overview

This page considers web browsers.

It covers -

• introduction - the technology and issues
• history - from the 'browser wars' to Firefox and Chrome
• markets - how many browsers, market share and demographics

It supplements the discussion of Networks & the Global Information Infrastructure, Accessibility, Design, Electronic Publishing, Metadata and Search.

     introduction

Web browsers - such as Internet Explorer, Safari and Mozilla Firefox - are software that enables a user to display and interact with 'web documents', including -

• items accessed on the internet (hosted on a web server)
• items accessed on an intranet
• physical format digital publications (eg on a CD-ROM, on a floppy disk or USB drive)
• items held on a personal computer.

They are thus not restricted to display of information that is on the web.

In essence a web browser does four things. It -

• 'fetches' web pages on the net using the hypertext transfer protocol (HTTP) or other protocols such as FTP and Gopher
• recognises particular code and files (eg HTML and XML code, GIF and JPEG image files)
• interprets that information for display onscreen (and through a printer) in a layout that is broadly what the author of that code intended
• offers additional functionality, eg bookmarks for keeping track of frequently accessed locations, tabbed browsing and even predictions (based on past browsing) about user destinations

A browser may be bundled with a search engine but in function they are separate entities. It may be designed for use on laptop/desktop personal computer or for browsing on a PDA or mobile phone.

Browsers interpret the code that underlies web pages and other resources. That interpretation - what you see on the screen - varies from browser to browser (and from type of personal computer or other device). It does not offer the verisimilitude of print, which is one reason for the use of PDF. Different browsers may not display a document fully or at all (eg some browsers are text-only).

Some browsers have greater functionality than others - early browsers were for example more restricted than recent generations. As part of commercial efforts to 'own the internet' (or merely the consumer desktop) some were 'optimised' for variants of code espoused by particular businesses, notably Microsoft's IE for code generated using its Frontpage editing tool.

It is noteworthy that there is no single 'official' browser: anyone is free to develop a new browser and let it contend in the market. Tim Berners-Lee's initial browser disappeared; the W3C browser Amaya has remained an academic exercise.

As the following paragraphs indicate, the past five years have been dominated by a browser monoculture, with Microsoft (through its hold on the desktop market in many nations) having the advantage that IE is the default browser for most people and indeed has often been mistaken for "the internet".

Apart from what has been denounced by judges, government agencies and competitors as sharp practice or anticompetitive activity that monoculture is of concern because

• it fosters security vulnerabilities
• at best blurs the global standards on which the web depends.

     history

The history of browsers has been driven by the interaction of technology and markets, with marketing decisions in practice having been more influential than the excellence of the code in particular browsers or their compliance with W3C standards.

The first web browser - confusing badged as WorldWideWeb - was devised by Tim Berners-Lee in 1990 and introduced to colleagues at CERN in March of the following year.

Uptake of the web in research institutions, by network operators and by enthusiasts was underpinned by NCSA Mosaic, a Unix-based graphical browser developed by the US National Center for Supercomputing Applications (NCSA) and released in 1993. Mosaic was subsequently released in versions for the PC (Microsoft Windows) and Apple Macintosh.

Mosaic team leader Marc Andreesen left to form what would later become Netscape Communications Corporation, with support from James Clark, an executive of minicomputer manufacturer Silicon Graphics (SGI).

Netscape released its Navigator browser in October 1994, becoming the dominant player by the middle of 1995. Its expectation was that it would build a sustainable business by selling server software, with the browser being freely available. That vision intially looked credible, given Microsoft dismissive view of the net (the first edition of Bill Gates' The Road Ahead famously ignored the web, reflecting perceptions that consumers would rely on proprietary networks such as AOL and MSN).

Media interest, government cheerleading and growing online populations in the West saw Microsoft respond by acquiring a web-authoring tool (FrontPage) for around US$100 million and release the Internet Explorer (IE) browser, with code ultimately drawn from the NCSA. That release was featured in Windows 95 Plus, with a default page set to MSN - a move that reinforced the 'legitimacy' of the web, gave Microsoft a major competitive advantage and led some users to equate IE/MSN with the net.

The market share of Netscape - and of the range of other browsers (such as Opera) that emerged in response to media coverage and Microsoft's new interest - accordingly declined.

Netscape's competition with Microsoft in what was tagged as "the Browser Wars" was lauded by many pundits, browser users and hardware/software manufacturers such as Sun and Oracle. However, Netscape's US and global market share continued to decline, down from an estimated global 90% in 1995 to 72% in 1997 to 33% in 1999.

Netscape came under increasing pressure as financial analysts forecast ongoing decline and noted that competition from Microsoft over server software (which MS variously gave away or released at 'special prices') eroded Netscape's capacity for major innovation and marketing.

Neither of the two major players gained major plaudits from some specialist observers, who criticised a perceived balkanisation of the web through emphasis on proprietary extensions.

Recurrent concerns about true interoperability were reflected in various "viewable with any browser" campaigns, responding to signals from web designers that a site was optimised for a particular browser (eg the 'best viewed in Netscape' label encountered by many surfers). That has most recently been evident in establishment of the Web Hypertext Applications Technology Working Group (WHATWG) under the auspices of the Mozilla Foundation.

