Cars & Trucks

[10970010] |

WordPerfect

[10970020] |'''WordPerfect''' is a [[proprietary software|proprietary]] [[word processing]] application. [10970030] |At the height of its popularity in the late 1980s and early 1990s, it was the ''[[de facto]]'' standard word processor, but has since been eclipsed in sales by [[Microsoft Office Word|Microsoft Word]]. [10970040] |Although the [[MS-DOS]] and [[Microsoft Windows]] versions are best known, its popularity was based on the fact that it had been available for a wide variety of computers and operating systems, including [[Mac OS]], [[Linux]], the [[Apple IIe]], a separate version for the [[Apple IIgs]], most popular versions of [[Unix]], [[OpenVMS|VMS]], [[Data General]], [[System/370]], [[AmigaOS]], [[Atari ST]], [[OS/2]], and [[Nextstep|NeXTSTEP]]. [10970050] |==WordPerfect for DOS== [10970060] |WordPerfect was originally produced by [[Bruce Bastian]] and Dr. [[Alan Ashton (executive)|Alan Ashton]] who founded Satellite Software International, Inc. of [[Orem]], [[Utah]], which later renamed itself WordPerfect Corporation. [10970070] |Originally written for [[Data General]] minicomputers, in 1982 the developers ported the program to the IBM PC as WordPerfect 2.20, continuing the version numbering of the Data General series. [10970080] |The program's popularity took off with the introduction of WordPerfect 4.2 in 1986, with automatic paragraph numbering (important to the law office market), and the splitting of a lengthy footnote and its partial overflow to the bottom of the next page, as if it had been professionally typeset (valuable to both the law office and academic markets). [10970090] |WordPerfect 4.2 became the first program to overtake the original microcomputer word processor market leader, [[WordStar]], in a major application category on the [[DOS]] platform. [10970100] |In 1989, WordPerfect Corporation released the program's most successful version ever, WordPerfect 5.1 for DOS, which was the first version to include [[Macintosh]] style pull-down menus to supplement the traditional F-key combinations, as well as support for tables, a spreadsheet-like feature. [10970110] |The data format used by WordPerfect 5.1 was, for years, the most portable format in the world. [10970120] |All word processors could read (and convert) that format. [10970130] |Many conferences and magazines insisted that you shipped your documents in 5.1 format. [10970140] |Unlike previous DOS versions, WordPerfect 6.0 for DOS could switch between its traditional text-based editing mode and a graphical editing mode that [[WYSIWYG|showed the document as it would print out]], including fonts and text effects like bold, underline, and italics. [10970150] |The previous text-based versions used different colors or text color inversions to indicate various markups, and (starting with version 5.0) used a graphic mode only for an uneditable print preview that used generic fonts rather than the actual fonts that appeared on the printed page. [10970160] |===Key characteristics=== [10970170] |To this day, WordPerfect's three major characteristics that have differentiated from other market-leading word processors are its streaming code architecture, its Reveal Codes feature, and its unusually user-friendly macro/scripting language, PerfectScript. [10970180] |====Streaming code architecture==== [10970190] |A key to WordPerfect's design is its streaming code architecture that parallels the formatting features of [[HTML]] and [[Cascading Style Sheets]]. [10970200] |Documents are created much the same way that raw HTML pages are written, with text interspersed by tags that trigger treatment of data until a corresponding closing tag is encountered, at which point the settings active to the point of the opening tag resume control. [10970210] |As with HTML, tags can be nested. [10970220] |Some data structures are treated as objects within the stream as with HTML's treatment of graphic images, e.g., footnotes and styles, but the bulk of a WordPerfect document's data and formatting codes appear as a single continuous stream. [10970230] |====Styles and style libraries==== [10970240] |The addition of styles and style libraries in WP 5.0 provided greatly increased power and flexibility in formatting documents, while maintaining the streaming-code architecture of earlier versions. [10970250] |Prior to that, WordPerfect's only use of styles (a particular type of programming object) is the Opening Style, which contains the default settings for a document. [10970260] |====Reveal codes==== [10970270] |The Reveal Codes feature is a second editing screen that can be toggled open and closed at the bottom of the main editing screen. [10970280] |Text is displayed in Reveal Codes interspersed with tags and the occasional objects, with the tags and objects represented by named tokens. [10970290] |The scheme makes it far easier to untangle coding messes than with styles-based word processors, and object tokens can be clicked with a pointing device to directly open the configuration editor for the particular object type, e.g. clicking on a style token brings up the style editor with the particular style type displayed. [10970300] |WordPerfect users forced to change word processors by employers frequently complain on WordPerfect online forums that they are lost without Reveal Codes. [10970310] |Because of their style dependencies, efforts to create the equivalent of Reveal Codes in other word processors have produced disappointing results. [10970320] |Note that WordPerfect had this feature already in its DOS incarnations: it could be brought forward by pressing the keys 'Alt' and 'F3' together. [10970330] |====Macro languages==== [10970340] |WordPerfect for DOS was notable for its Alt-keystroke macro facility, which was expanded with the addition of macro libraries in WP 5.0 that also allowed for Ctrl-keystroke macros, and remapping of any key as a macro. [10970350] |This enabled any sequence of keystrokes to be recorded, saved, edited, and recalled. [10970360] |Macros could examine system data, make decisions, be chained together, and operate recursively until a defined 'stop' condition was met. [10970370] |This capability provided an amazingly powerful way to rearrange data and formatting codes within a document, where the same sequence of actions needed to be performed repetitively e.g. for tabular data. [10970380] |Macros can also be edited using WordPerfect Program Editor. [10970390] |Unfortunately, this facility could not easily be ported to the subsequent Windows versions. [10970400] |A new and even more powerful interpreted token-based macro recording and scripting language was introduced for both DOS and Windows 6.0 versions, and that became the basis of the language named PerfectScript in later versions. [10970410] |PerfectScript has remained the mainstay scripting language for WordPerfect users ever since. [10970420] |PerfectScript was specifically designed to be user-friendly, thus avoiding far less user-friendly methods of scripting languages implemented on other word processing programs that require education in advanced programming concepts such as Object Oriented Programming in order to produce useful yet sophisticated and powerful macros. [10970430] |===Function keys=== [10970440] |Like its mid-1980s competitor, [[MultiMate]], WordPerfect used almost every possible combination of [[function key]]s with Ctrl, Alt, and Shift modifiers. [10970450] |([[WordPerfect 4.1|See example help screen on this page]].) [10970460] |This was in contrast to [[WordStar]], which used only Ctrl, in conjunction with traditional typing keys. [10970470] |Many people still know and use the [[function key]] combinations from the DOS version, which were originally designed for Data General Dasher VDUs that supported 2 groups of 5 plain, shift, control, and control shift function keys. [10970480] |This was translated to the layout of the 1981 [[IBM PC keyboard]], with two columns of function keys at the left end of the keyboard, but worked even better with the 1984 PC AT keyboard with 3 groups of 4 function keys across the top of the keyboard. [10970490] |With the 1981 PC keyboard, the Tab key and the related F4 (''Indent'') functions were adjacent. [10970500] |This plethora of keystroke possibilities, combined with the developers' wish to keep the user interface free of "clutter" such as on-screen menus, made it necessary for most users to use a keyboard template showing each function. [10970510] |Infamously, WordPerfect used F3 instead of F1 for ''Help'', F1 instead of Esc for ''Cancel'', and Esc for ''Repeat'' (though a configuration option in later versions allowed these functions to be rotated to locations that later became more standard). [10970520] |===Printer drivers=== [10970530] |WordPerfect for DOS shipped with an impressive array of printer drivers - a feature that played an important role in its adoption - and also shipped with a [[printer driver]] editor called PTR, which features a flexible [[Macro (computer science)|macro language]] and allows technically-inclined users to customize and create printer drivers. [10970540] |Internally, WordPerfect used an extensive WordPerfect [[character set]] as its [[internal code]]. [10970550] |The precise meaning of the characters, although clearly defined and documented, can be overridden in its customizable printer drivers with PTR. [10970560] |The relationship between different type faces and styles, and between them and the various sections in the WordPerfect character set, were also described in the printer drivers and can be customized through PTR. [10970570] |===WordPerfect Library/Office=== [10970580] |WordPerfect Corporation produced a variety of ancillary and spin-off products. [10970590] |WordPerfect Library (introduced in 1986 and later renamed WordPerfect Office) was a package of network and stand-alone utilities for use with WordPerfect, primarily developed for offices running [[Novell NetWare]]. [10970600] |WordPerfect Library/Office included the DOS antecedents of what is now known as [[Novell GroupWise]], a shareable package of contact management, calendaring, and related word processing utilities. [10970610] |WordPerfect Library/Office a brand name later revived by Corel after it acquired ownership of WordPerfect and other programs still bundled under that product name as of this writing – included amongst other utilities a local area network (LAN) email facility and was the most popular such package in its day. [10970620] |====WordPerfect Shell==== [10970630] |The Library/Office bundle also included a noteworthy task-switching program that ran as a shell atop DOS, branded as WordPerfect Shell. [10970640] |Task-switchers were a popular application type for the DOS operating system because of its lack of multi-tasking, making it impractical to have many applications running at once. [10970650] |Task-switchers were programs that allocated available memory between open applications, allowing fast switching between open applications whose actions were suspended when the user switched to a different program. [10970660] |WordPerfect Shell 4.0, which was also bundled with the WordPerfect 6.x versions, had most functionality of the Windows 3.x shell but was far more versatile. [10970670] |Its automated memory management was superior to that of the Microsoft Windows shell, and Microsoft's product generally performed with far less frequent memory glitches when Windows was run as a program under Shell 4.0. [10970680] |The user interface for Shell is based on a hierarchical menu metaphor rather than the windows/folders/icons metaphor used by Microsoft. [10970690] |Shell 4.0's menu structures could be individually hot-keyed as pop-ups and its powerful menu editor allowed fast creation and editing of menu structures and menu items, with each menu item quickly configurable for entry of command lines and menu names. [10970700] |Shell 4.0 included 80 programmable clipboards, and the menu structures and menu items were also programmable using a scripting language whose scripts could themselves be chained to and from WordPerfect macros. [10970710] |The scripting language also included a keyboard buffer stuffing tool for control and operation of non-WordPerfect applications. [10970720] |Microsoft Windows had no answer to such powerful features other than a glitz of windows, icons, pointing devices, and an overwhelming marketing strategy. [10970730] |WordPerfect Shell was laid to rest along with many other popular DOS character-based tools inundated by Microsoft's marketing of Windows 95. [10970740] |Novell later licensed Shell 3.0 and 4.0 for free distribution. [10970750] |As of this writing it is still downloadable from the DataPerfect Users Group. [10970760] |WordPerfect Library/Office also included a Calculator, a flat-file database called Notebook that could be used by itself or in WordPerfect document merges, an exceptionally powerful relational database - [[DataPerfect]] - that retains a small but dedicated following despite having been dropped by WordPerfect Corporation in favour of Borland's Paradox as a companion of WP for Windows. [10970770] |Additional features continue to be added from time to time by DataPerfect's author, Lew Bastian - Bruce Bastian's older brother - a brilliant programmer who had written some of IBM's earliest disk-caching patents, and DataPerfect can now run as web server. [10970780] |LetterPerfect was a scaled down version of WordPerfect with the more advanced features removed but with file and (for the most part) keystroke compatibility. [10970790] |An implementation of Microsoft Visual Basic for Applications (VBA), introduced with WordPerfect for Windows 9.0, provides a full-featured development environment for building advanced custom WordPerfect solutions. [10970800] |These solutions are often created by corporate developers or programmers and may not be easily accessible to the typical WordPerfect user. [10970810] |For these users, PerfectScript is the better option. [10970820] |People who code scripts for WordPerfect use the Macros & Merges forum at WordPerfect Universe as their primary meeting ground. [10970830] |That site is a collaboration among other WordPerfect-related web site operators and others and functions as a portal to WordPerfect resources on the web. [10970840] |The site also maintains an extensive clip library for use in PerfectScript programming, has the Web's largest metalink library for locating online WordPerfect resources, and has the only peer-to-peer forum on the Web for DOS WordPerfect. [10970850] |The WordPerfect template and document file formats have remained remarkably stable since the WordPerfect 6.x DOS and Windows versions. [10970860] |Complete backward compatibility has been maintained and all WordPerfect versions since 6.0 have included a feature that stores any unrecognized codes in stream location represented in Reveal Codes by an "Unknown" token. [10970870] |Documents generated on newer versions can thus be edited in older versions with the codes retained. [10970880] |Then, upon being reopened in a newer version of WordPerfect, the "unknown" tokens regain their functionality. [10970890] |None of the newer WordPerfect features reflected in the file formats cause data loss when opened in older versions. [10970900] |==WordPerfect for Windows== [10970910] |===History=== [10970920] |WordPerfect was late in coming to market with a Windows version. [10970930] |The first mature version, WordPerfect 5.2 for Windows, was released in November [[1992]]. [10970940] |Prior to that, there was a WordPerfect 5.1 for Windows, introduced a year earlier. [10970950] |That version had to be installed from DOS and was largely unpopular due to serious stability issues. [10970960] |By the time WordPerfect 5.2 for Windows was introduced, [[Microsoft Word|Microsoft Word for Windows]] version 2 had been on the market for over a year and had received its third interim release, v2.0c. WordPerfect's function-key-centered user interface did not adapt well to the new paradigm of mouse and pull-down menus, especially with many of WordPerfect's standard key combinations pre-empted by incompatible keyboard shortcuts that Windows itself used (e.g. Alt-F4 became ''Exit Program'' as opposed to WordPerfect's ''Block Text''). [10970970] |The DOS version's impressive arsenal of finely tuned printer drivers was also rendered obsolete by Windows' use of its own printer device drivers. [10970980] |Internally, WordPerfect for Windows still used the WordPerfect character set as its internal code. [10970990] |This caused WordPerfect for Windows to be unable to support some languages — for example [[Chinese language|Chinese]] — that were natively supported by Windows. [10971000] |WordPerfect became part of an [[office suite]] when the company entered into a co-licensing agreement with [[Borland|Borland Software Corporation]] in 1993. [10971010] |The offerings were marketed as Borland Office, containing Windows versions of WordPerfect, [[Quattro Pro]], [[Borland Paradox]], and a LAN-based groupware