Better Email Prioritization Going Mainstream?
from the maybe-possibly dept
For years, we’ve been hearing stories about how email overload was a problem. One potential solution that is often brought up is the idea of better prioritization of email. Rather than just a time-based list, shouldn’t our computers be smart enough to recognize which emails are likely to be more important? There have been a number of projects that try to do this, with varying degrees of success — but none have gone very far outside of small numbers of users. However, it appears that Microsoft is trying to get into the game, with a relatively simple prioritization system, based on separating out the emails of people you regularly email with, on the assumption that they’re more likely to be important than other emails. Again, this isn’t a new idea, but having Microsoft actually implement it suggests that it may finally break into the mainstream. The real question, though, is whether or not people have already become so conditioned to traditional email (including their own tricks for email management) that they’ll see this system as screwing things up more than helping.
Comments on “Better Email Prioritization Going Mainstream?”
Remote Action Interface - RAI
I wrote the following artice a couple of years back. Feedback is greatly appreciated, it relates nicely to “smart” email.
Remote Action Interface (RAI)?;?
::= action
::= {letter | digit}*.{letter | digit}*
As the Internet expands due to the addition of new systems, applications and users the need for efficient communication between all these entities increases. The Remote Action Interface Architecture (RAI) achieves loose coupling among internet entities. Any system which is network addressable is considered an entity. In order to simplify entity to entity communication there is a need to abstract away the complexity of the network. By eliminating the knowledge of the network from the implementation layer the RAI architecture is able to move data from one entity to another seamlessly. Implementing a solution built on the principles of Remote Action Interface (RAI) provides the ability for entities to self manage their own data without human intervention.
Within the RAI Architecture there are three classifications actions, entity and containers. The classifications are organized in a hierarchical tree structure. All leaf nodes are entities while all paths starting from the tree root (domain) are containers.
Actions:
The four basic actions which can be preformed on an entity are view, delete, update, and add. From these four actions additional actions can be derived. Data sent to entity is contained within xml tags.
Syntax: action@
Entity:
The entity defines what time of actions it will perform. The entity must reside within a container.
Syntax: action@entity
Container:
Containers hold zero or more entities. The classification changes between entity and container depending on the manner in which the entity or container is being referenced.
Syntax: action@entity.container
RAI Definition
A request is sent to discovery@entity.container the response is a RAI definition, which lists all entities and actions.
::=
{
[public | private] ?;?
[public | private]
}?;?
::= entity
::= {letter | digit}*.{letter | digit}*
RAI Architecture Implementation
RAI Server
Today the communication and programming are two distinct layers. The RAI server functions as a mail and web server, which allows fusion of the network and the implementation layers. The RAI server is not able to hold the requester?s state information; therefore security access verification is done on each request.
RAI Language Extension
In order to support the paradigm shift in computing architecture, programming languages must support embedded communication. Data within the application is passed to method using either the RAI request or conventional memory stack.
Current language construct
Object.method(parameter1, parameter2);
Future language construct
view@entity.com;
view@entity.container.com;
RAI Request
The first interaction directed to a entity is the RAI discovery request sent to discovery@entity.com. The response will contain the entity?s RAI Definition. The RAI definition is simple and universally available to all requesters. The requester is free to invoke any discovered action. The requester invokes an action by sending a request to action@entity.container. Data is sent to the RAI action in the format specified by the RAI definition.
RAI Response
The RAI response is the result of a RAI request made by a remote entity. The format of the RAI response is outlined in the RAI action definition.
Implementation Example
Wal-Mart purchases products from hundreds of vendor world wide. Many of Wal-Mart?s suppliers? change mailing addresses frequently. Every time a supplier changes its mailing address Wal-Mart must locate the supplier in their supplier database and manually update the address information. Having to manually find and update information costs Wal-Mart time and money. Costs are reduced by implementing a RAI solution.
Sample RAI Definition
entity FedEx.com
{
Entity address.FedEx.com; // entity declaration
public view; // action declaration
// input: none
// returns: website
};
entity address.FedEx.com
{
public view;
// input: none
// returns: company address information
// 43
// wall street
};
Fedex Source Code
Public view@address.Fedex.com
{
String streetNumber = ?7?;
String streetName; = ?Long Tail Drive?;
// load data in to entity
update@streetNumber.address.Fedex.com = streetNumber;
update@streetName.address.Fedex.com = streetName;
// from is global variable
// from variable can contains update@fedex.address.accounting.walmart.com
return from = address.Fedex.com;
}
Wal-Mart Source Code
Public update@fedex.address.account.walmart.com
{
// from is global variable
// from variable can contains view@address.fedex.com
String Number = streetNumber.from;
String Name = streetName.from;
Return;
}
No Subject Given
How many microseconds will elapse between the MS implementation of mail prioritization, and the discovery of an exploit by professional spammers?
Re: No Subject Given
I have G-mail and don’t seem to have any major troubles at all with spam making it into my inbox. I may lose a student loan to the trash box once in awhile but overall the types of prioritization already in place a doing a good job. this sounds as if it takes it to the next level and as I said here “Always best to support a smaller technology group who seems to be ahead of their time by a step or two.”
priori
Email is already a means of information management – it doesn’t need to be laden with more granularity. There’s already an immediacy hierarchy in place: most important = phone call. Less important = IM. Not very important = email. Next thing, “prioritization” will creep to cell phones…the presumable “important” call will get a louder ring!?? Feh. Gmail, tags, alerts. Done.
ps – this was kind of stream of consciousness. Usually i take the time to form a more coherent repost and of course, bash on Techdirt a bit ;).