Friday, 28. March 2008, 15:40:33
SVG, Web Standards, HSLA, Safari
...
We've just released a public build of our testing and development build that shows Opera passing 100/100 in the DOM tests. It also has pixel perfect rendering. The final test which has to be passed is the performance test. This puts Safari and Opera neck and neck in the race to be the first browser to pass Acid 3. I'd like to congratulate the Core Opera team for all the work they've done to pass the first two tests so quickly, and the WebKit team for making this a fun race so far. Both engines getting this far means that we both support important CSS3 properties like HSLA color model and Web Fonts, and pushes support for SVG forward. I'm looking forward to when Gecko and the new IE engine get this far.
Tuesday, 20. November 2007, 15:45:21
Firefox, CSS2.1, Safari, Opera
...
With the release of Firefox 3 beta today, and the recent release of Safari 3 final joining the beta release of Opera 9.5 kestrel, all the major players bar one have entered the battle field. The Internet Explorer team are still keeping their eight generation browser close to their chest. While Safari 3 has shown its full hand by shipping as a stable release, both Kestrel and Firefox still have time to add features, standards support and polish.
Ignoring IE8, as that would just be guess work, what does this generation of browsers have in store for us at this moment in time? For web designers and developers, standards support is one of the most important details. I'm in the process of comparing these browsers to see how their CSS support stacks up against each other, using the 2007 CSS snapshot as a base. I'm using the current internal build of Opera Kestrel (more or less the same as the last weekly), and the aforementioned Firefox 3 beta and Safari 3 final.
As previously reported, the CSS snapshot includes CSS2.1, CSS3 Selectors, CSS3 Namespaces and CSS3 Colour. While I haven't completed any tests in detail yet, I've done some tests and have a feeling that Opera Kestrel will come out on top. Excluding bugs, Kestrel supports all CSS 2.1 properties and values except one - visibility: collapse. I've not checked the latest changes to this spec however. I'm also not sure if the bugs in the CSS2.1 test suite have been fixed yet. The test suite is huge, so will take a good while to go through, especially for 3 browsers. For CSS3 Colour, I've made a preliminary support chart on CSS3.info. Firefox leads in this regard, with Opera behind due to not supporting the alpha channel on RGB and HSL. Once alpha channel support is added both will get fixed however. The browser support chart for Selectors is in need of an update, but running the selectors test on the same site shows that Opera passes all tests, while Safari and Firefox still have some issues. This isn't strictly accurate as Opera doesn't have ::selection support, and does have some known bugs. There is a more in-depth test suite that I'll go through. It is fairly certain Opera wins out here though. For CSS3 Namespaces, it is pretty much even, with all browsers supporting @namespace, although of the five tests I did find, Safari has one issue, while Firefox and Opera both pass all of the test. Overall I think that each of these browsers has fairly good support for the CSS 2007 snapshot, but Opera should be in pole position for the moment. I could be wrong, and things could change before Firefox and Opera release stable versions, and Safari could even release a stable point release before each vendor releases their final product. Who knows, IE could come and surprise us all as the dark horse in the pack.
Sunday, 9. September 2007, 19:02:02
iPod, iPhone, standards, Safari
...
The issues of designing foe one browser, never-mind one device, should be very clear to anyone that has promoted web standards and the open web. Apple themselves have made this even more clear by their latest move. They've just recently anouned the iPod Touch, with the included Safari browser. Now this doesn't suffer from many of the problems of designing for one browser or device; It has the same screen, the same engine and likely the same hardware. However, if we look at all those URLs that have been coming out in the last weeks. like digg.com/iphone, mediatemple.net/iphone et al. we have a problem. It is easy to imagine that iPod Safari users will think these sites are for iPhone only, when they'd in fact work just great on the iPod. Work great that is as long as the authors don't use browser sniffing to detect the iPhone (providing the iPod uses a different user agent string, I'm not sure). These sites will often also work great on other mobile browsers like the S60 browser, Opera Mini and Opera Mobile, but the URLs suggest otherwise.
Of course, the naming and the iPod issue isn't the only reason why it is short sighted. I've hi-lighted why one browser design is bad, in that it introduces cases where you rely (even accidentally) on browser bugs and vendor specific extensions. It may not matter to you, after all the Iphone/ipod is really hot, but it does matter to your potential customers. If we look who there are, then you'll probably see something like this:
Mobile/Handheld browser share for August 2007, as percentage of total browser market share
Data from Net Applications
- Opera Mini: 0.27%
- iPhone Safari: 0.05%
- PSP Internet Browser: 0.02%
- Series60 Browser: 0.02%
- Danger Web Browser: 0.02%
- ACCESS NetFront: 0.01%
- BlackBerry: 0.00%
Mobile/Handheld growth, July '07 to August '07, as percentage of total browser market share
- Opera Mini: + 0.03 (0.24 -> 0.27)
- iPhone Safari: + 0.01 (0.04 -> 0.05)
- PSP Internet Browser: 0
- Danger Web Browser: 0
- ACCESS NetFront: 0
- Series60 Browser: - 0.01%
- BlackBerry: Doesn't feature in July features
The iPhone and Series60 figures were taken from the Operating System figures. For iPhone this will be a accurate measure as Safari is the only browser available. For Series60 it will also include other browsers like Opera Mobile. As always with statistics, take them with a grain of salt. I don't know which sites were monitored to get these figures. As they are a US based company, I'd expect (but don't know) the stats to be biased towards the US or English speaking countries, where Apple tends to be stronger. Nokia tends to be stronger outside the US and Opera Mini historically is also more popular outside the states as Mini doesn't work on Verizon phones and has had issues with T-mobile US.
