Links
·
RSS Feed
Popular Pages
 |
 |
.. it is a really special feeling when you give someone a book and it changes the way they think and act. So I'm really pleased to have just finished reading a book that I know I'll be handing out ...
-Jeffery Fredrick |
 |
I was amazed that these five chapters only take about 160 pages and yet tell you all you need to know about successful projects. I’ve experienced a lot of these problems myself, and so did/do you, ...
-Javaddicts.net |
 |
If your shop has trouble shipping quality software on time -- and let's face it, most do -- then this book is for you. If you're a manager, I'd say that doubly so.
-Ernest Friedman-Hill "JavaRanch Sheriff" |


|
 |
 |
|
(Aug 19)
Last week I was talking with a friend about a common ailment on development teams today. And it seems to be getting worse.
Perhaps you've seen it already in your shop. Once one person catches an STD, it seems to spread quickly.
STD, of course, stands for Shiny Things Development.
Oh cool! Check that out... it's new and cool. Let's include it in the product! Why? Umm... it solves some problem. And didn't I just say it's shiny and new?
How many shops have you met that have insane development infrastructures "just in case" things get crazy? Anytime the shop has a list of tools and libraries where every single one requires a specific version for anything to work, someone there has STD.
What's the problem with STD? It generally indicates a lack of discretion and promiscuous use of technology. Rather than saving yourself for something that actually works, you're chasing down every new product and technology. Sure, it might be fun to try out new stuff, but over time the fun is far outweighed by the problems. It's never fun in hindsight.
If your product stack is just as complicated to understand as your product is, you're infected. If team members can't get parts of the product working locally, you've got it too. Have you every patched a single toolkit locally? That's okay. When you've patched more than a dozen, it's not the rest of the world, it's you. Have you rewritten large parts of a new toolkit, just to get it working? And it never occurred to you to just dump the new tech? It's just a matter of time until parts start falling off.
But there's still hope. Unlike some human STDs, software STDs can be cured. It takes time and discipline. You'll have to step back and eliminate some of the more trendy technologies from your product. A few doses of YAGNI usually clears things up nicely.
Here are a few guidelines for spotting trendy ideas:
- you've patched anything locally because it just doesn't work
- an update to a toolkit always breaks the application
- you spend more time isolating which area a bug is in than you do fixing the bug
- you have a list of tech for new developers to install that usually doesn't get them up and running
How will you know when you've eliminated the problem? When new team members are up and running in a day. When you spend time adding features and supporting the product instead of trying to figure out whether the latest bug is in LDAP, the database, the application.
When someone introduces you to a cool new toolkit and you can only see the potential problems it can cause in it's zero point one state, then you're on the road to recovery.
Category: Agile
(Jul 1)
The Agile It! Experience (aka Agile ITX) finished up this weekend and was a great success. The speakers were all top rate. The audience was educated and engaged. We had great interactions and great questions. I brought my daughter Hannah along and she walked the fine line between entertaining the audience without annoying them. Other than me bashing my head into the side of the swimming pool (ouch!) and having too many managers (and not enough management related talks), I can't think of a thing that wasn't stellar. Next time I'll bring goggles to the pool, and I can guaranty we'll have the management track topped out next time around.
Here are a few related links summing up the conference:
David Bock from a speaker's point of view: Agile IT Exchange Conference
And Mike Witter's attendee point of view: Agile IT Experience… 5 closing thoughts
Here's our expert panel discussion. I got to emcee a blue ribbon panel taking questions from attendees. It was late in the evening and I ended it about 8 pm, but it's good stuff.
Agile IT Experience 2008 Panel Discussion - Agile Marketing, Culture, Jokes
The entire conference went extremely well... and you'll be seeing more of them in the future.
Category: Agile
(Jun 30)
Jennette Mullaney was kind enough to attend my talk Continuous Integration, The Cornerstone of a Great Shop talk in Las Vegas. We spoke for a bit afterwards and she put it all together into a nice interview.
Continuous integration reduces bugs, increases productivity
Enjoy!
Category: Agile
(Jun 3)
Ken, one of NFJS's best speakers, speaks tonight at Agile RTP. He'll be giving his well-known Iteration Zero talk.
