Database Design Software

A group of software developers realized that engineers in academic disciplines and industries were using non-standardized diagrams and creating unnecessary confusion.  This group wanted to create a system where no matter what language you speak or which country you’re from, you can look at a standardized diagram and immediately understand its meaning.

These programmers merged their concurrent work into one standard method called the Unified Modeling Language 1.0  The point of this language was to have global leaders in software development agree on standardized practices that would apply to any willing participant. When you consider the globalization of the world economy–including the outsourcing boom–this is a very valuable tool.  In many ways, diagrams that help design databases and programs have aided globalization and global communication.

The UML standard has about 14 standard diagrams that are part of the current UML 2.0 version, but not all of them are used with the same frequency.  The most popular ones are, of course, the most useful ones: sequence diagram, activity diagrams, use case diagrams, and to a lesser extent, timing and ____________.

Activity diagrams show a flow of how a user interacts with the program and endows the flow with logical rules.  This allows use testers to see whether the program turned out as planned.

Sequence diagrams allow you to show the creation and structure of objects, e.g., when they’re active, what they’re doing, and how they interact with other objects.  True to their name, sequence diagrams display the sequence of events that execute code.  It isn’t uncommon to have multiple sequence and activity diagrams, which give a more granular vision of what the program is doing.

Use case diagrams are very simple and can demonstrate what tasks users might want a program to execute.  They basically provide a bird’s eye view of a program, without getting into the details of which buttons get clicked or which commands are needed.

Kinds of databases?

Different kind of databases are needed for different tasks.  There are two main types: ones that are operational and ones that measure statistics.  The first kind, operational databases, are also called online transaction processing, or OLTP, and are used in the operation of a process to allow data manipulation inside the database.  Analytic databases store static data that can be referenced to measure performance or to create dashboards.

Flat file vs relational databases

Think about the evolution of computers and how far we have come from vacuum tubes the size of person. Because technology is moving so fast, it has been hard to grasp where it was going. There is the famous quote by Thomas Watson, chairman of IBM, in 1943, ”I think there is a world market for maybe five computers.” As computers became less expensive, IBM was able to develop a mainframe computer that thousands of companies could afford, and the market for personal computers exploded soon after.  With the popularity of PCs, the demand for software became sky-high.  Anything from word processing, task management  data manipulation, basic calculation, computers were inextricably linked with work-time productivity.  Before long, every household and every company was using computers.

Today, if you develop a software program, you have a few more options than when the PC market first opened up.

Besides the cost of development, companies incurred the cost of printing hard copies, shipping, also pay reselletrs.  Advantageous because it gives users full access to software that they need without having to have fast internet connection, back when internet was slow and very costly.  You could have the same level of performance in a central location, installed on each PC.

As networks became more widely available and internet speeds became faster, the next mode was Software as a Service (SaaS).

-Don’t need to download/install anything
-available when you need it
-significantly cheaper than a downloaded product, because the distribution and overhead costs are so low and the market is so much larger

So getting back to network diagrams, now we live in a world where people deploy applications across a global stage.  INdividual users want an experience in the browser that’s on par with the experience of downloaded software.

Dealing with variable usage across the whole world.  Need to create a network with 1000s of servers that serve up the same experience to users in a geographic location.  Need contingency plans in case a server goes down, so that requires redundancy.  Need to plan for variability of usage–sometimes use is low, sometimes very high.  What you don’t want to do is pay a lot of money and have excess capacity…also don’t want your users left out in the cold if either no one can use it or a certain percentage aren’t getting a good user experience.

Some companies have the means and foresight (as well as luck and timing) of being able to buy excess server space.  Amazon web services is one of those companies–the product of Amazon planning for 4th quarter spike in shopping activity.

Amazon’s bread and butter is easy, convenient, reliable online shopping. Made most of their money in 4th quarter, wanted to make sure that server overloads wouldn’t interrupt shopping experience for user.s  Was latent for most of the year.  Thought to themselves, what can we do to offset the cost of these servers?  Starting selling space to other companies.  Now, Amazon and other companies are great third party providers of technology and stability that other companies need when they’re deploying software to the world.  Network diagrams get to the heart of that challenge.  Before you go out and spend a lot of money creating a network , you want to make sure your costs are justified and that you’re getting exactly what you need.