IE8 needs Canvas support

IE8 needs Canvas support

Summary: IE may be the quickest browser to load pages, but this is not a 100m dash; seems like someone has forgotten to tell Microsoft that there is another 300m of JavaScript to go until this race is over.

SHARE:
6

IE8 is due to hit the internet at 3am tomorrow morning (Australian time), but Microsoft's newest browser will remain a laggard as the other popular browsers continue to increase their functionality.

Last week Microsoft released a whitepaper claiming that IE8 was quicker in loading 25 popular sites than its competitors. At the time, I noted that none of these sites were particularly heavy in their JavaScript usage, and of course, a load time does not reflect the rendering ability of a browser.

If you want a browser with awesome page load times and zero rendering, I suggest you use Lynx (or even cURL if you're truly hardcore).

Today comes news that Google has launched a site to showcase Chrome's JavaScript capabilities called Chrome Experiments. On this site are 17 demonstrations of JavaScript which is typically mixed with Flash and/or HTML5 elements, particularly Canvas. A glimpse at the following table shows how the current IE8 beta stacks up against its rivals:

Legend:
 Fail (usually epic)
 Some errors
 Pass
IE8Firefox 3.1bChrome 2.0.169.1Safari 4Opera 10 alpha
Boombox     
Pixamix     
Chromedrones     
BallDroppings     
Shop     
Video & Picture Puzzle     
Homeostasis     
Ball Pool     
Canopy     
smalltalk     
Social Collider     
Twitch     
Browsermation     
Browser Talk[1]     
DOMTRIS     
Browser Ball     
Google Gravity     
Monster     
Colorscube     

[1]: Full test not conducted due to lack of microphone; however, IE failed to render the test properly.

In the vast majority of cases, Internet Explorer failed the demo because it does not support Canvas. I'm sure that if I were to install Mozilla's Canvas plug-in for IE the results would be better — but if Internet Explorer is to be a leading browser, it should come with Canvas support and not expect users to install plug-ins for HTML5 elements.

As more and more web applications like Bespin appear that need Canvas support, one could be forgiven for seeing the faint beginnings of a return to the bad old days of "Best viewed in xxx browser".

IE is simply eating a truckload of dust, and that is not interesting.

Microsoft is too smart and savvy to let something as terrible as that return to web pages, but IE8 is in danger of being forgotten before it has even left the starting blocks. All the hype is currently surrounding the open source browsers and Opera for one very good reason: they are pushing the envelope and bringing the web new functionality. IE is simply eating a truckload of dust, and that is not interesting.

IE may be the quickest browser to load pages, but this is not a 100m dash; seems like someone has forgotten to tell Microsoft that there is another 300m of JavaScript to go until this race is over.

A postscript note on general performance: as these were tests built for Chrome and constantly nagged me to install that browser, it's no surprise to learn that Chrome was the quickest at these. But second place was always a toss-up between Safari and, surprisingly, Opera. When the demo worked properly, Opera often "felt" as quick as Chrome — of course, this is all conjecture without any formal stats.

Third place in performance was typically Firefox, with IE always bringing up the rear.

Topics: Google, Browser, Open Source

About

Chris started his journalistic adventure in 2006 as the Editor of Builder AU after originally joining CBS as a programmer. After a Canadian sojourn, he returned in 2011 as the Editor of TechRepublic Australia, and is now the Australian Editor of ZDNet.

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Talkback

6 comments
Log in or register to join the discussion
  • A test designed around software isn't really a test

    "A postscript note on general performance: as these were tests that were built for Chrome and constantly nagged me to install Chrome, it's no surprise to learn that Chrome was the quickest at these."

    That is the biggest problem here. Naturally if a test is designed around a particular piece of software then that software will have the upper hand at all times and that has been shown to be the case here. Let's do a history lesson though - Look back to who rolled out the first popular web browser. It was Netscape. Look who had a hand in the launch of the W3C. It was Sir Tim Berners-Lee, the bloke who was behind the WWW. It's easy to see that the original set of web standards (and thus updates and upgrades to those standards) were always going to favour browsers with the Gecko rendering engine and rips there-of. The fact that the W3C recognises vendor-specific functionality in Netscape but not IE is ample evidence of this.

    Some may argue, and perhaps fairly, that Microsoft can take some of the blame because they decided to go it alone instead of following W3C guidelines to the letter, however it should be remembered that when IE4 came out and walloped Netscape there was not the following or even concern for coding standards that exists today with half a dozen popular browsers.

