5. Conclusion

The development of XML has opened up many possibilities for working with archaeological information, both on and off the Web. While print designers have used raster and vector technology equally and in combination for years, Web designers have been without a comprehensive vector graphics solution. Forced to work around the particular issues associated with vector images for so long, SVG has remained somewhat below the radar of many, but this will almost certainly change. In addition to possessing the useful qualities found generally in vector technology, SVG opens up a host of additional possibilities. When combined with other forms of technology like databases and JavaScript, it is capable of interactivity that is both simple and complex. SVG can be used to create high-quality graphical interfaces for input and analysis of archaeological information, and make primary data easier to understand. As archaeologists seek to preserve their data either for further active use, or in the form of an archive for future interpretation, forward migration of data is an important issue (Anon. 2005c; Zeldman 2003, 17). Using standards-based W3C technology and archaeological data in ways that can be easily brought to the Web will probably produce data that are better maintained, easily migrated and serve multiple purposes.

SVG is currently in its second release. SVG 1.1 became a W3C recommendation on 14 January 2003 (Anon. 2006j) and, although fundamentally similar to the original version, it addresses problems found in the first release along with some enhancements. Primarily, SVG 1.1 is a modularisation of the recommendation, which is being applied to other W3C technologies as well (including XHTML). By defining the functionality of each module in a way that can be freely combined, the W3C hopes to minimise the creation of proprietary tags that undermine its overall usability (Cagle 2002, 548). The mobile device industry is one of the primary users of SVG, and release 1.1 now consists of the full version of SVG and the new 'SVG Mobile', which comprises 'basic' and 'tiny' modules to accommodate different classes of mobile devices (Watt et al. 2003, 1032). Work is currently underway on SVG 1.2 (Anon. 2006k). While there are several enhancements under consideration, the most eagerly awaited will probably be a provision for text flow. As an XML grammar for images, SVG does not currently handle large amounts of text well. This is one of the primary obstacles hindering the use of SVG as a Web authoring tool. There are some awkward ways around the problem but, generally, text rendered in SVG does not reflow when changing the size or shape of a browser window (Cagle 2002, 552). One solution may be the 'foreign object' element, which allows (X)HTML to be housed inside of SVG documents instead of the other way round. Unfortunately, this has not been properly implemented in the most popular browsers or plug-ins (Watt 2002, 393). A better solution would be to address this problem directly in SVG. There are several improvements in progress, and one of the most promising is the 'flow text' element. Flow text would not only allow text to re-flow naturally as users have come to expect with (X)HTML, it would make whole text blocks in an SVG image selectable and searchable. Currently only single lines can be selected (Watt et al. 2003, 1038). This would greatly increase usability and accessibility, allowing the structure of a website to be created in SVG and controlled using CSS while retaining a similar level of functionality to (X)HTML.

