There are other bibliographic formats which could be deployed in this role, e.g. one of the numerous variants of the MAchine Readable Catalogue (MARC) record[42,43,44], or the Text Encoding Initiative (TEI) SGML DTD[45]. There is some evidence that these technologies are inappropriate for use as URCs - they are complex to implement and require significant computing resources in their own right. By contrast the formats which are being developed under the auspices of the IETF require very little computing power to process, and may be readily understood by human beings.
Specifically, MARC has the problem of multiple national (and vendor extended) formats. This situation is made more complex by the use of variant fields and sub-fields. Finally, MARC's use of numeric tags to identify field types means that MARC records are effectively useless to any software which does not know the meaning of each field ID it encounters.
This is particularly a problem for networked information retrieval, where it should be trivial to query any database on the network. The plethora of variants of MARC - often using the same tags to mean different things - makes this extremely impractical. This situation does not arise with the whois++ based URC scenario discussed above, as this can process the field names programatically if it chooses to, but is always free to return the contents of the template as plain text to the end user. There would be little point in returning raw MARC records to the end user. Clearly, the trivial URC syntax is immune from this problem, but whether it probides enough meta-information is a matter for debate.
The use of TEI headers in the URC role might appear attractive, particularly in the light of the World-Wide Web, an SGML based application which has enjoyed enormous success in recent years. Unfortunately there is a fundamental contradiction here - WWW uses a minimal HyperText Markup Language, HTML, which is coincidentally expressable as an SGML DTD. The TEI DTD is large and complex, and would require either hard coding in an application a la HTML, or a full SGML parser implementation such as sgmls.
It should be noted that in contrast to the whois++ and minimal URC scenarios discussed above, both TEI and MARC offer the opportunity to be highly precise about the information being represented - e.g. in specifying authors' names, dates, and so on. The SGML URC scenario also does this, but in a way which does imply the same overheads.