package called WordPerfect Office (not to be confused with the complete applications suite of the same name later marketed by Corel) based on the WordPerfect Library for DOS. [10971020] |The WordPerfect product line was sold twice, first to [[Novell]] in June [[1994]], who then sold it to [[Corel]] in January [[1996]]. [10971030] |However, Novell kept the WordPerfect Office technology, incorporating it into its [[GroupWise]] messaging and collaboration product. [10971040] |Compounding WordPerfect's troubles were issues associated with the release of the first [[32-bit application|32-bit]] version, WordPerfect 7, intended for use on [[Windows 95]]. [10971050] |While it contained notable improvements over the [[16-bit application|16-bit]] WordPerfect for Windows 6.1, it was released in May [[1996]], nine months after the introduction of Windows 95 and Microsoft Office 95 (including [[Word for Windows|Word 95]]). [10971060] |The initial release suffered from notable stability problems. [10971070] |WordPerfect 7 also didn't have a Microsoft "Designed for Windows 95" logo. [10971080] |This was important to Windows 95 software purchasers as Microsoft set standards for application design, behavior, and interaction with the operating system. [10971090] |To make matters worse, the original release of WordPerfect 7 was incompatible with [[Windows NT]], hindering its adoption in academia. [10971100] |The "NT Enabled" version of WordPerfect 7, which Corel considered to be Service Pack 2, wasn't available until Q1-[[1997]], over 6 months after the introduction of [[Windows NT 4.0]], a year and a half after the introduction of Office 95 (which supported Windows NT out of the box), and shortly after the introduction of Office 97. [10971110] |Corel charged its customers to receive, what amounted to, a bug fix. [10971120] |While WordPerfect retained a majority of the retail shelf sales of word processors, Microsoft gained marketshare by including [[Microsoft Word|Word for Windows]] in its Windows product on new PCs. [10971130] |Microsoft gave discounts for Windows to OEMs who included Word on their PCs. [10971140] |When new PC buyers found Word installed on their new PC, Word began to dominate marketshare of desktop word processing. [10971150] |Amongst the remaining avid users of WordPerfect are many law firms and academics who favor the WordPerfect features such as macros and reveal codes. [10971160] |Corel now caters to these markets, with, for example, a major sale to the [[United States Department of Justice]] in 2005 . [10971170] |In November 2004, Novell filed an antitrust lawsuit against Microsoft for alleged anticompetitive behavior (viz, tying Word to sales of Windows) that Novell claims led to loss of WordPerfect market share . [10971180] |===Corel WordPerfect=== [10971190] |Since its acquisition by [[Corel]], WordPerfect for Windows has officially been known as '''Corel WordPerfect'''. [10971200] |== Unicode and Asian language editing == [10971210] |WordPerfect also lacks support for [[Unicode]]. [10971220] |The absence of support for [[Unicode]] limits its usefulness in many markets outside North America and Western Europe. [10971230] |Despite pleas from longtime users, this feature has not been implemented as of yet. [10971240] |For users in WordPerfect's traditional markets, the inability to deal with complex character sets, such as Asian language scripts, can cause difficulty when working on documents containing those characters. [10971250] |However, later versions have provided better compliance with interface conventions, file compatibility, and even Word interface emulation. [10971260] |== "Classic Mode" == [10971270] |Corel added "Classic Mode" in WordPerfect 11. [10971280] |==WordPerfect for Macintosh== [10971290] |Development of WordPerfect for Macintosh did not run parallel to versions for other operating systems, and used version numbers unconnected to contemporary releases for DOS, Windows, etc. [10971300] |The first release reminded users and reviewers of the DOS version, and was not especially successful in the marketplace. [10971310] |Version 2 was a total re-write, adhering more closely to Apple's UI guidelines. [10971320] |Version 3 took this further, making extensive use of the technologies Apple introduced in Systems 7.0–7.5, while remaining fast and capable of running well on older machines. [10971330] |Corel released version 3.5 in 1996, followed by the improved version 3.5e. [10971340] |It was never updated beyond that, and the product was eventually discontinued. [10971350] |[[As of 2004]], Corel has reiterated that the company has no plans to further develop WordPerfect for Macintosh (such as creating a native Mac OS X version). [10971360] |For several years, Corel allowed Mac users to download version 3.5e from their website free of charge, and some Mac users still use this version. [10971370] |The download is still available, along with the necessary OS 8/9/Classic Updater that slows scroll speed and restores functionality to the Style and Window menus. [10971380] |Like other Mac OS applications of its age, it requires the Classic environment on [[PowerPC]] Macs. [10971390] |While Intel Macs do not support Classic, emulators such as [[SheepShaver]], and [[vMac]] allow users to run WordPerfect and other Mac OS applications. [10971400] |Users wishing to use an up to date version of WordPerfect can run the Windows version through [[Boot Camp (software)|Boot Camp]] or a Windows emulator, and through [[Darwine]] or [[CrossOver Mac]] with mixed results. [10971410] |==WordPerfect for Linux== [10971420] |In 1995, WordPerfect 6.0 was made available for [[Linux]] as part of [[SCO Group|Caldera]]'s internet office package. [10971430] |In late 1997, a newer version was made available for download, but had to be purchased to be activated. [10971440] |Hoping to establish themselves in the nascent commercial Linux market, Corel also developed their [[Corel Linux|own distribution]] of Linux. [10971450] |Although the Linux distribution was fairly well-received, the response to WordPerfect for Linux was varied. [10971460] |Some Linux promoters appreciated the availability of a well-known, mainstream application for the OS. Developers of other Linux-compatible word processors questioned the need for another application in the category. [10971470] |Advocates of [[open-source software]] scoffed at its proprietary, closed-source nature, and questioned the viability of a commercial application in a market dominated by free software, such as [[OpenOffice.org]] and numerous others. [10971480] |The performance and stability of WordPerfect 9.0 (not a native Linux application like WP 6-8, but derived from the Windows version using the [[Wine (software)|Wine]] compatibility library) was highly criticized. [10971490] |WordPerfect failed to gain a large user base, and as part of Corel's change of strategic direction following a (non-voting) investment by Microsoft, WordPerfect for Linux was discontinued and their Linux distribution was sold to [[Xandros]]. [10971500] |In April 2004, Corel re-released WordPerfect 8.1 (the last Linux-native version) with some updates, as a "proof of concept" and to test the Linux market. [10971510] |[[As of 2005]], WordPerfect for Linux is not available for purchase. [10971520] |==Versions== [10971530] |(* - Part of [[WordPerfect Office]]) [10971540] |Known versions for VAX/VMS include 5.1, 5.3 and 7.1 , year of release unknown. [10971550] |Known versions for SUN include 6.0, requiring SunOS or Solaris 2, year of release unknown. [10971560] |Known versions for IBM System/370 include 4.2, released 1988. [10971570] |Known versions for OS/2 include 5.0, released 1989. [10971580] |Known versions for the DEC Rainbow 100 include version (?), released November 1983. [10971590] |In addition, versions of WordPerfect have also been available for Apricot, Atari ST, DEC Rainbow, Tandy 2000, TI Professional, Victor 9000, and Zenith Z-100 systems, as well as around 30 flavors of unix, including AT&T, NCR, SCO Xenix, Microport Unix, DEC Ultrix, Pyramid Tech Unix, Tru64, AIX, Motorola 8000, and HP9000 and SUN 3. [10971600] |==Current versions== [10971610] |On [[January 17]], [[2006]], Corel announced WordPerfect X3, the newest version of this office package. [10971620] |Corel is an original member of the [[OASIS (organization)|OASIS]] Technical Committee on the [[OpenDocument|Open Document Format]], and Paul Langille, a senior Corel developer, is one of the original four authors of the OpenDocument specification. [10971630] |In