Launching really old code written in 2006

1 points by vaibhav228 10y ago ↗ HN

We have a project written in 2006 with old tech stack like Struts1. It has lot of frameworks written based on hash maps which can be replaced with good frameworks available. For example they have written workflows frameworks for validating a use case by executing different struts actions.

For some reason management did not gave launch signal at that time in 2006 and project was closed.

Now we are starting that project again and the plan is to use the same code base and migrate to struts 2.3.24.

They have written different modules as separate projects and creating jars and referring them which is good, but they have cyclic dependencies as well, which is my concern.

I have gone through code and the code is not good. Not written properly and lots of performance issues. All UI related written in plain javascript, no jQuery.

The old code does not work and not getting up, because some part of the code is missing. No documentation. No old team member.

Everyone is new. New team is very much frustrated with working on such code and delivery schedule is very tight.

I think even if we migrate and work out that code and get it up and running. The maintenance of that code will be huge task and will be difficult to add new features.

We already had discussion with client And looks like he does not want to use new technologies or re design it.

How should I convince my client to write using new technologies and re architect the solution. We can use some part of code they have written.

How should I come up with new solution which he will like and can at least think about doing a POC to demonstrate.

Please anyone faced with such challenges. How you guys convinced your management or client?

3 comments

[ 3.4 ms ] story [ 22.1 ms ] thread

qubex 10y ago ↗

It all depends on the benefit your client expects to get out of this project. What you are proposing will increase the ‘cost’ associated with this project. Your client, if rational, will seek the best cost/benefit outcome. If raising the cost makes the net benefit vanish, they won't do it.

vaibhav228 10y ago ↗

Thanks for your reply. I agree that it will increase a cost little. But maintaining the code will have more cost.

Many smaller businesses have little understanding of concepts such as total cost of ownership and maintenance over the lifetime of a product or system.