This page and effort is failing to follow the microformats process. If you are the author of this page, please join the #microformats IRC channel and/or microformats-discuss mailing list so that the community can help walk you through the process. - Tantek
This page is now abandoned due to non conformance. Some of this material may appear on new pages following the proper process. - DimitriosZachariadis
4d: A geo microformat alternative proposal
The term 4d stands for four dimensions and is used interchangeably with spacetime in this page, as heaving the same meaning. When in bold, it means the proposed microformat 4d.
- DimitriosZachariadis 17:03, 17 Jan 2007 (PST)
A compact microformat for defining locations and events is proposed, as an alternative for the draft geo microformat. A tolerant five part, semicolon separated, spacetime and reference-system value, is argued to be sufficient to provide accurate information for uniquely identifying a place or an event in a four dimensional universe.
Spacetimes are the arenas in which all physical events take place — for example, the motion of planets around the Sun may be described in a particular type of spacetime, or the motion of light around a rotating star may be described in another type of spacetime. The basic elements of spacetime are events. In any given spacetime, an event is a unique position at a unique time. Examples of events include the explosion of a star or the single beat of a drum. Spacetime, Wikipedia (annotation by the author)
Location and geo referencing microformats proposed on this site, namely geo, luna and mars have at present (Jan 2007) a status of draft. Dealing with the issue of geo tagging requires a consensus on the issues of importance, if the microformat is to be successful and used without problems. To that extend, a proposal for a new microformat might not be superfluous, if it can present and offer for discussion some of the issues at hand. With that ambition, the 4d microformat is presented and discussed herein.
Most of the location related information, whether about the Earth or another celestial body, have a time related aspect attached; historical places and events, geographical features, but also street names and addresses are tightly related to an act or an observation made at some point in time; addresses used today may have not been in use a century ago (assuming the place identified by the coordinates did exist), or may well change to something different tomorrow. Landmarks existing for centuries have an age, which means they have a birthday, but they may also one day disappear.
A vast amount of information used by people on a daily basis in the web is in fact an aggregation of events; stories and descriptions are most of the time nothing more than annotations of events. As such, most of this information can be accurately tagged, stored and used for as long as it exists, if its spacetime dimensions are known.
Default values and ambiguities
Disambiguation of the values used for tagging is an important issue, if the data tagged are to be valid, accurate and and useful outside the assumptions made at the time of their creation.
Omitting dimensions from a 4d coordinate system introduces assumptions about their default values. To demonstrate the issues that need to be dealt with, the following example of a markup, tagged with the geo microformat, might prove useful:
In 1687, a Turkish ammunition dump <a class="geo" href="http://maps.google.com/maps?q=37.971508,23.72658&ie=UTF8&z=18&ll=37.971508,23.72658 &spn=0.002584,0.006083&t=k&om=1&iwloc=addr"> inside the building <abbr class="latitude" title="37.971508"></abbr>, <abbr class="longitude" title="23.72658"></abbr> </a> was ignited by a Venetian cannonball. The resulting explosion severely damaged the Parthenon and its sculptures.
which may be rendered as:
In the case of the markup above, a human reader would assume that the coordinates involved in the code have been taken using WGS86, the most known and used datum today, due to its global validity and the proliferation of the GPS receivers. The ambiguity of altitude leads to an assumption of a default value of 0, which is acceptable for 2D cartography. No assumption can be made about time.
If the geo portion of the above markup is reduced to:
<a class="geo" href="http://maps.google.com/maps?q=37.971508,23.72658&ie=UTF8&z=18&ll=37.971508,23.72658 &spn=0.002584,0.006083&t=k&om=1&iwloc=addr"> inside the building <abbr class="latitude" title="37.971508"></abbr>, </a>
the markup becomes useless for identifying locations, since it no more represents a location (but rather a geometric locus). If this is the case, then it seems that there is no point in having latitude and longitude defined by separate tags; indeed, the meaning of a coordinate is that it is an ordered list of numbers, or a tuple. Wikipedia, Coordinates. RFC2426 also specifies the GEO type as a single lat;long entity.