At least by these figures, Mini is not only far more popular than all the other mobile browsers (put together) but is also growing faster. This is something to bear in mind when thinking of making a device/browser specific web site, or optimisation. I'm certainly not saying to make Mini only sites though, digg.com/mini would just be as bad. Imagine how much work it would be to update and create the regular site, and a Mini, iPhone/iPod and Wii version for example. Admittedly, there is not much to go on with just two months data, and things will likely change as time goes on. The global role out of the iPhone will help its marketshare, when that eventually happens. Mini will always be available on more devices, and more carriers, and in markets where Apple doesn't focus. There are markets where devices like the iPhone will be too expensive, such as the developing world, where feature phones rule. There are a lot of people in these markets. I think iPhone will eventually get a significant percentage of the mobile market and be a major mobile browser, but I don't think it will take over the industry like it has for digital music players. Another advantage Mini has is that for the most part, the users of Mini are people that downloaded a mobile browser because they want to surf the web. For many other mobile browsers, they are pre-installed applications that people don't always use, or know what the application does. Safari wont often have this problem either as Apple will market internet on the iPhone heavily and people buying the iPhone will probably be see it as a reason for purchasing the phone. Who knows who will eventually become the major player beyond the desktop, but I don't think it will be a one horse race like happened on desktop.
Tuesday, 12. June 2007, 20:13:46
Safari, standards, speed, Opera
...
Congratulations to the Safari team on the beta release of their browser on Windows. Welcome to the party. It's a great day for Opera. Of of the reasons why it is so good for Opera is web standards. Many web developers don't test in Safari as it was only available on Mac. With another standards aware browser available on Windows it reinforces to developers that standards matter. Especially to the crowd that just test on IE and Firefox, and assume that Firefox equals web standards. This leads to many issues where developers use Mozilla extensions to the DOM or Mozilla bugs without realising it. In many cases, sites that break in Safari break in Opera and the other way around. I know there has been work I've done that has benefited Safari, and I'm sure that developers that find issues in Safari will also help Opera (providing they don't use browser sniffing to just give Safari the fix). I'm active now and again in the WebKit bug tracking system, and I hope we can work closer together both ways in the future. This kind of work benefits every body. Safari have been fairly quiet n the browser community of late and it'd be nice to change that. It is not only important to work closely with the Safari team, but also the IE team and the Mozilla team. To be honest, the IE team have probably been the most helpful recently, and we are building great relations there.
Safari have been laying down the smack, with the claims of the fastest browser. The results they show, particularly hi-lighting the HTML rendering speed (Which I personally think JavaScript speed is more important these days) don't look too flattering to Opera, but as always you can take these sorts of results with a grain of salt, especially as Safari have been optimising for iBench. Use another test framework and you'll probably get different results. None of this takes away from Safari being a great product. WebKit is a very nice, fast, standards compliant rendering engine. But, while we often keep things close to our chest, we are not standing still on the development of Opera. While the speed comparisons were done with Opera 9, we are well into the development of Opera Kestrel. Opera has always been fast, and it is a design goal of both Kestrel and Peregrine to improve the speed further. I believe we are making good progress in this. We have a target to aim for now. As quoted in the Opera Desktop team blog, The Kestrel falcon is able to see ultraviolet, which helps them spot prey while hovering 10-20 meters over the ground.
, while The Peregrine Falcon is the fastest creature on the planet in its hunting dive, the stoop
. Just like Apple, Opera has innovation in its corporate DNA. Many of the new features found in the new Safari were first found in Opera, such as sessions, and we'll continue to innovate at a fast pace. I think it is an exciting time, where there is some strong competition in the desktop space that drives the industry forward.
I look forward to working with the Safari team and the other vendors to solve compatibility issues, and hopefully we can sit around a table soon to discuss this. Sharing test cases is one area where we can work together, as well as setting a baseline in what standards we implement to drive the adoption of important standards such as the mature parts of CSS3. Congratulations again to the WebKit team on a fine product, and to Apple for what looks like a really exciting Leopard release.
WIDGETIZE