Saturday, November 30, 2013

Leadership

At thanksgiving you always find out what the rest of the family is doing. This year I learned that one of my cousins was in the early stages of of a technology start-up. The fact that she is leading a start-up isn't really as surprise, but how she's doing it is a bit unusual for your average start-up traditionalist. She has a group of developers creating her application for free. Why are they doing this? First, because they think is 'cool' and second because they trust her. There are no contracts or lawyers. They this think this is a great idea, and that my cousin will take care of them if it works out. What a great way to manage lead a team.

Tuesday, November 26, 2013

The future of desktops

Everyone is talking about the future of desktops. If you read the headlines you would assume that 5 years from now everyone will be using a tablet, and we will never see a desktop again. I think that assumption is a bit silly.  We will still have to go to work, sit at desks, and type things into  computers, Its a heck of a lot easier to do that with a keyboard, monitor, and mouse rather than a touch screen. The real question is what will the desktop itself (the hardware) be doing. Most likely it will be a platform for delivering cloud based technology. Your main system (which may be MS Windows) will be on a virtual machine, your web/cloud bases apps will be in a browser. You desktop PC will be  primarily a portal. Since its not doing much, you won't need tons of hardware, nor will you want to spend a fortune on an operating system. This is a great opportunity for Linux. For those of you who enjoy irony, imagine an office full of Linux desktops running VDI clients with MS Windows as the guest O/S.

Saturday, November 16, 2013

Tools are Tools

We are constantly bombarded with new tools that offer the promise of better productivity, quality, profit...,etc; there's lean, agile, project management, TQM and many more.  While all of these tools can provide great benefits, frequently they are confused as the end rather than the means. No matter what tool(s) you use, they are only as useful as the people who use them and the plans they support. Great tools don't fix bad management, poor planning, or lack of vision (but they sure look good). Before you invest in a tool make sure you have your fundamentals (people and plans)  in place.  Remember, quality buildings were created before the invention of power tools, and great businesses were founded before books on  'modern management techniques' were published.

Saturday, November 9, 2013

My pet-peeve list of productivity sucking technology for developers.


  1. Eclipse - Why isn't this as easy as Netbeans ?.
  2. Windows Version Upgrades  - You had to move user directories and admin tools. ?
  3. Windows Server 2008. - How can you release a 'server' operating system that, for security reasons, can't run scheduled tasks out of the box ?
  4. Gnome3. - Really, a task bar wasn't good enough?
  5. Software without installs for major OSs. - I have to build my own stuff, why do I have to build yours ?




Friday, November 8, 2013

The Joy of a Big Grails Version Upgrade

I had the joy of upgrading from Grails 1.3.7 to 2.3.2 this morning (new computer). After three hours of upgrade fun I finally got everything to work.  For those of you who are making a big jump with grails here's what I had to do: 

1. Upgrade grails.
2. Attempt a compile.
3. Uncomment all of the maven update locations in buildconfig.groovy
3. Uninstall each plugin that failed.  Note, you don't use a script to uninstall the plugin, you need to remove the plugin entry from the application.properties file in the in your application's root directory.
4. Install the new plugin. Again no script, you create an entry in the buildconfig.groovy file. 
(By the way if you use the script to try to install/remove a plugin,  it does give you instructions about the new way to get them to work.)
5. Guess what, scaffolding is now a plugin, so you'll need to add it if you want any scaffolded controllers to work.

(Its good practice to run the clean script between each step to save you some greif.)

The good news is my other application only took 3 minutes to upgrade once I knew what I was doing.

I'm sorry that I don't have any examples of the changes that I made. I'm not near my work computer right now. I'll try and add some later.

Wednesday, November 6, 2013

Personal Rule-of-Thumb.

I frequently have to deal with requests from users.  Over time I've come up with a four part method of deciding if I should classify a request as  a 'project' and put it through the project management process, or  a 'task' that I can do without too much ceremony. After some creative brainstorming I even came up with an acronym. It's not a project if its SLIM

Short Time. It will take less than two hours.
Limited Parties. Not more than 2 other people involved.
No Process Impact. It won't change how an existing process  works (no training or user doc , etc)
Mostly Self Contained. It won't require a lot of interaction with other program and data flows, data in or out does not need to be changed.

What types of 'personal' rules-of-thumb do you have to organize your time ?

Sunday, November 3, 2013

5 Reasons Why The Ops Team Won't Load Updates

5. You changed the interface
4. You added a feature
3. You didn't test everything
2. The last time we loaded an update you didn't test anything
1. Loading an update will require missing the football game this weekend


See also: 5 Reasons Why The Dev Team Creates Updates

5 Reasons Why The Dev Team Creates Updates

5. You said there was a problem with the user interface
4. You said we needed a new feature
3. Something was broken and you wanted it fixed
2. You said you had to have it 'as soon as possible' .. or else
1. Writing code means missing a trip to the ballet


See also: 5 Reasons Why The Ops Team Won't Load Updates