January 2006, subscribers to Corel's electronic newsletter were informed that WordPerfect 13 was scheduled for release later in 2006. [10971640] |The subsequent release of X3 (identified as "13" internally and in registry entries) has been met with generally positive reviews, due to new features including a unique PDF import capability, metadata removal tools, integrated search and online resources and other features. [10971650] |Version X3 was described by [[CNET]] in January, 2006 as a "winner", "a feature-packed productivity suite that's just as easy to use – and in many ways more innovative than – industry-goliath Microsoft Office 2003." [10971660] |CNET went on to describe X3 as "a solid upgrade for longtime users", but that "Die-hard Microsoft fans may want to wait to see what Redmond has up its sleeve with the radical changes expected within the upcoming Microsoft Office 12." [10971670] |While the notable if incremental enhancements of WordPerfect Office X3 have been well received by reviewers, a number of online forums have voiced concern about the future direction of WordPerfect, with longtime users complaining about certain usability and functionality issues that users have been asking to have fixed for the last few release versions. [10971680] |Although the released version of X3 does not support the [[Office Open XML|OOXML]] or [[OpenDocument]] formats, a beta has been released that supports both. [10971690] |Reports surfaced late in January 2006 that Apple's [[iWork]] had leapfrogged WordPerfect Office as the leading alternative to Microsoft Office. [10971700] |This claim was soon debunked after industry analyst Joe Wilcox described JupiterResearch usage surveys that showed WordPerfect as the No. 2 office suite behind Microsoft Office in the consumer, small and medium businesses, and enterprise markets with a roughly 15 percent share in each market. [10971710] |In April [[2008]] Corel released their WordPerfect Office X4 [[office suite]] containing the new X4 version of WordPerfect which includes support for [[PDF]], [[OpenDocument]] and [[Office Open XML]]. [10980010] |

Word sense disambiguation

[10980020] |In [[computational linguistics]], '''word sense disambiguation''' (WSD) is the process of identifying which [[word sense|sense]] of a [[word]] (having a number of distinct senses) is used in a given [[Sentence (linguistics)|sentence]]. [10980030] |For example, consider the word ''bass'', two distinct senses of which are: [10980040] |#a type of fish [10980050] |#tones of low frequency [10980060] |and the sentences: [10980070] |#''I went fishing for some sea bass'' [10980080] |#''The bass line of the song is very moving'' [10980090] |Explanation [10980100] |To a human it is obvious that the first sentence is using the word ''bass'' in the first sense above, and that in the second sentence it is being used in the second sense. [10980110] |Although this seems obvious to a human, developing [[algorithm]]s to replicate this human ability is a difficult task. [10980120] |==Difficulties== [10980130] |One problem with word sense disambiguation is deciding what the senses are. [10980140] |In cases like the word ''bass'' above, at least some senses are obviously different. [10980150] |In other cases, however, the different senses can be closely related (one meaning being a [[metaphor]]ical or [[metonymy|metonymic]] extension of another), and in such cases division of words into senses becomes much more difficult. [10980160] |Different dictionaries will provide different divisions of words into senses. [10980170] |One solution some researchers have used is to choose a particular dictionary, and just use its set of senses. [10980180] |Generally, however, research results using broad distinctions in senses have been much better than those using narrow, so most researchers ignore the fine-grained distinctions in their work. [10980190] |Another problem is inter-judge [[variance]]. [10980200] |WSD systems are normally tested by having their results on a task compared against those of a human. [10980210] |However, humans do not agree on the task at hand — give a list of senses and sentences, and humans will not always agree on which word belongs in which sense. [10980220] |A computer cannot be expected to give better performance on such a task than a human (indeed, since the human serves as the standard, the computer being better than the human is incoherent), so the human performance serves as an upper bound. [10980230] |Human performance, however, is much better on coarse-grained than fine-grained distinctions, so this again is why research on coarse-grained distinctions is most useful. [10980240] |==Approaches== [10980250] |As in all [[natural language processing]], there are two main approaches to WSD — deep approaches and shallow approaches. [10980260] |Deep approaches presume access to a comprehensive body of [[commonsense knowledge|world knowledge]]. [10980270] |Knowledge such as "you can go fishing for a type of fish, but not for low frequency sounds" and "songs have low frequency sounds as parts, but not types of fish" is then used to determine in which sense the word is used. [10980280] |These approaches are not very successful in practice, mainly because such a body of knowledge does not exist in computer-readable format outside of very limited domains. [10980290] |But if such knowledge did exist, they would be much more accurate than the shallow approaches. [10980300] |However, there is a long tradition in Computational Linguistics of trying such approaches in terms of coded knowledge, and in some cases it is hard to say clearly whether the knowledge involved is linguistic or world knowledge. [10980310] |The first attempt was that by Margaret Masterman and her colleagues at Cambridge Language Research Unit in England in the 1950s. [10980320] |This used as data a punched-card version of Roget's Thesaurus and its numbered "heads" as indicators of topics and looked for their repetitions in text, using a set intersection algorithm: it was not very successful (and is described in some detail in (Wilks, Y. et al., 1996) but had strong relationships to later work, especially Yarowsky's machine learning optimisation of a thesaurus method in the 1990s (see below). [10980330] |Shallow approaches don't try to understand the text. [10980340] |They just consider the surrounding words, using information like "if ''bass'' has words ''sea'' or ''fishing'' nearby, it probably is in the fish sense; if ''bass'' has the words ''music'' or ''song'' nearby, it is probably in the music sense." [10980350] |These rules can be automatically derived by the computer, using a training corpus of words tagged with their word senses. [10980360] |This approach, while theoretically not as powerful as deep approaches, gives superior results in practice, due to computers' limited world knowledge. [10980370] |It can, though, be confused by sentences like ''The dogs bark at the tree'', which contains the word ''bark'' near both ''tree'' and ''dogs''. [10980380] |These approaches normally work by defining a window of ''N'' content words around each word to be disambiguated in the corpus, and statistically analyzing those ''N'' surrounding words. [10980390] |Two shallow approaches used to train and then disambiguate are ''[[Naïve Bayes classifier]]s'' and ''[[decision tree]]s''. [10980400] |In recent research, kernel based methods such as [[support vector machine]]s have shown superior performance in [[supervised learning]]. [10980410] |But over the last few years, there hasn't been any major improvement in performance of any of these methods. [10980420] |It is instructive to compare the word sense disambiguation problem with the problem of [[part-of-speech tagging]]. [10980430] |Both involve disambiguating or tagging with words, be it with senses or parts of speech. [10980440] |However, algorithms used for one do not tend to work well for the other, mainly because the part of speech of a word is primarily determined by the immediately adjacent one to three words, whereas the sense of a word may be determined by words further away. [10980450] |The success rate for part-of-speech tagging algorithms is at present much higher than that for WSD, state-of-the art being around 95% accuracy or better, as compared to less than 75% accuracy in word sense disambiguation with supervised learning. [10980460] |These figures are typical for English, and may be very different from those for other languages. [10980470] |Another aspect of word sense disambiguation