In 1998 AOL (subsequently to engulf the Time Warner media conglomerate) acquired the ailing Netscape for US$4.2 billion. Acquisition was followed by the controversial release of code that came to form the basis of the Mozilla browser. It was also highlighted in US federal/state government and European Commission antitrust litigation against Microsoft, featuring arguments that integration of IE with Windows was inappropriate, that MS required hardware manufacturers to favour IE and that MS responses about 'unbundling' were at best disingenuous.

Microsoft's share of the desktop in Australia, the US and other markets continued to grow. By 2002 some estimates suggest that Netscape was down to around 4%, with IE having over 87%. (Uncertainties about market share are discussed below.) Competition came from niche browsers such as iCab, Opera and Lynx (a text-only browser attractive to some people facing accessibility barriers).

That year saw release of Mozilla, later enhanced for non-specialist users as Firefox (acclaimed as quicker and safer that IE). It was followed by Apple's Safari, based on the open source Konqueror browser (a competitor of Mozilla on Unix-based systems.

In December 2007 AOL announced that it would stop supporting Netscape as of 1 February 2008. The announcement reflected the disappearance of Netscape's market share, down to a mere 0.6% by late 2007.

In 2008 Google launch its own browser - Google Chrome. It commented that

We can add value for users and, at the same time, help drive innovation on the web. We realised that the web had evolved from mainly simple text pages to rich, interactive applications and that we needed to completely rethink the browser. What we really needed was not just a browser, but also a modern platform for web pages and applications, and that’s what we set out to build.

Chrome was initially released for Windows users, with an expected release of versions for Apple and Linux. It is written with Webkit, the open-source engine used by Apple's Safari and Google's Android. The browser is also getting a new Javascript virtual machine, V8, which is said to be better for complex and rich Web applications. Prior to release of Chrome Google provided financial support for other browsers, notably Firefox from Mozilla, an open-source product.

     markets

Although Microsoft's IE is the dominant browser in the West, it is important to recognise that other browsers exist, are in use and may displace IE in future.

The success of IE is arguably a matter of effective marketing rather than technological pre-eminence, so rhetoric from Redmond about a digital 'manifest destiny' should be regarded with scepticism. 2004 is reported as having seen the MS share of the market decline for the first time - albeit marginally - with uptake among youth and enthusiast demographics of Firefox (which claimed over 25 million downloads in 100 days from its initial release) and Safari.

In principle, many people with basic software skills are capable of developing a rudimentary browser. Since 1993 we have thus seen the release (and often disappearance) of a succession of browsers that include

• Amaya
• AOL
• Camino
• Chimera
• Compuserve
• Dillo
• Epiphany
• Firefox
• Galeon
• iCab
• K-Meleon
• Links
• Lynx
• Netscape
• Omniweb
• Opera
• Prodigy
• Safari

Figures for day to day use of those browsers (and IE) are uncertain. Statistics about browsers, like other internet metrics, are often a matter of eye of newt, leg of toad - inference and hocus pocus - rather than truly concrete data.

One reason is that there are no authoritative global surveys of use. Analysts instead often rely on figures of browsers shipped with personal computers (eg all Windows personal computers are counted as being IE machines) or downloaded from vendor sites. That is an inexact measure, since many people appear to have several browsers on their machine/s but rely on a single browser for daily use.

Other analysts have inferred the shape of the overall user population by examining server logs - evident for example in the alerts, when trying to access particular sites, that the user is not using the current edition of IE (and in egregious instances thus cannot access the site/page).

It uncertain whether those inferences accurately reflect the practice of the overall online population in the 'anglosphere'. Much writing about browser use arguably ignores or understates practice in Asia - now a substantial part of the global online population - where many people are apparently relying on browsers developed for particular domestic markets and non-ASCII character sets (eg in China).

The brief on Measuring Web Site Usage: Log File Analysis by Susan Haigh & Janette Megarity notes particular issues, complemented by Kim Bartlett's 1999 discussion How Server Statistics Undercount Text Browsers and Dan Tobias' comparison of results from three counters.

It is important to recognise that demographics for browser use vary.

The preferred browser for some technical populations or national groups may not be IE. Detailed figures are uncertain, with much of the information being anecdotal and thus properly treated with caution. It appears that graphics specialists - in particular those using Apple personal computers - tend to favour Safari or other non-IE browsers for their personal use. What a MS representative characterised to us as "hard core geeks" often pride themselves on avoidance of MS products.

Surveys on some ICT sites thus indicate that over 40% of the preferred browser is a non-IE product. That is reflected in site by site figures; a Wikipedia article for example claims that Opera (which some reports suggest has a fraction of 1% of the overall market) accounts for around 20% of contributions to the Norwegian and Russian Wikipedia. The 'klingon demographic' is similarly less likely to use IE.

A study of some 3,500 unique visitors to this site over three days in 2005 indicated that browsers were as follows -

• 80.5% MS Explorer
• 12.5% Firefox
• 4.18% Safari
• 1.09% Opera
• 0.78% Netscape
• 0.75% Mozilla
• 0.09% Camino
• 0.03% Konqueror

For the corresponding period in 2007 some 6,230 visitors used the following browsers -

• 72.3% Internet Explorer
• 19.8% Firefox
• 4.08% Safari
• 2.52% Mozilla
• 0.90% Opera
• 0.18% Netscape
• 0.06% Camino
• 0.04% Mozilla Compatible Agent
• 0.02% Konqueror
  

    next page  (studies and landmarks)