Archaeology occurs in three dimensions, and many vector-drawing tools used by archaeologists reflect this. Much of the value in CAD and GIS programs lies in their ability to render information in 3-D. While SVG images can be created from 3-D sources, the result is always a flattened two-dimensional vector image. For example, a 3-D model created in AutoCAD can be placed and rotated into a particular 'view' that highlights whatever visual information the archaeologist is trying to convey, and that view can be exported into Illustrator with layers intact. Illustrator can then create a flattened 2-D image in SVG showing the 3-D view chosen in AutoCAD. The SVG image can then be used like any other, but the 3-D view cannot change. Unsurprisingly, the power and versatility of XML has gained the attention of the Web3D Consortium. Under its former guise as the VRML Consortium, this non-profit organisation has been responsible for the creation and promotion of open-source 3-D technologies for nearly a decade. After developing VRML 1.0 and 2.0, they have chosen XML for the next implementation of VRML. Working in conjunction with the W3C and over 100 commercial and non-commercial concerns, the consortium has embraced XML and is in the process of creating an XML-based VRML application called X3D. The Web3D Consortium intends X3D to be compatible with VRML documents and tools, but with the ability to integrate with other XML technologies. Web3D has also opted for a modular approach, with different levels of implementation known as 'profiles', which can be modified individually to conform to the changing needs of developers. X3D is meant to replace VRML, but also to remain compatible. Web3D cites several factors for moving to an XML-based standard. XML has become the dominant syntax currently in use, whereas VRML uses its own unique syntax. Moving to a more universal syntax is meant to address the problems of forward migration and, of particular interest to archaeology, archival preservation of data. Web3D also sees XML as a way to simplify the way VRML is created, and to provide a tighter integration of VRML with other XML technologies. In particular, Web3D sees X3D as part of the XML family, to be used in conjunction with SVG, Synchronised Multimedia (SMIL) and XHTML (Anon. 2006l). Another addition to this list for archaeologists could be the Geography Markup Language (GML). GML is an XML application developed by the Open Geospatial Consortium for describing Geographic information (Anon. 2006m). As a specific tool for storing and transporting geographic data, GML is not designed for visualisation. It is intended for use with the other XML applications designed for this purpose, like X3D and SVG (Lake nd).

As part of the suite of W3C recommendations, SVG is consistent with the vision of the Web as non-proprietary, platform-independent and accessible. Adoption of W3C standards represents the best chance for fully realising healthy and sensible Web development, and it is important that archaeologists play an active role by investing time and resources into exploring applications like SVG. This is remarkably easy, as anyone can join in public forums like the Yahoo SVG Developers Group (Anon. 2006n), and most of the individuals currently involved with shaping the future of SVG can be found there. There are a wide variety of participants, including Chris Lilley and other members of the W3C working group, the authors of SVG titles, commercial and non-commercial developers, and other interested individuals. As such, archaeologists can make their problems and preferences known to other SVG developers and informally to the W3C. It is also important to promote the technology both within the discipline of archaeology and without. As a new medium, widespread use will only be possible if adopted by a significant number of people. If this is achieved, the result will be native implementation of SVG into all current Web browsers, and more sophisticated tools for developers (Laaker 2002, 16). The momentum of SVG has been recently demonstrated by an announcement from ESRI, the premier commercial GIS software developer, to support SVG in its upcoming release of ArcGIS, version 9.2 (Anon. 2006o).

XML provides a very flexible format for working with any kind of data that can be encoded as text (Harold and Means 2002, 8), and as a result of this, it has been chosen for use in many projects that are not Web-based. As the XML application for describing vector graphics, SVG has also been taken up for use outside of the Web. Whether a project was created for the Web or not, the ease with which a non-Web XML project can be made Web ready is also an advantage. Use of SVG is growing by leaps and bounds, as illustrated by the proceedings of 'SVG Open 2005'. Held in Enschede, the Netherlands, in August 2005, the fourth annual conference on Scalable Vector Graphics (Anon. 2005b) shows the breadth of creative ideas and applications already in use. With presentations of over 100 papers and posters, the conference reflects the great variety of people and projects already using SVG. Projects included commercial endeavours and academic explorations, with practitioners concentrated in, but not limited to, North America, Europe and Japan.

The future of SVG and archaeology looks bright. Basic incorporation of SVG into archaeological projects whenever vector images are appropriate can be implemented now with broad operating system and browser support. Adobe Illustrator makes creation of SVG images easily accessible at a reasonable cost, while allowing archaeologists to continue using the same vector-based programs. As work with SVG progresses, it will find uses beyond the Web like other XML applications. Recent professional conferences show the breadth of possibilities already being explored. SVG is an open-source, platform-independent technology that can be used for input of new archaeological data, retrospective conversion of data, as well as interpretation and visualisation of data. Used in combination with other members of the XML family and other forms of technology, it is a very powerful tool. SVG is not just a way of creating vector images; it is an elegant solution to archaeological visualisation.