The Imperative of Open Source Apps

An IT application cost consists of two components. There are the licensing fees associated with a product and there are the service costs associated with customisation of the product and installation, training and support costs.

Before proceeding further, let me relate a story. A new marketing head had asked me for help in getting region-wise, product-wise sales. It did not seem to be an issue till we looked into the data. Finance had simplified their task by entering only the summary value of the invoices. The sale value for each customer was known but not what he bought.

Finance department was only concerned with the payments due from a customer and had not bothered to use the application the way it was intended.

I was helpless and the marketing head was furious and very unhappy. IT department was useless as far as he was concerned! Sales order should have been entered into the same system. One should not have to rely upon accounts to enter the details.

The key point is that it is important to have data in place. No amount of money spent on an application is going to give meaningful reports without reliable data. Process is important and that leads to the need for integrated ERP-like applications.

Open Source ERP

Any organisation can save the costs of license fees by using an Open Source solution. The benefit of the saving will be far more dramatic for small and mid-sized firms. Furthermore, the availability of Internet at a reasonable cost with high enough speeds has made it possible for offices or distributors in multiple locations to use a centralised application. An application with a web interface is ideal. However, desktop applications are also possible using a VPN. There are a number of Open Source solutions available, including;

How does one choose? Since the licensing costs are not there, the organisation can evaluate the customisation costs. Ensuring that the application data flow and the organisation's processes tally, may require customisation of the application or a change in the processes of the organisation. Hence, you can evaluate the available solution and choosing the one which best fits the organisation.

You need to keep in mind that the cost of customisation depends on the changes needed to the application and not on the number of people using it. Excessive customisation costs clearly put the mid-sized companies at a major disadvantage compared to larger organisations.

On the other hand, the cost of changing internal processes will clearly be to the advantage of the smaller organizations.

Far too often, there is a tendency to focus on the reports and their formats; whereas, the critical component is whether the required data is available and being captured correctly at the point of origin.

Open Source applications hide nothing. This is a huge advantage. The database models and information about how to create the reports is available. Any missing report can be added, even by an external application accessing the same database. Correct data can be mined for information. No matter how good a mining application is, it can never find information in the absence of correct data.

Additional requirements

In the old days, if the requirements have increased, an application would be enhanced or a new one developed. There may even be a downtime, while the upgrade or replacement was in progress. However, especially in the web-world, there is no scope for downtime. Gmail is not available for a few hours to some people and it is international news! If Amazon would go down for a few hours, it would loose millions. Even though a mid-sized company may not need such hard uptime conditions, there is no reason not to use the same techniques as used by the large companies.

Additional requirements or changed requirements should not require a change in the core application. These changes can be incorporated in additional applications or modules for the existing application. A user looking on a web site cannot tell whether the pages are from a single application or from multiple apps. He does not even care as long as the user interface is consistent.

There are wonderful frameworks available in Open Source to enable an organisation to add applications with speed and agility.

A few well known examples are:

Open source specialised reporting and business intelligence tools include:

No short-cuts

Often, simpler applications will allow short-cuts to processes or ability to 'fix' problems caused by not using the required process. This is a recipe for disaster.

A far better approach is to see if one can adapt the organisation to the practices implemented within the application system. The implemented practices are often among the best practices followed across the world and, hence, worth considering.

Not taking a decision is also a decision. The risk of choosing a wrong application is fairly low with Open Source. The cost of getting started is very-very low. The cost of services is manageable. No single long-term contract is mandatory.

An example comes to mind. We were working with PostgreSQL database and found that one application needed an extra index than was supported by PostgreSQL. We posed a query and one of the developers responded and informed us that it was just a soft limit in one of the configuration files and pointed us to how to change it. We had never expected to make changes to the database, but using the above guidance, an software engineer fresh out of college could build the updated version from the source code and we could overcome the constraint. All this within a day or two.

As cannot be emphasised enough, there is no secret sauce which makes an Open Source application work. Typically, Open Source developer worry a lot about security. It is normally an integral part of the system and the default settings out of the box are secure.

In case a wrong service provider is chosen, one can switch to another. The knowledge required is openly available. Should a wrong application have been chosen, the data from one Open Source application can be transferred to another should such a need arise.

The presence of exit options usually ensures that they are not needed but act like an insurance policy which, at least, honest folks never hope to encash.

Try an Open Source App. You will not regret it.