    I'll add that no matter how much a distributor of a web browser (any of them) says their product meets the requirements of coding standards, not one of them renders a web page (standards-compliant or not) in precisely the same way. In some cases this difference is a big one and in other cases I am talking about one or two pixels in any direction but it is a difference none-the-less and therefore no less of a headache to those of us left to produce websites that look the same for everyone regardless of what they are using.
    anonymous
  • IE8

    It is too expensive to develop for Microsoft products. Their determination to chop and change while desperately trying to avoid standards means that developers and their customers are trapped into an endless and pointless upgrade/update cycle. I'll stick with standards compliant products that have the best chance of supporting web code for the long term - Firefox and Opera.
    anonymous
  • IE Walloped Mozilla

    Mel, what are you thinking? IE walloped Netscape because the company behind IE placed IE as part of the operating system & pressured companies to not use Netscape. This has been proven in a court of law. That IE walloped Netscape had nothing to do with the capabilities of the browser itself. IE added in stuff that worked with the operating system & not with other operating systems. IE had as many holes in it then as what Netscape did. In those days Netscape Communicator was around & IMO it was a much better browser than IE was.

    Despite IE's prevalence I have been an ardent detractor from it & avoided it at all times. It gets used only when sites are M$ specific but these happily are disappearing.
    anonymous
  • M$ claim that IE8 is WWW3 compliant

    I had a version running on a netbook & found it interesting that it was quite good on compliance. So much so that to connect to an IE-specific site then that site would not display correctly. I did laugh mockingly.
    anonymous
  • Can you please explain !!!

    You said -
    "... At the time, I noted that none of these sites [top 25] were particularly heavy in their JavaScript usage ..."

    Does it not imply that javascript is currently not the most important workload to optimize on. With that data in mind, where would you deploy your engineering workforce ? Doesn't the investments that the IE team make a lot of sense !!! That is not to say that the IE team did not improve JS perf. They did, but they focussed on a much larger scope of things to bring in real perf gains.

    you said -
    "...IE may be the quickest browser to load pages, but this is not a 100m dash; seems like someone has forgotten to tell Microsoft that there is another 300m of JavaScript to go until this race is over.."

    What is that supposed to mean ? IE is the fastest to load pages and "get you going". Despite this fact, you would argue that an average user should worry how fast the javascript engine or the renderer or the many other components individually be !!!

    I'm interested to know how you describe perf from a user point of view - raw javascript speed, renderer performance or overall speed of browsing.

    Regarding Canvas. HTML5 spec is not yet ready. It's not expected to be signed off till 2021 according to w3c. There are not many sites using that tag yet. There are pluggins that do the job now. If you managed the engineering resources of the IE team, would you prioritize this investment for this release ?

    I'm surprised at the state of journalism these days. Somebody publishes a report that chrome has a fast javascript engine, and the rest of the "journalists" start rehashing the same stuff over and over again. The original articles are searchable, you do not need to rewrite it.

    The white paper that the IE team published should be a must-read for everybody writing about a performance characteristics of any product. Any serious engineer/journalist writing about performance would understand how difficult it is to describe performance. to measure and interpret the perf data. I would request you to do your homework before trying to influence the public opinion in any way.
    anonymous
  • IE8

    Bruce, your argument is flawed in many ways. Like it or not (and I assume a lot of people don't) IE is by far the most used web browser out there. It can't be avoided, especially by big business and big websites. I own a few small websites and I don't dare even avoid IE for them because I read the logs and I know what most people use.

    If you build a website solely to render in Firefox and Opera, the latter having such a small share of the browser market despite its capabilites that its not even worth you bringing to anyone's attention, then most people who view the site(s) you are in charge of will not be able to use the site effectively.

    It is neither expensive or time-consuming to develop websites that render well in all browsers and it has to be remembered that a standard is only a standard when it is universally adopted and recognised and because of the great number of standards that exist and the fact that many older ones such as HTML4.0 are still in popular use. This is an issue that has plagued the W3C since its beginnings. The W3C know this to be true and that is why they offer both transitional and strict arguments in their DOCTYPE tags. The bottom line is that the W3C doesn't have the political or financial clout to make all their older standards redundant and force all browser vendors to make their software 100% compliant with the current standards.

    I once had an acquaintance of mine say that strict standards-compliance was more important than how a site renders. I was almost beside myself with disbelief at the time since this bloke was a professional web developer. I beg to differ with his opinion based on my own priorities - when I visit a website I look for things like ease of navigation and how well the site is laid out. Content delivery is very important but not at the expense of ergonomics just for the sake of standards-compliance.
    anonymous