The fast lady’s song

It made public news that Opera is sidelining the Presto engine for Webkit, something I expected before I knew.

What is good for business and what is good for the Web are not the same. Opera’s business case for Presto has not been so strong. Most of the development effort had the least perceptible effect for the users (and the profit), and what has had the most perceptible effect has taken the least effort.

Those quirks that have annoyed me as an Opera user, and those that have annoyed you if you use Opera too, are likely not what an Opera developer has spent much time to solve. I am no greater fan of Webkit than of Presto, and there is no reason to believe that Opera users will be more satisfied with Webkit than with Presto, but the onus is no longer on Opera Software.

If Presto had 5 bugs and Webkit had 10, but web developers worked around the Webkit ones, it would be Opera that appeared like the buggy one, and who had to expend developer time to handle those. Now bugs would not put Opera at a comparable disadvantage because the others in Opera’s world would share them too.

Presto Gecko!

In a matter of engine lineage Konqueror konquered through the Webkit branch, but who knows Konqueror today? A decade ago it was losing an OSS mindshare battle with Mozilla, but technically it was more of a threat to Opera on mobile devices than Mozilla was. The assessment back then was that Mozilla would pose no real threat on mobile, and a decade later that has proven to be right.

My view was that while Opera could take the device hills but they couldn’t win the war unless they also could hold the rich valleys below, meaning at the time and to this day Windows for PCs. That Opera has consistently failed to do through the years, while Mozilla took much that valley and some smaller Linux side valleys but not the higher ground.

Rivalry notwithstanding, in standards politics the closed-sourced for-profit Opera hill browser and open-sourced not-for-profit Mozilla valley browser may have been the closest of allies. It is not so surprising, the other competitors are the company men, whether for the goals of Apple, or Google, or Microsoft. It may not be a good omen for the Open Web that next year will be the year of the wooden horse.

The business interests of Opera haven’t changed fundamentally, an engine change affects the Web more than it changes the way a company operates. But Mozilla have gotten a lessened ally, not because Opera will be a -webkit- mouthpiece, but because a -webkit-Opera is likely to retreat from the standards battlefield, leaving a few guerilla warriors.

Fighting the Open Battle

I spent some of my opening years at Opera arguing for that it was Open Standards rather than Open Source that is crucial to keep the Internet free. Open Source is well and good, and we are better off that 2 of 3 remaining major engines are open sourced, but it isn’t the source code make for a healthy ecosystem. GNU Emacs is an Open Source editor-cum-operating-system, but had Emacs alone been be the editing environment, with different forks and customisations to adapt to different needs, next to nobody’s need would actually be fulfilled, and managing that haystack of forks would not be feasible.

Open Standards on the other hand, of which the Internet is the prime example, can handle any participant, open or closed source, and use their products. I can make something on the Web, like this blog entry or a data API, without knowing who will be using it or what software they use for it. You, and the billons of other people and machines that read this, can use this not caring who I am or what software I used. You can process this for your own purpose, again without knowing anything about who your producers and consumers are, as long as they follow open standards.

We may only realise how unusual what we have is when we lose it, when the contract of open standards is broken, and we no longer speak the same language. The web apps for different platforms is an example of losing that contract. If you develop an app for Apple, you would have to redo it for Android, and then again for Windows. It was really bad in the days of Netscape vs Internet Explorer, but it didn’t quite come to this.

Closed-sourced Opera has been ported to more devices, taken more hills, than I believe any open-sourced browser has. Basically Opera turned over the years into a device porting implementation. In a Closed Standards world that would not be enough. Were you to make the next great device first you would have to decide which feudal master you would have to declare your allegiance to: Apple, Google, or Microsoft. Independent would not be an available option.

Originally published on

Join the Conversation

  1. Yep, Presto has really good memory management, I could do 175 tabs in 25 windows easy with only 2GB of RAM, yet i still had most of the ram free :)I have been to many sites see that I was on Opera & get a go away message, other sites send broken code on purpose (some use old Yahoo that thing Opera 10+ is 1)I helped a bit with browserjs testing (partly finding the browser sniffers), browserjs helps a lotNow all the IE only browser sniffers block IE 11 (“IE” is not in the UA string)It is very hard to develop a web browser that gets ignored for vendor prefixing of webkit & moz, many sites vendor prefixed code, so Opera could not do those parts as it was block (not sure why you would add 3 to 4 times the code to do the same thing, one set of code would do)Opera Mail is still using Presto (I think) it has very few bugs (its crash reports are open in the web browser (presto) and the crash reporter talks about the browser not the mail program (wording is wrong)


Your email address will not be published.