I may not be there (my wife's under the weather), but Ken's a great speaker. If you're in the RTP, NC area, I'd encourage you to come out.
Category: Agile
(May 6)
Just a quick reminder, if you're in the RTP, NC area, come join us tonight and hear Andy Hunt.
We're also raffling off a pass to the Research Triangle Software Symposium, the local incarnation of the popular NFJS Java conference tour.
More details are on the meetup page
Rumor has it that Andy might exercise the speaker's privilege to change the topic to the future of Agile. See you tonight.
Category: Agile
(May 1)
I frequently hear the question at conferences... sure, we all know we should write tests, but it's like exercise. I know I should, but I don't. It would take too much time.
Recently I've been telling people that it takes only 10% of their time to create automated tests. And then I tell them about this blog entry. I decided to link to it so it'd be easier to find.
Automated Testing -- It's the thought that counts
At 6th Sense Analytics, we collect process metrics, so this is just an example of the type of insight we can pull out. After the tests were created, we knew exactly how much time was required.
Category: Agile
(Apr 30)
I wanted to let everyone know that we'll be giving away a free pass to Research Triangle Software Symposium. The very popular No Fluff Just Stuff traveling software conference will be in your backyard if you're in North Carolina.
It's primarily a Java conference, but we've got a strong Agile track as well. And there are classes in other topics, like Scala, Technical Debt and more.
I'm a bit biased... I'm one of the speakers. :) But I really like the conference and the people it draws. It's a great chance to get a quick introduction to a broad collection of topics and network with a great national and local crowd.
And if you don't win the free pass, you could always try writing a haiku about Atlassian. :)
Category: Agile
(Apr 22)
Last weekend I spoke at the Great Lakes Software Symposium in Chicago and an audience member verbalized a great tip about driving the team towards small code commits.
When you want a team a to add one feature or fix one bug, you use tools like continuous integration to train yourself (and your team) to check in a single feature. But how big are those features? If you've got features lasting weeks or months, then adding a single feature is a huge amount of code.
The key to small code commits is small estimates.
My definition of small is between half a day and a week's worth of work. Anything longer than a week probably has a surprise or two hiding in it. You haven't decomposed that feature down far enough and when you encounter this hidden surprise it might blow up in your face. Anything shorter than half a day is too small. And you can't do anything in half a day anyway... try it sometime and see the interruptions appear from out of thin air.
It's simply easier to estimate smaller chunks of work. It also minimizes the code collisions between you and other developers. It makes your integration tests more effective as well. If your integration test covers 150 classes, you need to use small code commits and continuous integration to isolate what code was broken.
So where are you? Still working on figuring out how to estimate? Start with small units of work and be sure to look back and see how you're performing over time. Fast feedback leads to fast fixes. Be sure to work in smaller amounts of work, commit often, and look back over what you've done as often as possible, whether it's your continuous integration system providing the feedback or the estimate errors.
But remember....
Smaller estimates lead to smaller code changes.
Category: Agile
(Apr 17)
Agile ITX will be in DC this June.
I was fortunate enough to be involved with the organization of this conference and I think it'll be a great one. The speaker list is out of this world.
The topics were selected to help both the developer and their managers. I'm constantly hearing from managers who just don't know how to manage or introduce Agile practices. In addition, we selected topics that hit the entire software development life cycle, not just the execution phase.
We're featuring three awesome books as a part of the conference. One of my favorites, Ship It! as well as Release It! and Manage It! Every attendee will get a copy of all three books.
Category: Agile
(Feb 25)
Clarke Ching has conducted email interviews of a number of personalities in the agile space and he's publishing them a few at a time. So far he's got interviews from Johanna Rothman, Jim Shore, Shane Warden, and me.
My page is here.
Enjoy!
Category: Agile
|
|
|
|
|
|
|
 |
|
Back in Boston! I'll be speaking on Saturday and Sunday.
|
 |
 |
|
 |
|
Back on the west coast. I'll speak (and keynote) on Friday and Saturday in Redmond.
|
 |
 |
|
 |
|
Any conference in Orlando is good, but this one is great!
|
 |
 |
|
 |
|
The first Agile Experience was a huge success. Tell your manager about this one.
|
 |
|