It could be argued that having a simpler, albeit verbose, format, helps content formating tools like XSLT processors, locate data easier. However, Microformats are a way of adding simple markup to human-readable data items such as events, contact details or locations, on web pages, so that the information in them can be extracted by software and indexed, searched for, saved, cross-referenced or combined. Microformats are not meant to help format content for humans, but to help machines gather information in a Semantic Web. On the other hand, a machine can split the data of a 4d tag in one single line of code.
As argued above, the inclusion of a time dimension in geo markup can add significantly to the semantics of a piece of geo information, as it moves the focus of geo tagging from the realm of two dimensional cartography to the realm of events.
This should come as no surprise: events are about stories; cartography is about navigation. People are interested in stories, news and events; few people can successfully navigate using a map and even fewer are interested in navigation and cartography as a science or art. Geographic coordinates shown in the text are quite useless without a map, and are even more useless if an electronic map is available for use. Humans use names to identify places, not numbers. The reason there is a need for showing geographic coordinates in text, is mostly to help humans manually identify places on maps. If this can be done electronically, then numbers don't seem to matter.
A human visiting the page with the markup above (which was chosen for the length of the time dimension and the connotations involved), will have a chance to read the text (and the year of the event, which however is un-tagged), click at the link and get a Google Map centered at the Acropolis and the Parthenon. Semantic connotations for humans enhance their understanding of information presented to them. However, a machine seeking information in the Semantic Web, will have no way of finding out if it is the 5th century BC Athenian state or the contemporary Greek state that this particular piece of information is referring to. Semantic information about that particular markup that could otherwise be cataloged by the machine, will pass unnoticed.
Had this geo markup been amended with a time dimension, a robot crawling this and other similar pages could create a list of historical events, based on nothing else but the information provided by the 4d microformat.
Humans have learned to deal with time efficiently, in their locality, through the use of clocks. Wrist watches remind people of upcoming events in their lives. An event in a remote place is more difficult to conceive, but not entirely impossible. Although it involves a simple addition or subtraction, only people traveling a lot can do it with ease.
Counting time in multiple reference systems is however a difficult task. It might not be necessary for humans in their daily life, but it is important when dealing with information on a global, or universal basis. Different cultures, use different reference systems to tell the time. Islamic calendar These need to be taken into account in a Semantic Web.
In a similar manner, trying to addapt to a Martian reference system, while living on Earth, involves more than a simple addition/subtraction of a few hours. A Martian day, is not of the same duration as a Terran day, and the same is true for the duration of the seasons; a number of connotations, that would otherwise help in getting a gut feeling, fail helplessly.
The reference system
It should be noted that the assumption about the datum used in expressing the lat-long coordinates mentioned above, is not a trivial one: Only 20 years ago, the same coordinates would possibly point hundreds of meters away from the Parthenon, since the reference system used at that time was different than WGS84. Furthermore, assuming WGS84 was the datum for the markup above, these coordinates may not be accurate 10 years from today, when the WGS84 reference system will have been revised once again.
A markable indication of the importance of the reference system, when expressing geo coordinates, is the fact that the Greenwich Observatory, which was by definition the origin for the longitude coordinate for more than a century, lies now about 102.5m West of the WGS84 0.0 meridian, at N 51° 28' 36.71, W 0° 0' 5.18", (in WGS84 datum) according to Wikipedia, Prime_Meridian. Interestingly, Google maps and Wikipedia do not seem to agree on these coordinates map
It should be clear that, for a microformat to be able to convey accurate information, well defined and known reference systems should be used. By performing transformations among the various reference systems, the content referenced in one system can be understandable and usefull on another, on a global or universal basis.
