1.1 Will dasII support CYMK color model in postscript output?

Not trivially. We could support this by using a special RGB colortable that is converted to CMYK as an automatic post-process.

1.2 Where can we see demonstrations of dasII in its current state?

There are a few applications at http://www-pw.physics.uiowa.edu/das2/demo-apps.html. You will need to have java 1.4.x and webstart installed on your computer. Keep looking here periodically for new demos. We may also release some demos as applets, but presently we use webstart to constrain the platform.

1.3 How will dasII perform for very large datasets, like the Cluster wideband data?

Java is quite efficient nowadays, and it is often reported that it can outperform C++ under some circumstances because it can perform run-time optimizations. And since das2 is tightly integrated to the web, calculations that are too costly for Java to perform can be performed on the server by a more efficient means.

1.4 Will dasII write efficient, portable postscript output?

We have verified that the output is efficient and easily modified. We have not tested the postscript on a variety of platforms, but we have no reason to think there will be problems.

1.5 Will dasII support a "dynamic colorbar" that can be used to rapidly set optimal colorbar settings?

We use a 24-bit color model that prevents rescaling the colorbar at an animated level of interactivity, as you might have seen done in IDL under an 8-bit color model. The colorbar is interactively modifyable, and we've found that one can quickly select an optimal colorbar range using this method. We could explore supporting a 8-bit color mode in das2 if there is a desire for this ability.

2.0 Just what are levels of interactivity?

• animation-interactive: the operation can be performed very quickly by the system so that a sequence of the operations results in a fluid output. An example of this is the spectrogram slicer, which can be dragged across the spectrogram to quickly locate unusual features.
• interactive: feedback is immediate and the operator can make many modifications per minute. For example, after zooming on an axis, you will see the result within a half-second.
• sub-interactive: operations that should be interactive are not, which causes irritation and is a failure mode. Often X window applications running over slow link are sub-interactive. This is often used to describe failure mode of an interactive process, but when its unavoidable, some feedback should be given to indicate that condition is expected, such as an hourglass.
• patient-interactive: the operator is aware that he must wait for feedback. Interactive functions can be performed, as long as the operator has other things to do while waiting. Examples are X window apps over a phone line, debugging source code. These operations should be accompanied by a progress bar that indicates that something is happening and might give an idea of when the task might be completed.
• batch-interactive: operations are expensive enough that it is worthwhile to completely set the configuration or define a set of operations before they are performed. Faster modes of interactivity allow for a sequence of incremental changes to the configuration, but speed makes that prohibative in this case.
• batch: operations do not take place while the operator is waiting.

2.1 What's a Granny string?

The strings used in IDL for formatting equations, like E=mc!e2, where invented by Grandle and Nystrom. We call them Granny strings, and the strings you used in IDL will work in das2.

2.2 What's webstart?

Webstart is Sun Microsystems' solution for delivering Java applications over the web. A java applet is a program with limited access to the machine it runs on, so you can safely run programs found on the web. They are quite limited, so Sun devised a system where can run an application from a trusted party on your computer. When you download a java application with webstart, you will be shown the digital signature on the application, and asked if you trust the source. Once you okay the source, you can freely run the application on your computer. The application is stored on your computer for later use, and Webstart also automatically checks for updates each time you run it. Webstart is distributed with Sun's JRE 1.4.0 and up, but must be installed manually on Unix systems.

2.3 Can I store my plots for later use, and send them via email?

Plots and all the controls that affect the plot are stored together in a dasML file. This file is in ascii and is in a format that is quite tolerant of the whitespace changes that often when sending a file. Since das2 is tightly integrated to the web, we are also considering server-side storage of dasML files, so you can get to your plots on virtually any computer that has access to the web. We can also imbed the dasML within the .png files produced by das2, so that an image shared on the web can be modified in the future.

2.4 How about annotations?

We plan include annotation objects for the plots in the near future. These will include scalable text, arrows, and lines. Our goal is to provide output that is journal-ready needing no further modification.

3.0 What's a Datum?

A datum is the smallest bit of information that is passed around the system, and can be generally thought of as a double and units identifier. These are java objects that are communicated between parts of the system. For example, when you click a point on a spectrogram, it's not the pixel location that's communicated to the slicer, but the data coordinates as read off the X and Y axes. It became clear early on that there'd be great benefit in this as a temporal location can be handled with the same code that handles quantitative data. What's more this allows us to handle units conversions automatically, so that the point "800 km/s" is trivially located on an axis with units of "cm/s." Datums also allow us to accept annotative data along with quantitative data, so a dataset could consist of the tuples like (UT Time, Frequency, S/C Name).

3.1 What's a dataset identifier?

This string that looks a bit like a URL uniquely identifies a datasets. For example, the id "http://www-pw.physics.uiowa.edu/das/dasServer?voyager1/pws/sa-4s-pf" identifies the dataset voyager1/pws/sa-4s-pf served by the DasServer at http://www-pw.physics.uiowa.edu/das/dasServer. The id's can also refer to files on a local file system, or java codes that generate data sets.