Helium is an email navigation client. Use it to explore your email.
You cannot harm your email by using Helium so an appropriate strategy for learning how to use it is simply to play around and see what happens. If you'd like a little more guidance than that, the following provides an overview of how to configure settings, view mail, perform searches and manipulate the environment.
But first we start with a description of what you see when the application starts up.
The instructions below are not an exhaustive description of how to use Helium. If you would like more information please see the FAQ, the developer's documentation and check out the mailing list.
Take a look at the Helium interface. If you don't have one handy try this screenshot. The interface is divided up into five internal windows, shown and explained below.
This is the window where you enter configuration information.
In this window, enter search strings for text searches.
Press this button to compose a new message from scratch. A composition window will be displayed.
This window lists all the messages currently in the system and provides a mechanism for sorting the messages (click on the table headers) and displaying an individual message (double click on a message).
This window provides a visual overview of the mail messages, shows clusters of messages (when that feature is available) and allows message viewing with a double-click.
The shutdown window exists outside the primary Helium window. If you click on this window the system will shutdown and exit.
Configuring Helium to retrieve and send email is a fairly straigthforward process, especially if you get all your mail from one place. If you get it from more than one it is slightly more complicated.
When Helium first starts, the upper left of the window contains the configuration window with seven fields. They are:
When these fields are filled with the correct information and the okay button is pressed, Helium will retrieve mail from the remote server. Any mail that is both on the remote server and already in Helium will not be transfered. As mail is retrieved summary information will be displayed in the list view and a dot representing the message will shown in the visualization display. If nothing happens check the configuration information to be sure it is correct. Also check the Web Start console for errors. See the FAQ for more info on the console.
Once mail retrieval is complete you may retrieve mail from another location by changing the information in the configuration window and pressing the okay button again. To retrieve new received mail from the current configuration, press the okay button again.
Pressing the help button will lead to this page.
Helium's persistence system will save your configuration settings between sessions. The next time you start the system any new mail will be retrieved automatically.
Individual messages may be viewed with one of two methods:
When viewing a single message, in addition to reading the message, you have three options. You may:
Only some message formats may be viewed. At this time plain text messages and HTML messages that are only HTML (not part of a multipart MIME message) may be viewed.
Helium provides very basic support for visualizing clusters of potentially similar messages. Once messages have been loaded the visualization window can be told to perform calculations to determine similarity amongst messages.
The calculations can be very memory intensive so there is a preferred process to follow:
Do not put too much stock in the clusters that are generated. The analysis being performed is quite experimental. At this time latent semantic analysis is being used. For more information please see the Latent Semantic Analysis site at Colorado University for more information.
Helium provides two types of searches: text searches and mail header metadata searches. Other types of searches are in the works but have not yet been implemented.
Each type of search returns its results in a new window that looks the same as and provides the same features as the list view. Results may be sorted by clicking on the table headers and single messages may be viewed by double-clicking on a row.
Underneath the hood, searches are performed using Constraints. Contraints provide a very flexible tool for retrieving items in Helium. If you would like to learn more about Constraints, see the developer documentation.
Text searches are performed by using the search window. When text is entered in the text field a case sensitive search of the email message bodies and headers is performed. So, for example, a search for "helium" and "Helium" will return different results. At this time searches are very simplistic: matches are made on exactly what is provided. A search on "hel" will return messages containing both "hello" and "helium".
Header metadata searches are performed while viewing an individual message. A double click on address field (From, To, Cc) will return a list of all messages that are from, to, or cc (respectively the address that matches. A double click on a subject field will return a list of all messages that have the identical subject.
As with text searches the address and subject metadata searches search for exact matches on the entire header value. This can lead to some less than perfect results since subjects change slightly as replies are added to a thread and many mailers present from addresses differently.
Double-clicking on a date header will return all messages that were sent between 12 hours before this message and 12 hours after this message.
Helium provides a mechanism for saving the layout of the graphical user interface. Any internal window that is used in the system may have its position and size adjusted and then saved. The next time the system is used the saved position and size will used to place the window. This holds true for the five primary windows shown above as well as any windows which are created on the fly, such as search results views, messages views, and message composition windows.
To use this feature, move around internal windows to something you like, and then press the "Save Interface" button.
The primary problem that you might find with Helium is that it can be a bit slow. The flexibility provided by KnownSpace and Java comes with some cost in performance. Helium requires a fast processor and a large amount of memory. Under certain circumstances, especially when using the semantic clustering features, there's a chance the system will either run out of memory, or spend so much time attempting to acquire memory that it will be completely unusable. This is a problem that is being addressed in future releases.
When the LSA clustering system is used with a large number of Entities it can cause an out of memory error. The system will continue working but the clustering system will not produced any results.