that differentiates it from part-of-speech tagging is the availability of training data. [10980480] |While it is relatively easy to assign parts of speech to text, training people to tag senses is far more difficult . [10980490] |While users can memorize all of the possible parts of speech a word can take, it is impossible for individuals to memorize all of the senses a word can take. [10980500] |Thus, many word sense disambiguation algorithms use [[semi-supervised learning]], which allows both labeled and unlabeled data. [10980510] |The [[Yarowsky algorithm]] was an early example of such an algorithm. [10980520] |Yarowsky’s [[Unsupervised learning|unsupervised algorithm]] uses the ‘One sense per collocation’ and the ‘One sense per discourse’ properties of human languages for word sense disambiguation. [10980530] |From observation, words tend to exhibit only one sense in most given discourse and in a given collocation. [10980540] |The corpus is initially untagged. [10980550] |The algorithm starts with a large corpus, in which it identifies examples of the given polysemous word, and stores all the relevant sentences as lines. [10980560] |For instance, Yarowsky uses the word ‘plant’ in his 1995 paper to demonstrate the algorithm. [10980570] |Assume that there are two possible senses of the word, the next step is to identify a small number of seed collocations representative of each sense, give each sense a label, i.e. sense A and B, then assign the appropriate label to all training examples containing the seed collocations. [10980580] |In this case, the words ‘life’ and ‘manufacturing’ are chosen as initial seed collocations for sense A and B respectively. [10980590] |The residual examples (85% - 98% according to Yarowsky) remain untagged. [10980600] |The algorithm should initially choose seed collocations representative that will distinguish sense A and B accurately and productively. [10980610] |This can be done by selecting seed words from a dictionary’s entry for that sense. [10980620] |The collocations tend to have stronger effect if they are adjacent to the target word, the effect weakens with distance. [10980630] |According to the criteria given in Yarowsky (1993), seed words that appear in the most reliable collocational relationships with the target word will be selected. [10980640] |The effect is much stronger for words in a predicate-argument relationship than for arbitrary associations at the same distance to the target word, and is much stronger for collocations with content words than with function words. [10980650] |Having said this, a collocation word can have several collocational relationships with the target word throughout the corpus. [10980660] |This could give the word different rankings or even different classifications. [10980670] |Alternatively, it can be done by identifying a single defining collocate for each class, and using for seeds only those contexts containing one of these defining words. [10980680] |A publicly available database called WordNet can be used as an automatic source for such defining terms. [10980690] |In addition, words that occur near the target word in great frequency can be selected as seed collocations representative. [10980700] |This approach is not fully automatic, a human judge must decide which word will be selected for each target word’s sense, the outputs will be reliable indicators of the senses. [10980710] |A decision-list algorithm is then used to identify other reliable collocations. [10980720] |This training algorithm calculates the probability P(Sense | Collocation), and the decision list is ranked by the log-likelihood ratio: '''Log( P(Sense_A | Collocation_i) / P(Sense_B | Collocation_i) )''' [10980730] |A [[smoothing]] algorithm will then be used to avoid 0 values. [10980740] |The decision-list algorithm resolves many problems in a large set of non-independent evidence source by using only the most reliable piece of evidence rather than the whole matching collocation set. [10980750] |The new resulting classifier will then be applied to the whole sample set. [10980760] |Add those examples in the residual that are tagged as A or B with probability above a reasonable threshold to the seed sets. [10980770] |Apply the decision-list algorithm and the above adding step iteratively. [10980780] |As more newly-learned collocations are added to the seed sets, the sense A or sense B set will grow, and the original residual will shrink. [10980790] |However, these collocations stay in the seed sets only if their probability of classification remains above the threshold, otherwise they are returned to the residual for later classification. [10980800] |At the end of each iteration, the ‘One sense per discourse’ property can be used to help preventing initially mistagged collocates and hence improving the purity of the seed sets. [10980810] |In order to avoid strong collocates becoming indicators for the wrong class, the class-inclusion threshold needs to be randomly altered. [10980820] |For the same purpose, after intermediate convergence the algorithm will also need to increase the width of the context window. [10980830] |The algorithm will continue to iterate until no more reliable collocations are found. [10980840] |The ‘One sense per discourse’ property can be used here for error correction. [10980850] |For a target word that has a binary sense partition, if the occurrences of the majority sense A exceed that of the minor sense B by a certain threshold, the minority ones will be relabeled as A. According to Yarowsky, for any sense to be clearly dominant, the occurrences of the target word should not be less than 4. [10980860] |When the algorithm converges on a stable residual set, a final decision list of the target word is obtained. [10980870] |The most reliable collocations are at the top of the new list instead of the original seed words. [10980880] |The original untagged corpus is then tagged with sense labels and probabilities. [10980890] |The final decision list may now be applied to new data, the collocation with the highest rank in the list is used to classify the new data. [10980900] |For example, if the highest ranking collocation of the target word in the new data set is of sense A, then the target word is classified as sense A. [10990010] |

XHTML

[10990020] |The '''''Extensible [[Hypertext]] Markup Language''''', or '''XHTML''', is a [[markup language]] that has the same depth of expression as [[HTML]], but also conforms to [[XML]] syntax. [10990030] |While HTML is an application of [[Standard Generalized Markup Language]] (SGML), a very flexible markup language, XHTML is an application of [[XML]], a more restrictive subset of SGML. [10990040] |Because they need to be [[XML#Well-formed_documents|well-formed]], true XHTML documents allow for automated processing to be performed using standard XML tools—unlike HTML, which requires a relatively complex, lenient, and generally custom [[parsing|parser]]. [10990050] |XHTML can be thought of as the intersection of HTML and XML in many respects, since it is a reformulation of HTML in XML. [10990060] |XHTML 1.0 became a [[World Wide Web Consortium]] (W3C) [[W3C recommendation|Recommendation]] on [[January 26]], [[2000]]. [10990070] |XHTML 1.1 became a W3C Recommendation on [[May 31]], [[2001]]. [10990080] |== Overview == [10990090] |XHTML is "a reformulation of the three HTML 4 document types as applications of XML 1.0". [10990100] |The [[W3C]] also continues to maintain the HTML 4.01 Recommendation and the specifications for [[HTML5]] and XHTML5 are being actively developed. [10990110] |In the current XHTML 1.0 Recommendation document, as published and revised to August 2002, the W3C comments that, "The XHTML family is the next step in the evolution of the Internet. [10990120] |By migrating to XHTML today, content developers can enter the XML world with all of its attendant benefits, while still remaining confident in their content's backward and future compatibility." [10990130] |=== Motivation === [10990140] |The need for a reformulated version of HTML was felt primarily because [[World Wide Web]] content now needs to be delivered to many devices (like [[Mobile computing|mobile device]]s) apart from traditional desktop [[computer]]s, where extra resources cannot be devoted to support the additional complexity of HTML syntax. [10990150] |In practice, however, HTML-supporting browsers for such constrained devices have emerged faster than XHTML support has been added to the desktop browser with the largest market share, [[Internet