This week I attended a formal training class on DSView 3 software. This is the primary application that my team develops and supports. It is also the application which has the User Interface that I support.
Now the DSView3 UI looks dated at best. However, updating and making changes to this UI is going to hold several challenges for me. But before I get into that, let me go into some background about the application. The company that I work for is mainly a hardware manufacter. We manufacture KVM (keyboard, video and mouse) switches, MergePoint Appliances (these group and control Service Processors), Power devices (SPC – Power Distribution Strip), and 3rd party products (sensor hubs, & auditors). DSView 3 is the application that allows the remote access and control of these appliances to be traditionally used in a lights-out datacenter.
Definition of a lights out data center as per Webopedia.com:
Also called a
lights out datacenter, a room that contains a number of
servers under lock and key and kept in the dark that under normal operation is not entered by human administrators, and all operations in the room are automated. The
computers in a lights out server room typically are controlled by the use of
KVM switches to help ensure the security of the locked room.
So the target user of DSView 3 software is someone who wants to access, manage and monitor target devices remotely. DSView 3 offers Hub & Spoke architecture which allows for redundance/disaster recovery, performance improvement, and load sharing. The different DSView 3 servers use replication to keep their information in synch across locations.
So some of the design challenges of DSView 3 stem from the fact that the functionality of the application is driven by the various hardware in which it communicates. Each piece of hardware comes with it’s own Onboard Web interface in which DSView offers centralized management. Many pieces of the application are written using plug-ins based on the hardware. This allows DSView to quickly add new functionality as new appliances are available, but it also means that it limits the flexibility of the UI.
Another design challenge that I face is that DSView 3 users have become used to the strange navigational structure of the application, so any large sweeping changes may be resisted. Sure the current UI may be difficult to learn, but many people have already done that, and it’s a lot easier than learning each individual Onboard Web Interface for each piece of hardware.
Due to the design challenges I mentioned, some analysis should probably be done to the current application to see which pieces can be made more intuitive without upsetting current users, or the business. I need to start gathering customer data to find out what the current users base likes about the application and what can be improved on. It was also suggested by Lane Halley from
Cooper that I create a Concept Model (as per Dan Brown) to help establish a new interface structure.
Yesterday I downloaded IHMC Cmap Tools which is software used to assist in the creation of Concept Models, but I still need to figure out Dan Brown’s process before I can impliment it.