posted by tenno-seremel:

It's worse, not better at best. Spaghetti markup :}

I suppose your mileage may vary, but for me it's like this.

The best is some kind of plain-text config/manifest like operaprefs.ini and all the *.conf files in /etc:
```
name = An extension
version = 1.0
```
JSON is like an annoying version of that:
```
{
'name': 'An extension',
'version': '1.0',
}
```
But XML is even more annoying and less readable:
```
<extension>
<name>An extension</name>
<version>1.0</version>
</extension>
```
mubaidrWednesday, February 13, 2013 3:55:08 PM
Webkit bugs would become the standard: there would be no way for developers to test on multiple engines to determine whether an unexpected behavior is a bug or intended.

Considering above statement, this decision might prove to be worst. RIP Presto + Carakan
ahoj1234Wednesday, February 13, 2013 3:57:38 PM
Well, if you can tell me answers to:

http://my.opera.com/ODIN/blog/300-million-users-and-move-to-webkit?startidx=50#comment103786442

than I will vote in your poll My opinion on Presto/WebKit depends on changes what we will have to be respecting.
Mark A.Wednesday, February 13, 2013 4:00:04 PM
Where's the Gecko option?
Sin2xWednesday, February 13, 2013 4:00:51 PM
Haavard, Webkit is only a layout engine - what is going to happen to Opera's Javascript engine? Currently it is lagging behind its competitors speed-wise.

Also, I really-really hope that Opera will not lose any of its configurability and its awesome mouse selection behaviour (Chrome's mouse selection sucks!).
Knut Remi "DrLaunch" LøvliWednesday, February 13, 2013 4:02:10 PM
Before I start, I must say that that poll in the sidebar there is going to be very biased. It's not going to represent the userbase at all. It represents a very vocal minority, with strong feelings about this change. Most of which will probably be the the old "I'm afraid of change so I have to be a total douche about it", people.

I answered "Bad idea". Not because I'm afraid of change. Actually, the people who had that reason are wrong. I doubt Opera will be much different to general end user. Things will probably work the same way they do today. Probably even extensions and Dragonfly. I belong to the small subset that thinks this will have a bad impact on the future of the web. I think David Storey speaks for a lot of us on this matter.

What happened during the last browser monopoly, was that developers stopped following the web specs, and started making code that targeted MSIE-specific features. The problems caused by this were around for many years, and the Mozilla Foundation and Opera Software spent a lot of effort on changing things for the better. During the rise of WebKit on mobile, mobile devs started targeting WebKit-specific features, especially in CSS only. And during the rise of Chrome on PC/Mac, developers started making web applications that used features only available in Chrome, especially V8, and started using browser sniffing to block browsers again. What just happened is that Opera Software seemingly stopped wearing its blue beanie, and started rewarding these devs for trying to make a new browser monopoly.

Left are gecko and trident, used in Firefox and MSIE respectively. As far as Mozilla goes, I think they have yet to release an official statement about it. If they switch to WebKit, which I kind of doubt, we're left with trident, from the old monopolist MSIE. Then I doubt trident will survive much longer. Way to go Opera Software. Replacing that monopoly you fought so hard with another.

I for one am going to keep using Opera until the switch has been done. Then I'm going to have to decide whether to stay with an old version of Opera, or if I'll be forced to take my blue beanie with me and switch to Firefox. Am I moaning about this? Yes. Am I entitled to? Absolutely.

Follow @dstorey for more about web standards and the open web.
mubaidrWednesday, February 13, 2013 4:03:29 PM
Originally posted by Sin2x:

Haavard, Webkit is only a layout engine - what is going to happen to Opera's Javascript engine? Currently it is lagging behind its competitors speed-wise.

V8 is replacing Crakan, much speedy!

Originally posted by Sin2x:

Also, I really-really hope that Opera will not lose any of its configurability and its awesome mouse selection behaviour (Chrome's mouse selection sucks!).

Yes, that is concern ofr most of the Opera users.
Sin2xWednesday, February 13, 2013 4:19:31 PM
Originally posted by mubaidr:

V8 is replacing Crakan, much speedy!

Is this from a confirmed source or speculation? It doesn't seem likely to me.
FransWednesday, February 13, 2013 4:24:27 PM
"Opera will use WebKit as its rendering engine and V8 as its JavaScript engine." http://my.opera.com/ODIN/blog/300-million-users-and-move-to-webkit
M Shabeer AliWednesday, February 13, 2013 4:31:45 PM
Seems more like a "management" decision.

" ... it ends up taking up a lot of resources - resources that could have been spent on innovation and polish instead."

If "freeing up resources" isn't management speak for firing a bunch of developers, then there is certainly some merit in the idea.

It could also finally give you guys a part of the lucrative ios pie and a real chance to compete with Safari.

Of course, the right thing to do would have been to open source the Presto / Carakan source under a dual - license. But then there is the real threat that former Opera employees, who left because they were disillusioned with the current management, might just create a better product from it.

The direction towards adopting webkit however, also seems to be an attempt to dress up Opera and make it more attractive to Google and Apple - companies very interested in purchasing Opera.

Hope this isn't a short term solution and Opera won't go the Netscape and Nokia way.

On the whole I feel sad ... this isn't innovation - it's a weak compromise.
s/ashWednesday, February 13, 2013 4:38:28 PM
Originally posted by mubaidr:

Originally posted by Sin2x:
Haavard, Webkit is only a layout engine - what is going to happen to Opera's Javascript engine? Currently it is lagging behind its competitors speed-wise.
V8 is replacing Crakan, much speedy!

But at the same time it is slower than Mozilla's IonMonkey.
seriousWednesday, February 13, 2013 4:54:37 PM
Originally posted by AnuarSh:

If you change only what is below the address bar and above the status bar, then "as you want" ...
+1
mubaidrWednesday, February 13, 2013 5:00:26 PM
Originally posted by AnuarSh:

If you change only what is below the address bar and above the status bar, then "as you want" ...

+1, Enough said!
Tenno SeremelWednesday, February 13, 2013 5:16:36 PM
Originally posted by Frenzie:

But XML is even more annoying and less readable:

I'd say useless quotes and commas are worse. However, for plain 1 level configuration you don't need anything other than INI-style. When you've got trees, that's a different matter entirely. It's easier to parse XML with my eyes rather than JSON. Well, for me anyway
Kamil PiwowarskiWednesday, February 13, 2013 5:26:15 PM
What with Opera Dragonfly? ;(