Web browsers (part 1)

By Grant Jacobs 15/06/2010

For a diversion, I’m offering a brief round-up of web browsers with a one-paragraph introduction. I am not offering a review of all the differing features: that would take many days, time I don’t have. (Nor do I have enough interest.)

My main aim is to encourage people to try the alternatives. Download them, give them a run.

In a second post, I will introduce the real alternatives; here I’m going to cover the better-known players.

There is all sorts of noise about what is the ’best’ web browser out on the internet. Personally, what is best – beyond issues of standards compliance – is what works best for you.

For those that must, some benchmarks for JavaScript performance* can be found in a recent ArsTechnica article. Benchmarks are dependent on a lot of different, rather fiddly things that make interpreting them more complex than they look at face value.

I generally don’t get overly worried about benchmarks unless there are marked differences across a wide range of tests, and I would encourage readers to do the same. You are better to worry about features that help you and stability – the browser not crashing! – in my opinion.

In practice most browsers do well at some performance tests and less so at others. My casual eye-balling of the benchmarks in the ArsTechnica article suggest Opera may be doing best overall in that it while it only sometimes is the fastest, it rarely does poorly.

First, four browsers that run on at least Windows and Mac OS X (some also have Linux versions):

safariSafari. ComputerWorld and ArsTechnica have a good reviews of the main new features added to version 5, just released a few days ago. Aside from the push towards HTML 5, developers can now create extensions to Safari which will be sandboxed. (That is, the code of each extension is isolated from the remainder of the browser and other extensions, so that havoc in one extension will not affect others or the web browser.) I’m looking forward to seeing what developers create.

The new Reader Mode has been creating a little fuss, with some saying that it will block advertising and is therefore ’against’ the model of the internet and that site developers will lose control. This feature scans a web page and if locates a element with more than 2,000 words of text it offers to present it as an article ’stripped out’ from the page it came from.

Like all recent browsers, the new version offers faster JavaScript. Naturally there are arguments all over the place about which browser is faster… For myself, it is great to see increased JavaScript performance, as this should unlock creative potential of developers.*

opera10Opera. I have a little bit of a bias here. Opera feels like the perennial underdog, despite at one point being the most standards-complient, fastest and most stable browser on the market. In particular, it regularly introduced new features before it’s competitors.*** It’s competitors have since made in-roads and now the differences are now lesser. It is a very professional development effort run by a well-established commercial company. The ’desktop’ browser has always been free to the best of my knowledge; I believe they make their income from the ’device’ versions of their browser. Unlike most other browsers, it also has built-in mail programs and whatnot. (The download is still a modest size.) It’s stability has been particularly important to me, as I leave the browser up for weeks at a time with many open tabs. (50 wouldn’t be unusual.)

firefox35Firefox. Developed on an open-source model, this browser is the successor to the NetScape/Mozilla family of web browsers that made up the main portion of the market outside of Internet Explorer. It is particularly well-known for it’s large collection of extensions, which add considerable functionality over-and-above the ’basic’ browser (which is not to say that it is basic). Writers and scientists, for example, might be interested in Zotero, the bibliographic reference extension.

chromeChrome is a relatively recent addition to the browser market, a much trumpeted offering from Google. Initially launched as the fastest web browser on the market, the others are already catching up. Some users express mixed feelings over it’s user interface; I’m not entirely convinced in this aspect of it myself but it’s a personal preference thing and others will have different ideas aout it. Try it out for yourself. It is both true and a bit unfair to say that it lacks the wide range of features found in, say, Opera – I come away with a ’fast but stripped down’ feel – but this will in part reflect that it is quite new with more features yet to come.

Finally, one browser that only runs on one platform:

internet-explorerInternet Explorer only runs on Windows. Several years ago, prior to Apple releasing Safari, there were Apple versions. I have little experience of recent iterations of this browser. I have nothing against it, but I am not able to to comment on it.**** I have little doubt that it does the job. It seems to fare poorly in speed benchmarks other than where it makes use of hardware graphics acceleration. Developers (that is, programmers, have loathed earlier versions for not following the international standards with the upshot that more complex web pages often required specific code that detected the user was using ‘IE’ and using alternative code. Most of this should now be well in the past; it will be interesting to see if it can regain favour among developers.


* Why JavaScript performance? Most modern web browsers are very fast at presenting ’static’ web pages. Two of the main places users see pauses are when the browser has to go off to fetch additional content in order to present a page,** or where some dynamic contents needs to be manipulated by JavaScript. JavaScript is a interpreted programming language that runs inside your web browser. As far as programming languages go, it’s slow. Speeding it up will unlock the creative potential of web developers (programmers) and allow (some) sluggish dynamic websites to become snappy.

** The very latest browsers use clever schemes to resolve the locations of the extra content (DNS pre-fetching) and behind-the-scenes downloads to try reduce these pauses. It’s part of the reason for their apparent increase in speed: they’re not actually faster as such, but are less often held up on waiting because they try fetch information ahead of it being needed.

*** I’d list some of them, but it seems a bit moot now that the other competitors offer similar features.

**** My virtualisation software (that allows users to run Windows on non-Windows machines) needs updating, a job that is low priority at the moment, otherwise I would.

Other articles on Code for life:

That Ben Goldacre fuss

Retrospective–The mythology of bioinformatics

iPads for the disabled

Wikipedia project available for a good home

0 Responses to “Web browsers (part 1)”