Explorer]]. [10990160] |Another goal for XHTML and XML was to reduce the demands on parsers and [[user agent]]s in general. [10990170] |With HTML, user agents increasingly took on the burden of "correcting" errant documents. [10990180] |Instead, XML requires user agents to give a "fatal" error when encountering malformed XML. [10990190] |In theory, this allows for vendors to produce leaner browsers, without the obligation to work around author errors. [10990200] |A side effect of this behavior is that those authoring XHTML documents and testing in conformant browsers should be more readily alerted to errors that may have gone otherwise unnoticed if the browser had attempted to render or ignore the malformed markup. [10990210] |A feature XHTML inherits from its XML underpinnings is XML [[Namespace (computer science)|namespaces]]. [10990220] |With namespaces, authors or communities of authors can define their own XML elements, attributes and content models to mix within XHTML documents. [10990230] |This is similar to the semantic flexibility of the class attribute in an [[HTML element]], but with fewer restrictions. [10990240] |Some W3C XML namespaces/schema that can be mixed with XHTML include [[MathML]] for semantic math markup, [[Scalable Vector Graphics]] for markup of vector graphics, and [[RDFa]] for embedding [[Resource Description Framework|RDF]] data. [10990250] |=== Relationship to HTML === [10990260] |HTML is the [[antecedent]] technology to XHTML. [10990270] |The changes from HTML to first-generation XHTML 1.0 are minor and are mainly to achieve conformance with XML. [10990280] |The most important change is the requirement that the document must be [[well-formed element|well-formed]] and that all [[HTML element|elements]] must be explicitly closed as required in XML. [10990290] |In XML, all element and attribute names are [[case-sensitive]], so the XHTML approach has been to define all tag names to be lowercase. [10990300] |This contrasts with some earlier established traditions which began around the time of HTML 2.0, when many used uppercase tags. [10990310] |In XHTML, all attribute values must be enclosed by quotes; either single (') or double (") quotes may be used. [10990320] |In contrast, this was sometimes optional in SGML-based HTML, where numeric or boolean attributes can omit quotes. [10990330] |All elements must also be explicitly closed, including empty (aka [[singleton]]) elements such as img and br. [10990340] |This can be done by adding a closing slash to the start tag, ''e.g.'', <img /> and <br />. [10990350] |Attribute minimization (e.g., <option selected>) is also prohibited, as the attribute selected contains no explicit value; instead this would be written as <option selected="selected">. [10990360] |HTML elements which are optional in the content model will not appear in the [[Document Object Model|DOM]] tree unless they are explicitly specified. [10990370] |For example, an XHTML page ''must'' have a <body> element, and a table will not have a <tbody> element unless the author specifies one. [10990380] |The XHTML 1.0 recommendation devotes a section to differences between HTML and XHTML.. [10990390] |The WHATWG wiki similarly considers differences that arise with the use of (X)HTML5.. [10990400] |Because XHTML and HTML are closely related technologies, sometimes they are written about and documented in parallel. [10990410] |In such circumstances, some authors conflate the two names by using a parenthetical notation, such as (X)HTML. [10990420] |This indicates that the documentation and principles can be considered to apply generally to both standards. [10990430] |=== Adoption === [10990440] |The similarities between HTML 4.01 and XHTML 1.0 led many web sites and content management systems to adopt the initial W3C XHTML 1.0 Recommendation. [10990450] |To aid authors in the transition, the W3C provided guidance on how to publish XHTML 1.0 documents in an HTML-compatible manner, and serve them to browsers that were not designed for XHTML. [10990460] |Such "HTML-compatible" content is sent using the HTML media type (text/html) rather than the official Internet media type for XHTML (application/xhtml+xml). [10990470] |When measuring the adoption of XHTML to that of regular HTML, therefore, it is important to distinguish whether it is media type usage or actual document contents that is being compared. [10990480] |Most web browsers have mature support for all of the possible XHTML media types. [10990490] |The notable exception is [[Internet Explorer]] by [[Microsoft]]; rather than rendering application/xhtml+xml content, a dialog box invites the user to save the content to disk instead. [10990500] |Both Internet Explorer 7 (released in 2006) and the initial beta version of Internet Explorer 8 (released in March 2008) exhibit this behaviour, and it is unclear whether this will be resolved in a future release. [10990510] |Whilst this remains the case, most web developers avoid using XHTML that isn’t HTML-compatible, so advantages of XML such as namespaces, faster parsing and smaller-footprint browsers do not benefit the user. [10990520] |Microsoft developer Chris Wilson explained in 2005 that IE7’s priorities were improved security and [[Cascading Style Sheet|CSS]] support, and that proper XHTML support would be difficult to graft onto IE’s compatibility-oriented HTML parser. [10990530] |Recently, notable developers have begun to question why Web authors ever made the leap into authoring in XHTML. [10990540] |In October 2006, HTML inventor and W3C chair [[Tim Berners-Lee]], explaining the motivation for the resumption of HTML (not XHTML) development, posted in his blog: "The attempt to get the world to switch to XML, including quotes around attribute values and slashes in empty tags and namespaces all at once didn't work. [10990550] |The large HTML-generating public did not move, largely because the browsers didn't complain." [10990560] |== Versions of XHTML == [10990570] |=== XHTML 1.0 === [10990580] |December 1998 saw the publication of a W3C Working Draft entitled ''Reformulating HTML in XML''. [10990590] |This introduced Voyager, the codename for a new markup language based on HTML 4 but adhering to the stricter syntax rules of XML. [10990600] |By February 1999 the specification had changed name to ''XHTML™ 1.0: The Extensible HyperText Markup Language'', and in January 2000 it was officially adopted as a W3C Recommendation. [10990610] |There are three formal [[Document Type Definition|DTDs]] for XHTML 1.0, corresponding to the three different versions of HTML 4.01: [10990620] |* '''XHTML 1.0 Strict''' is the equivalent to strict HTML 4.01, and includes elements and attributes that have not been marked deprecated in the HTML 4.01 specification. [10990630] |* '''XHTML 1.0 Transitional''' is the equivalent of HTML 4.01 Transitional, and includes the presentational elements (such as center, font and strike) excluded from the strict version. [10990640] |* '''XHTML 1.0 Frameset''' is the equivalent of HTML 4.01 Frameset, and allows for the definition of [[frameset|frameset documents]]—a common Web feature in the late 1990s. [10990650] |The second edition of XHTML 1.0 became a W3C Recommendation in August 2002. [10990660] |=== Modularization of XHTML === [10990670] |The initial draft of ''Modularization of XHTML'' became available in April 1999, and reached Recommendation status in April 2001. [10990680] |[[XHTML Modularization|Modularization]] provides an abstract collection of components through which XHTML can be subsetted and extended. [10990690] |The feature is intended to help XHTML extend it’s reach onto emerging platforms, such as mobile devices and Web-enabled televisions. [10990700] |The first XHTML Family Markup Languages to be developed with this technique were XHTML 1.1 and XHTML Basic 1.0. [10990710] |Another example is XHTML-Print (W3C Recommendation, September 2006), a language designed for printing from mobile devices to low-cost printers. [10990720] |In 2008 ''Modularization of XHTML'' is expected to be superseded by ''XHTML Modularization 1.1'', which adds an [[XML Schema (W3C)|XML Schema]] implementation. [10990730] |=== XHTML 1.1—Module-based XHTML === [10990740] |XHTML 1.1 evolved out of the work surrounding the initial ''Modularization of XHTML'' specification. [10990750] |The W3C released a first draft in September 1999; Recommendation status was reached in May 2001. [10990760] |The modules combined within XHTML 1.1 effectively recreate XHTML 1.0 Strict, with the addition of [[ruby