Tag: web

Vector web

here are a few data points

maybe we will finally see rich vector graphics, and new user interfaces that do not suck, and are accessible? also, a google search that could interpret vector graphics would be far superior to the fuzzy matches we get today one presumes. back in 1997, when dave was still a thought leader, he called for the lineto moveto web.

Beyond CMS

HYPERTEXT: from Text to Expertext outlines many of the concepts i have been thinking about in the evolution of cms. developing and greatly extending on the rather vague ideas of hypertext, it starts with small document collections and quickly goes over to massive collections.

To facilitate the handling of massive document collections, institutions collect and provide access to the documents. A Macrotext system may store 10s of millions of documents and principally supports search. After a document is found, browsing or reading is performed on paper or with a Microtext system. Human-human, as well as human-computer communication, must be supported by a Grouptext system rather than by a Macrotext system.

it then goes beyond content towards issues of metadata and expert knowledge.

The major problem with microtext, macrotext, and grouptext is that in dealing with text that contains many links people become disoriented. Expert guidance in the form of Expertext may lessen the disorientation. Instilling life into document systems, by adding procedures which will respond to different users in different ways, one has a greater chance of creating documents which will succeed in influencing others. The combination of expert system and hypertext system features increases the utility of each system.

wireless transaction servers

occasionally, there is still a gem to be found in the slashdot crowd. in response to the announcement of an apache port to windows ce, len produced this nugget:

The WinCE apache thing, but then it hit me. It isn’t about the general type of web server that we know and love. It’s about a portable transaction platform!

Imagine that you are a salesman, contractor or other individual who travels and does business with many different people in many different places. If you had a 802.11 (I don’t think bluetooth will make it) enabled Jornada or iPaq, then you could carry a web-based transaction server with you, wherever you go. You wouldn’t need a fixed IP, nor a constant broadband link to the internet.

You could hop on to your client’s wireless LAN, or establish a peer-to-peer link to make your transaction server (e-business whatever) appear local to your clients. They could do business with you, and you take your web-site and data with you when you leave. In this way, you could use a near universal interface (web browser) to handle customer interactions, without having to scrawl all of the information in by a flaky pen-based interface.

This would be a very cheap way of doing business, with less threat of being cracked by some script kiddie.

Web services metering

IBM makes a distinction between free web services with no service level agreements, and web services for business transactions. They envision a wide variety of business models for this, including

  • The pay-per-click/fee-for-use model
  • The subscription model
  • The lease model

their resource counter can create a wealth of information and output it in a new xml format for usage records developed by ipdr.org

obviously this needs to be integrated with uddi, and open api’s need to be developed so that this can be seamlessly integrated with the overall architecture. i hope it does not mean a worldwide roll out of a pki infrastructure, though. that would suck given the pki adoption track record. Their paper is at developerworks.