Tuesday, April 22, 2008

Towards a computerized methodology for computerization

We build software solution that run on computers to help people better run their businesses. We use computers to build these software solutions. Yet, computers help us to better run or computer enablement business (it is a vicious circle). But i am excited about some advances that will help us manage building software solutions using computers.
During testing, we find bugs, when the system becomes big and the no of developers grows, it is not easy to follow up on bugs. A couple of year ago, we started using a software tool called Mantis. It is an open source bug tracking software than runs PHP, Apache and MySQL (The LAMP combination, you know). It is a very helpful tool, where developers are assigned tasks to solve bugs. Unsurprisingly, we also use software like MS Project to follow up our project plan. There are Open Source Project planning software like DotProject. Not very exciting so far. What is more exciting however is the fact that you can integrate Mantis with DotProject. This will effectively link bug fixing task with your project plan. In other words, once a bug is fed into Mantis and assigned to a developer. The project plan shall immediately reflect this as a new task in the project plan and shall also reflect this task in terms of time plan in your project plan (Gantt Chart)
You have no clue how much joy is brought to my soul whenever an attempt is made towards reaching a Chaos free software development endeavor.
Mantis is already implemented in our office. DotProject is not, neither is the integration between both. Any volunteers?


Aseel Hmood said...

Being one of the teams who used Mantis extensively. I would recommend any enhancements in this domain. Mantis is very handy and proved its usefulness. The merge, if works as you mentioned, will be good but I think this will cause a lot of shifts in the plan during testing beside the fact that plan will contain a lot of subtasks (CBJ had nearly 650 issue which means 650 task on plan) is this reasonable?

Amani said...

I also like any attempt to ease the managerial and administrative tasks during software projects life cycle, because I've been through this mess before, but as an industrial engineer, I believe that a human can't get benefits from tools like computers and software programs unless he himself knows what he wants and how he wants the tool to help him. This demands that we have a well defined administrative procedure where we know exactly what tasks to be computerized and what is expected from this computerization, then we can compare between the tools and decide.

I do agree with Aseel that a lot of plan shifts might occur, this might cause more headache. At last, It's a good idea to have a merge that ease our work, but we have to decide what we want first.

Getting Ref of the View Object referenced by the current Iterator binding for One iterator page without knowing the name of the iterator

Getting Ref of the View Object referenced by the current Iterator binding for One iterator page without knowing the name of the iterator ...