MHTML to Enable E-mailing of Functioning Web Pages
With a recommendation from the World Wide Web Consortium (www.w3.org), XML is on its way toward becoming the next-generation Internet markup language. But the conversion from HTML to XML will be a slow one, and, in the meantime, HTML is still evolving into a more useful language.
The latest amendment to HTML is the MIME Encapsulation of Aggregate HTML Documents, or MHTML. MHTML enables users to attach to e-mail messages not just a captured Web page, but a fully functional Web site. "MHTML will be used to collect archives of snapshots of transitory Web pages, for one thing, and that is what it does to send Web pages," says Einar Stefferud, chair of the MHTML Working Group, a division of the Internet Engineering Task Force charged with developing MHTML.
Currently, single Web pages can be attached to e-mail documents but only as screen captures, and the receiver can open the attachment but not surf through other pages within a site without Internet access. The sent Web page templates act only as pointers, but an Internet connection is needed to actually access the information.
With MHTML, however, even users not connected to the Internet can navigate throughout an entire site, just as if they had access. The MHTML attachments literally contain the Web site content, scripts and components. However, without an Internet connection, the receiver of an MHTML attachment is limited to only the Web content included as an attachment. So in this sense, it works similarly to on offline browser, only the user doesn't need to connect and download prior to being able to browse without a connection.
MHTML feasibly could combine messaging and browsing, if the proposed standard is accepted and deployed. However, there are minor complications with browsing through MTHML attachments. "In most cases, there are no problems at all," explains Jacob Palme, an application area director of the MHTML Working Group. "There can be problems in special cases, for example, a Web site where documents are retrieved by searches through a search engine, or for Web documents containing Java code that dynamically creates URLs when executed."
Although MHTML can be extremely useful in the case of presentations and telephone briefings, the real value will not be in sending messages to one or two people. "The more readers there are to a message, the more value there is in spending much work on making the message look neat and readable. So one can expect MHTML to be used for messages with many recipients, such as electronic newspapers, magazines and journals," says Palme.
MHTML is designed for use with the current Internet markup language, HTML. But MHTML is not now compatible with XML. "It should be trivial to draft a new RFC that does for XML what MHTML does for HTML. This is mentioned in the MHTML standard. It would use Multipart/Related in much the same way as does MHTML," says chair Stefferud.
While MHTML may be symphonic to end users' ears, the notion of downloading 2-GB files sent in the new format is likely to ring concern in the ears of IT managers. Unfortunately, the proposed draft of MHTML does nothing to remedy the length of time spent downloading large files. But, says Palme, adjustments can be made to shorten download time. "One might modify the IMAP and NNTP protocols, so that the recipient can choose to get only the plain text part of a multipart/alternative message."
Microsoft Corp. and Lotus Development Corp. each plan to instill this technology into their messaging products and browsers. Microsoft already has implemented early versions of MHTML into Outlook 98 and Internet Explorer 4.0. Meanwhile, Lotus plans to include support for the new language in the upcoming version of the Notes client.