character|ruby annotation]] elements (ruby, rbc, rtc, rb, rt and rp) to better support East-Asian languages. [10990770] |Other changes include removal of the lang attribute (in favour of xml:lang), and removal of the name attribute from the a and map elements. [10990780] |Although XHTML 1.1 is largely compatible with XHTML 1.0 and HTML 4, in August 2002 the W3C issued a formal Note advising that it should not be transmitted with the HTML media type. [10990790] |With limited browser support for the alternate application/xhtml+xml media type, XHTML 1.1 has so far proven unable to gain widespread use. [10990800] |XHTML 1.1 Second Edition is expected in the third quarter of 2008. [10990810] |=== XHTML Basic and XHTML-MP === [10990820] |To support constrained devices, ''[[XHTML Basic]]'' was created by the W3C; it reached Recommendation status in December 2000. [10990830] |XHTML Basic 1.0 is the most restrictive version of XHTML, providing a minimal set of features that even the most limited devices can be expected to support. [10990840] |The [[Open Mobile Alliance]] and it’s predecessor the WAP Forum released three specifications between 2001 and 2006 that extended XHTML Basic 1.0. [10990850] |Known as [[XHTML Mobile Profile]] or XHTML-MP, they were strongly focussed on uniting the differing markup languages used on [[mobile phone|mobile handsets]] at the time. [10990860] |All provide richer form controls than XHTML Basic 1.0, along with varying levels of scripting support. [10990870] |''XHTML Basic 1.1'' became a W3C Proposed Recommendation in June 2008, superseding XHTML-MP 1.2. [10990880] |XHTML Basic 1.1 is almost but not quite a subset of regular XHTML 1.1. [10990890] |The most notable addition over XHTML 1.1 is the inputmode attribute—also found in XHTML-MP 1.2—which provides hints to help browsers improve form entry. [10990900] |=== XHTML 1.2 === [10990910] |The XHTML 2 Working Group is considering the creation a new language based on XHTML 1.1. [10990920] |If XHTML 1.2 is created, it will include [[WAI-ARIA]] and role attributes to better support accessible web applications, and improved [[Semantic Web]] support through [[RDFa]]. [10990930] |The inputmode attribute from XHTML Basic 1.1, along with the target attribute (for specifying [[Framing (World Wide Web)|frame]] targets) may also be present. [10990940] |=== XHTML 2.0 === [10990950] |Between August 2002 and July 2006 the W3C released the first eight Working Drafts of XHTML 2.0, a new version of XHTML able to make a clean break from the past by discarding the requirement of backward compatibility. [10990960] |This lack of compatibility with XHTML 1.x and HTML 4 caused some early controversy in the web developer community. [10990970] |Some parts of the language (such as the role and RDFa attributes) were subsequently split out of the specification and worked on as separate modules, partially to help make the transition from XHTML 1.x to XHTML 2.0 smoother. [10990980] |A ninth draft of XHTML 2.0 is expected to appear in 2008. [10990990] |New features introduced by XHTML 2.0 include: [10991000] |* HTML forms will be replaced by [[XForms]], an XML-based user input specification allowing forms to be displayed appropriately for different rendering devices. [10991010] |* HTML frames will be replaced by [[XFrames]]. [10991020] |* The [[DOM Events]] will be replaced by [[XML Events]], which uses the XML [[Document Object Model]]. [10991030] |* A new list element type, the nl element type, will be included to specifically designate a list as a navigation list. [10991040] |This will be useful in creating nested menus, which are currently created by a wide variety of means like nested unordered lists or nested definition lists. [10991050] |* Any element will be able to act as a [[hyperlink]], e.g., Articles, similar to [[XLink]]. [10991060] |However, XLink itself is not compatible with XHTML due to design differences. [10991070] |* Any element will be able to reference alternative media with the src attribute, e.g., London Bridge is the same as . [10991080] |* The alt attribute of the img element has been removed: alternative text will be given in the content of the img element, much like the object element, e.g., HMS Audacious. [10991090] |* A single heading element (h) will be added. [10991100] |The level of these headings are determined by the depth of the nesting. [10991110] |This allows the use of headings to be infinite, rather than limiting use to six levels deep. [10991120] |* The remaining presentational elements i, b and tt, still allowed in XHTML 1.x (even Strict), will be absent from XHTML 2.0. [10991130] |The only somewhat presentational elements remaining will be sup and sub for superscript and subscript respectively, because they have significant non-presentational uses and are required by certain languages. [10991140] |All other tags are meant to be [[semantic]] instead (e.g. for strong or bolded text) while allowing the user agent to control the presentation of elements via CSS. [10991150] |* The addition of RDF triple with the property and about attributes to facilitate the conversion from XHTML to RDF/XML. [10991160] |=== HTML 5—Vocabulary and APIs for HTML and XHTML === [10991170] |[[HTML 5]] initially grew independently of the W3C, through a loose group of browser manufacturers and other interested parties calling themselves the [[WHATWG]], or Web Hypertext Application Technology Working Group. [10991180] |The WHATWG announced the existence of an open mailing list in June 2004, along with a website bearing the strapline “Maintaining and evolving HTML since 2004.” [10991190] |The key motive of the group was to create a platform for dynamic web applications; they considered XHTML 2.0 to be too document-centric, and not suitable for the creation of forum sites or online shops. [10991200] |In April 2007, the Mozilla Foundation and Opera Software joined Apple in requesting that the newly rechartered HTML Working Group of the W3C adopt the work, under the name of HTML 5. [10991210] |The group resolved to do this the following month, and the First Public Working Draft of HTML 5 was issued by the W3C in January 2008. [10991220] |The most recent W3C Working Draft was published in June 2008. [10991230] |HTML 5 has both a regular text/html serialization and an XML serialization, which is known as XHTML 5. [10991240] |In addition to the markup language, the specification includes a number of [[application programming interfaces]]. [10991250] |The [[Document Object Model]] is extended with APIs for editing, drag-and-drop, data storage and network communication. [10991260] |The language can be considered more compatible with HTML 4 and XHTML 1.x than XHTML 2.0, due to the decision to keep the existing HTML form elements and events model. [10991270] |It adds many new elements not found in XHTML 1.x, however, such as section and aside. [10991280] |(The XHTML 1.2 equivalent of these structural elements would be and .) [10991290] |The specification is expected to add WAI-ARIA support in a future draft. [10991300] |There is currently no indication as to whether HTML 5 will support RDFa, or be limited just to [[microformats]]. [10991310] |== Valid XHTML documents == [10991320] |An XHTML document that conforms to an XHTML specification is said to be ''valid''. [10991330] |Validity assures consistency in document code, which in turn eases processing, but does not necessarily ensure consistent rendering by browsers. [10991340] |A document can be checked for validity with the [[W3C Markup Validation Service]]. [10991350] |In practice, many web development programs such as [[Dreamweaver]] provide code validation based on the [[W3C]] standards. [10991360] |=== DOCTYPEs === [10991370] |In order to validate an XHTML document, a [[Document Type Declaration]], or ''DOCTYPE'', may be used. [10991380] |A DOCTYPE declares to the browser which [[Document Type Definition]] (DTD) the document conforms to. [10991390] |A Document Type Declaration should be placed before the [[root element]]. [10991400] |The [[system identifier]] part of the DOCTYPE, which in these examples is the [[Uniform Resource Locator|URL]] that begins with ''http://'', need only point to a copy of the DTD to use if the validator cannot locate one based on the [[public identifier]] (the other quoted string). [10991410] |It does not need to be the specific URL that is in these examples; in fact, authors are encouraged to use local copies of the DTD files when possible. [10991420] |The public identifier, however, must be