Naming a geo microformat 4d helps avoid the problem of the geo-centric (i.e. Earth bound) root of the word geo; 4d can be easily used on any celestial body for which a reference system, even rudimentary, has been defined. To that extend, Mars related content can be readily microformatted and displayed on Google Mars. Obviously, matters regarding reference systems of other celestial bodies require expert knowledge, however, reading the news on NASA's site, e.g. for Titan, Two Sides of Dunes, reveals that reference systems are in place for all celestial bodies visited by a man made spacecraft, and probably for a lot more than those.
4d is a rather common acronym; Google yields about 53.9 million hits, while for 3d it yields 310 million hits and for geo 82.9 million.
It is proposed that the 4d microformat be formed as follows:
Root Class Name
The root class name for a 4d location/event is 4d
The 4d microformat has the following no properties.
The value of a 4d location/event resides in the title attribute of an (X)HTML tag as a semicolon separated 5-tuple:
v1;v2;z;t;u where: v1 is either latitude or x, depending on u, mandatory v2 is either longitude or y, depending on u, mandatory z is the altitude, optional t is the time, optional u is the reference system code, optional
Dimensions z, t, and u can be omitted. If abiguities arrise from such an omition, e.g. omiting the z and t dimensions but defining the u dimension, then the relevant field values must be left empty, e.g.: v1;v2;;;u
Default Reference System
The default reference system is WGS84 and the Gregorian calendar.
The default reference system may be set, using the name and content attributes of the meta (X)HTML element, as follows:
<meta name="4d.reference" content="reference-system"/>
Default coordinate values
When a dimension is omitted, the following default values are implied:
u: WGS84 v1: latitude, expressed in decimal degrees, no default v2: longitude, expressed in decimal degrees, no default z: altitude, expressed in meters, defaults to 0.0 t: time, no default
The default reference system for an (X)HTML page can be set by using meta data properties.
It is strongly recommended that web pages include a meta data element that defines the default reference system, in their HTML head section. The inclusion of the meta data element ensures the validity of the 4d data for as long as the page exists, while keeping 4d values compact. By adding the 4d metadata element, a conversion from the geo microformat to the 4d would only require the change of the root class, from geo to 4d.
Indicative reference systems include:
- v1 is x, expressed in meters
- v2 is y, expressed in meters
- t is expressed in ISO-8601
The markup used in the discussion, transferred in the 4d microformat:
In 1687, a Turkish ammunition dump <a class="4d" href="http://maps.google.com/maps?q=37.971508,23.72658&ie=UTF8&z=18&ll=37.971508,23.72658 &spn=0.002584,0.006083&t=k&om=1&iwloc=addr"> inside the building <abbr class="4d" title="37.971508;23.72658"></abbr> </a> was ignited by a Venetian cannonball. The resulting explosion severely damaged the Parthenon and its sculptures.
Hidden 4d coordinates, no content
The Parthenon was ruined by a Venetian cannonball in 1687 <abbr class="4d" title="37.971508;23.72658;;1687"></abbr>
Hidden 4d coordinates, free text content
<span class="4d" title="37.971508;23.72658;;1687"> The Parthenon was ruined by a Venetian cannonball in 1687 </span>
may be rendered as:
The Parthenon was ruined by a Venetian cannonball in 1687
Visible 4d coordinates
<abbr class="4d" title="37.971508;23.72658;;1687">N 37° 58' 17.43, E 23° 43' 35.69</abbr>
may be rendered as:
N 37° 58' 17.43, E 23° 43' 35.69
Minimum HTML page:
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> <html> <head> <meta http-equiv="Content-Type" content="text/html; charset=utf-8" /> <meta name="4d.reference" content="WGS86"/> </head> <body> <div class="4d" title="51.476864;-0.000518"> The Greenwich Observatory </div> </body> </html>
A fictitious Mars reference system, using actual Mars coordinates from Google Mars:
<span class="4d" title="18.302380;-133.472900;;;MARS:J2000"> Olympus Mons, Mars: The highest mountain in the solar system. </span>
Olympus Mons, Mars: The highest mountain in the solar system.
Real world example
[http:// A crafted page with the examples appering here and a small map demonstration]
Pages for discussing this proposal: