Fog Creek Software
Discussion Board

Project Management - Microsoft PM 2000

Is there a good site to learn on Microsoft's Project Management 2000?

I have had many attempts to start the learning before - but it was just plain boring. I am not directly involved in project management, is there a site that goes deep into Project Management and also touches the lowly developer like me?

Wednesday, September 10, 2003

Heh, well, this isn't really going to help you, but I personally take the view that if you need a really complicated project management tool to track progress either your project is too complicated or you're just deluding yourself.

Seriously, MS PM is really only useful if you work on a traditional waterfall project. An iterative or, even worse, agile project, is just not even going to be worth the bother.

The really sinister thing about PM tools like that is, the more detail you put in the more work you have to do to maintain it. All this effort adds no value to your project.

Now, a really simple tool like a burndown chart (used in XP) is a great way to easily track progress without all the self-delusion. It's also immediately obvious to any viewer whether your project is on track or not.

Just my opinion.

Wednesday, September 10, 2003

I took a one day class through Horizons, and found it to be really useful.  They didn't try to turn me into a pro, but they did show me the basics of what it could do and how I might use it.

As to it's utility, MS Project is very good if you have to coordinate the work of several teams, where progress is non-linear.  We found it very useful for construction management, where several teams would be working at the same time and it was necessary to track overall progress towards the deadline.

Clay Dowling
Wednesday, September 10, 2003

What is used to keep track of the project in XP? I am guessing XP is better suited to smaller development teams.

What if one is working on a larger implementation?

Ram Dass
Wednesday, September 10, 2003

I guess the point in XP is NOT to track your project that closely. In most cases, detailed project management tools simply give you a sense of control that is illusory.

In our project (an agile project), we track in this way:

1) We have a list of stories (features) which our customer (product manager in our case) has selected to be included in our next release. These stories are estimated at a high level in ideal working days. The total for the stories is the total effort required to deliver the release.

2) Each iteration our customer selects stories to be implemented. We allow our developers to "sign up" for tasks. The amount they can sign up for is based roughly on what they've tended to accomplish in previous iterations.

We track the tasks in a simply twiki (Wiki based communal tool with a plug in for XP iterations).

Developers then update their progress in the twiki as tasks are completed.

3) Each iteration I update a "burndown" chart. The chart starts with the total ideal days from 1 and subtracts the amount for stories actually completed in each iteration.

After a couple of iterations you can extrapolate your likely end date based on your past work completed per iteration (called velocity). If that extrapolation says you're not going to make your date then you have to adjust scope.

Ok, a lot of people are going to yell and wail that this is voodoo or something. I maintain it gives you about as accurate a view of your projects progress as MS project. The plus here for us is that this tracking system takes us all of a total of 15 minutes or so per week to maintain.

Project management IS a lot of guesswork and witchcraft. It also adds absolutely NO value to your project or company. It's wasted effort. And since it's wasted effort you should try to do as little of it as is absolutely necessary.

Can you track a project with MS project? Of course. For our part, we've made our last four release dates, with no death march like behavior. It works for us.

You have brought up a good point about teams. Our team is roughly a dozen people. You have decide for yourself if that is large or not. We find that it works, but larger teams using XP in our company have had difficulties. We seem to be about the max size you'd want to run XP with. This may seem like a large limitation, but project tracking for teams larger than this is problematic anyway.

If I had a larger team and wanted to do XP I'd probably split the team into smaller sub-teams. If there's opportunity to do this in your project it should work relatively well. If the separation is clean they can even do iterations of differing length.

Even if I weren't doing XP the first thing I'd want to do is split the team down into more manageable sized groups anyway.

Again, just my opinion.

Wednesday, September 10, 2003

Don't be surprised at the number of Anti-MS Project folks you hear from on this forum. MS Project is based on 'classical' project management, and the prevailing sentiment is that it isn't helpful to software development.

Joel slams it himself:

Anyways, I have a tip for you. Think of MS Project as being part of the Office suite. Every one of the Office programs has a jillion features and can go into excruiciating details.

Does anyone use every feature of Excel to do their cash flow forecasts for their business? Nope. Too complicated, too much detail, and all you get is a seat-of-the-pants estimate accurate to twenty decimal places.

So when it comes to MS Project, don't be sucked into learning every feature to model every contingency. Learn the 10% of the features that will help you understand who has to do what by when. That's all you need.

Reginald Braithwaite-Lee
Thursday, September 11, 2003

I second the motion to attend a one day course.

I would stay away from the 1 hour a night over 3 months type courses. You lose steam after 2 days.

In one day, or indeed one afternoon, you learn enough about Project to allow you to begin using it, and by extension learn.

The only way to learn an app is by using it.... unless you figure out which of your little project nuances it does or does not support, no book or website will do you any good.

Thursday, September 11, 2003

I guess I'm "thirding" the motion to attend a course.

I took a two-day course on Project along with my manager and the rest of our team-leads. Most of us didn't want to be there - we were in the "give me the software and I'll figure it out" mindset. That's the wrong thing when starting with complex project management software (MS-Project, Superproject, ...) - there are just too many ways to get messed up.

A good course will give you an introduction to the easier aspects of the software and point you to the elements you might need as your projects become more complex. You might be able to do the same with a book, but a good instructor will be able to get you up to speed faster.

Thursday, September 11, 2003

"Learn the 10% of the features that will help you understand who has to do what by when. "

Yep. Tried Joel's method and kept thinking "I wish this thing knew about 8 hour days and weekends... wait MS Project does that..."

Some of the more complex stuff eluded me, and still does, and I'd probably benefit from a course in it if i really cared about it. Just read Joel's article and think of MS Project as a spreadsheet that knows about weekends and holidays.
Thursday, September 11, 2003

MS Project is not the enemy of a successful project.

While it might not be useful for tracking the day-to-day activities of your programmers, it is still useful in planning and organizing a project as a whole.

There's more to your project than just your lines of code....

Mark Hoffman
Thursday, September 11, 2003

*  Recent Topics

*  Fog Creek Home