News for hackers. Stuff that matters.
Re:The latest build of Webkit "has to be seen to be believed." by pg on Saturday February 09, 2008 @10:26PM (#112876)
It's pretty impressive that Apple can compete with open source. I forgot that was possible.
Re:The latest build of Webkit "has to be seen to be believed." by tolmasky on Saturday February 09, 2008 @10:33PM (#112879)
Apple has open source figured out. It's the prefect fusion: the core technologies (http://webkit.org) are open source, but they manage the UI like no one else can. Software with style, the way it should be. The WebKit team is absolutely fantastic too, I miss them.
Re:The latest build of Webkit "has to be seen to be believed." by halo on Sunday February 10, 2008 @01:51PM (#113041)
You make it seem like it's open source thanks to their own good will - it's based on KHTML, the Konqueror rendering engine, which was LGPLed.
Re:The latest build of Webkit "has to be seen to be believed." by tlrobinson on Sunday February 10, 2008 @05:20PM (#113096)
Yes, it started as KHTML, but Apple has taken it to a new level.

The KHTML and WebKit guys have had their differences in the past, but I believe KHTML is trying to merge all of the WebKit improvements back into KHTML, hopefully eventually unforking the two:

http://arstechnica.com/journals/linux.ars/2007/07/23/the-unf...

Re:The latest build of Webkit "has to be seen to be believed." by boucher on Sunday February 10, 2008 @06:17PM (#113114)
It's also worth noting that Apple could have used a different engine, or started from scratch.

Similarly, they could have chosen not to open source the darwin kernel, and they could have based it on something other than BSD.

The point was that Apple seems to know when to choose open source technology and leverage it, and when to start from scratch and build their own proprietary system.

Re:The latest build of Webkit "has to be seen to be believed." by boucher on Saturday February 09, 2008 @10:35PM (#112880)
Apple competes with open source all the time. Linux is the most obvious example, but most consumer open source software competes with either Apple or MS, and open source is almost always losing. I think what makes this unique is that Firefox is an unusually well designed product that appeals to the average consumer in ways most open source software does not. It also had a head start, and has a huge user base.

In fact, you could argue that unlike usual, Apple is actually winning on technology, and not usability, since perhaps the biggest advantage to firefox is its ability to install plugins.

Re:The latest build of Webkit "has to be seen to be believed." by anewaccountname on Sunday February 10, 2008 @03:18PM (#113061)
I don't know if I'm reading this right, but think they aren't competing with "open source" in a general sense, since Webkit is open source. Or maybe you meant "with" in the "using" sense: "It's pretty impressive that Apple can compete using open source. I forgot that was possible."

Or maybe you were just talking about Firefox...

Re:The latest build of Webkit "has to be seen to be believed." by Nicolay77 on Sunday February 10, 2008 @12:17PM (#113026)
Opera has been doing that for ages.

AFAIK Opera 9.5 Beta is still faster than these two browsers.

Re:The latest build of Webkit "has to be seen to be believed." by tlrobinson on Sunday February 10, 2008 @06:01PM (#113108)
Where did you come up with that assumption? A quick (very unscientific) test of a big web application (over 100 JavaScript and image files) shows you're wrong...

(first column is seconds with empty cache, second column is seconds after cache was filled)

    WebKit     - 15 / 9
    Firefox 3  - 18 / 12
    Safari 3   - 19 / 12
    Opera 9.24 - 20 / 15
    Opera 9.5b - 20 / 19
    Firefox 2  - 21 / 21
Again, these are very rough numbers (I took the best out of about 3 runs) on a pretty slow machine (1.33GHz G4). Perhaps the numbers would be different for simple static pages, but I don't have a good way to time that.
Re:The latest build of Webkit "has to be seen to be believed." by Nicolay77 on Monday February 11, 2008 @09:54AM (#113355)
I got my data from here: http://celtickane.com/webdesign/jsspeed2007.php

Of course, the scientific method means this tests should be reproducible by everyone, and YMMV, but I wasn't talking bullshit, just see the page.