character-for-character the same as in the examples. [10991430] |These are the most common XHTML Document Type Declarations: [10991440] |;XHTML 1.0 Strict [10991450] |:<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN""http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> [10991460] |;XHTML 1.0 Transitional [10991470] |:<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN""http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> [10991480] |;XHTML 1.0 Frameset [10991490] |:<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Frameset//EN""http://www.w3.org/TR/xhtml1/DTD/xhtml1-frameset.dtd"> [10991500] |;XHTML 1.1 [10991510] |:<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.1//EN""http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd"> [10991520] |;HTML 5 [10991530] |HTML5 does not require a doctype, and HTML 5 validation is not DTD-based. [10991540] |;XHTML 2.0 [10991550] |XHTML 2.0, [[As of April 2008]], is in a draft phase. [10991560] |If an XHTML 2.0 Recommendation is published with the same document type declaration as in the current Working Draft, the declaration will appear as: [10991570] |:<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 2.0//EN" "http://www.w3.org/MarkUp/DTD/xhtml2.dtd"> [10991580] |A placeholder DTD schema exists at the corresponding URI, though it currently only includes the character reference entities from previous recommendations. [10991590] |XHTML 2 contemplates both a version attribute and an xsi:schemalocation attribute on the root HTML element that could possibly serve as a substitute for any DOCTYPE declaration. [10991600] |==== XML namespaces and schemas ==== [10991610] |In addition to the DOCTYPE, all XHTML elements must be in the appropriate [[XML namespace]] for the version being used. [10991620] |This is usually done by declaring a default namespace on the root element using xmlns="namespace" as in the example below. [10991630] |For XHTML 1.0, XHTML 1.1 and HTML5, this is [10991640] |:<html xmlns="http://www.w3.org/1999/xhtml"> [10991650] |XHTML 2.0 requires both a namespace and an [[W3C XML Schema|XML Schema]] instance declaration. [10991660] |These might be declared as [10991670] |:<html xmlns="http://www.w3.org/2002/06/xhtml2/" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"xsi:schemaLocation="http://www.w3.org/2002/06/xhtml2/ http://www.w3.org/MarkUp/SCHEMA/xhtml2.xsd"> [10991680] |This example for XHTML 2.0 also demonstrates the use of multiple namespaces within a document. [10991690] |The first xmlns default namespace declaration indicates that elements and attributes whose names have no XML namespace prefix fall within the XHTML 2.0 namespace. [10991700] |The second namespace prefix declaration xmlns:xsi indicates that any elements or attributes prefixed with the xsi: refer to the XMLSchema-Instance namespace. [10991710] |Through this namespace mechanism XML documents allow the use of a mixture of elements and attributes taken from various XML vocabularies while avoiding the potential for clashes of naming between items from independently developed vocabularies. [10991720] |Similar to the case of DOCTYPE above, the actual URL to the [[W3C XML Schema|XML Schema]] file can be changed, as long as the [[Universal Resource Identifier]] (URI) before it (which indicates the XHTML 2.0 namespace) remains the same. [10991730] |The namespace URI is intended to be a persistent and universally unique identifier for the particular version of the specification. [10991740] |If treated as a URL, the actual content located at the site is of no significance. [10991750] |==== XML Declaration ==== [10991760] |A [[character encoding]] may be specified at the beginning of an XHTML document in the XML declaration when the document is served using the application/xhtml+xml MIME type. [10991770] |(If an XML document lacks encoding specification, an XML parser assumes that the encoding is [[UTF-8]] or [[UTF-16]], unless the encoding has already been determined by a higher protocol.) [10991780] |For example: [10991790] |: [10991800] |The declaration may be optionally omitted because it declares as its encoding the default encoding. [10991810] |However, if the document instead makes use of XML 1.1 or another character encoding, a declaration is necessary. [10991820] |[[Internet Explorer]] prior to version 7 enters [[quirks mode]] if it encounters an XML declaration in a document served as text/html. [10991830] |=== Common errors === [10991840] |Some of the most common errors in the usage of XHTML are: [10991850] |* Failing to realize that documents won’t be treated as XHTML unless they are served with an appropriate XML [[Internet media type|MIME type]] [10991860] |* Not closing empty elements (elements without closing tags in HTML4) [10991870] |** Incorrect: [10991880] |** Correct: [10991890] |Note that any of these are acceptable in XHTML: <br></br>, <br/> and <br />. [10991900] |Older HTML-only browsers interpreting it as HTML will generally accept <br> and <br />. [10991910] |* Not closing non-empty elements [10991920] |** Incorrect: This is a paragraph.
This is another paragraph. [10991930] |** Correct:
This is a paragraph.
This is another paragraph.
[10991940] |* Improperly nesting elements (Note that this would also be invalid in [[HTML]]) [10991950] |** Incorrect: This is some text. [10991960] |** Correct: This is some text. [10991970] |* Not putting quotation marks around attribute values [10991980] |** Incorrect: [10991990] |** Correct: [10992000] |** Correct: [10992010] |* Using the ampersand character outside of entities [10992020] |** Incorrect: Cars & Trucks [10992030] |** Correct: Cars & Trucks [10992040] |* Using the ampersand outside of entities in [[Uniform Resource Locator|URL]]s (Note that this would also be invalid in [[HTML]]) [10992050] |** Incorrect: News [10992060] |** Correct: News [10992070] |* Failing to recognize that XHTML elements and attributes are case sensitive [10992080] |** Incorrect: The Best Page Ever [10992090] |** Correct: The Best Page Ever [10992100] |* Using attribute minimization [10992110] |** Incorrect: READ-ONLY [10992120] |** Correct: READ-ONLY [10992130] |* Mis-using CDATA, script-comments and xml-comments when embedding scripts and stylesheets. [10992140] |** This problem can be avoided altogether by putting all script and stylesheet information into separate files and referring to them as follows in the XHTML head element. [10992150] |
<link rel="stylesheet" href="/style/screen.css" type="text/css" /> <script type="text/javascript" src="/script/site.js"></script>
[10992160] |::Note: The format <script …></script>, rather than the more concise <script … />, is required for HTML compatibility when served as MIME type text/html. [10992170] |** If an author chooses to include script or style data inline within an XHTML document, different approaches are recommended depending whether the author intends to serve the page as application/xhtml+xml and target only fully conformant browsers, or serve the page as text/html and try to obtain usability in Internet Explorer 6 and other non-conformant browsers. [10992180] |== Backward compatibility == [10992190] |XHTML 1.x documents are mostly backward compatible with HTML 4 user agents when the appropriate guidelines are followed. [10992200] |XHTML 1.1 is essentially compatible, although the elements for [[ruby character|ruby annotiation]] are not part of the HTML 4 specification and thus generally ignored by HTML 4 browsers. [10992210] |Later XHTML 1.x modules such as those for the role attribute, [[RDFa]] and [[WAI-ARIA]] degrade gracefully in a similar manner. [10992220] |HTML 5 and XHTML 2 are significantly less compatible, although this can be mitigated to some degree through the use of scripting. [10992230] |(This can be simple one-liners, such as the use of “document.createElement()” to register a new HTML element within Internet Explorer, or complete JavaScript frameworks, such as the [[FormFaces]] implementation of [[XForms]].) [10992240] |=== Examples === [10992250] |The followings are examples of XHTML 1.0 Strict. [10992260] |Both of them have the same visual output. [10992270] |The former one follows the [http://www.w3.org/TR/xhtml1/#guidelines HTML Compatibility Guidelines] in Appendix C of the XHTML 1.0 Specification while the latter one breaks backward compatibility but provides cleaner codes. [10992280] |Example 1. XHTML 1.0 Example
This is an example of an XHTML 1.0 Strict document.
[10992290] |Example 2. XHTML 1.0 Example
This is an example of an XHTML 1.0 Strict document.