As a junior programmer, how do I quickly make changes to large unfamiliar software? [duplicate]

Question

Possible Duplicate:
Tips or techniques to use when you don’t know how to code something?

I recently graduated on a Computing course, and luckily got a job as a junior programmer. Now, I have been given a series of tasks, from a bigger project.

My company has (supposedly) bought a revenue/sales software written in PHP. Basically this whole big piece of software needs to be modified to meet my country standards.

I have been given a large number of tasks to do in a relatively small time span. Now, this might be my very own understanding of the whole timing problem, but this seems a very big thing to do because I need to fully understand:

1. the whole system structure
2. the "style" in which it has been written, classes, functions, etc.
3. the database structure

Considering the above, I need to add some functionality, some of which can be considered "core" ones which, in my opinion, is a too long of a job to be done in 6 days.

I understand that I have given too little information for anyone to understand the level of difficulty of said tasks, since for some people might be impossible while for others might be a breeze, so I want to ask the following:

To the people who went past a similar situation, what was your outcome? Can I start considering myself unprepared or maybe they are asking too much? What was your experience on a similar situation?

Answer 1

You do not need to know the entire system to make changes. In fact, most systems are too large for one person to fully understand. I suggest you do your best to understand the impact of your changes -- your changes will only impact a small slice of the system. Creating tests will help ensure the correct functionality.

Does it matter if you are ready for this project? Your company thinks you are. I assume if you are a junior programmer that there are people whom can help you out. I'd go to them, I doubt they put you on the project with the expectation that you'd complete the changes without guidance. If that was the case they'd put a more experienced developer on the project.

Answer 2

Can I start considering myself unprepared or maybe they are asking too much? What was your experience on a similar situation?

Stop worrying, and just prioritize the tasks that you are currently assigned to do. Do the best out of the time that you are given by implementing changes and analyse its potential impact on the current system. Thus, having set of testing strategy before pushing your change-set to production is the very important part to Keep in mind.

The 6 days time-frame/deadline might be set just arbitrarily to see your response/potential, in terms of your effort to tackle the tasks. My advice would be - just focus on your tasks and do higher priority tasks as much as possible.

Answer 3

I'd second ElYusubov's suggestion of just prioritize and do the work. There are a lot of unknowns and this only gets resolved by diving in and getting things done.

In my first job out of university, I spent a few weeks getting my first task done which required the assistance of a couple co-workers to navigate me through the work to get the task completed. A key point here is to have communication with whoever assigned this tasks so that a realistic timeline can be made. You have to communicate how long you think things will take and possible explain where you came up with that estimate in cases with a wide difference between what someone may think it'll take and what it actually takes.

If the work has been estimated by a fellow developer, then ask for a list of the pieces to do that may be a different breakdown than what you thought is necessary. In a way, this makes me think of how to get from A to B that likely has more than a few answers. Some people will see how to make a straight line from one to another and get there in the fastest way and others may take an unconventional yet possibly more worthwhile route to get the same job done.

Answer 4

I've been giving deadlines which were created based on when business people wanted/needed something done: maybe it's a big trade show and the sales guys need a new feature or you're working on some integrations and a different dev team needs something from you so they can start their coding on time.

You have six days to work on these items, I would spend the first two and see how much you get done. If it still seems like too much let your boss know that you're going to work as hard as possible but will probably only get X% of the requirements complete. They might be able to help prioritize so that the most important tasks are finished on time.

You also may be able to bring in some extra help if it's 100% necessary that everything be finished in six days.

Answer 5

Ask for documentation. Not necessarily the project has the actual analytic docs, but it should have at least user and administrative ones. From this find what components you should change, than go down to classes and functions. Make one functional change, after that you can come with your guess for expected timeline to your boss.