A random blog
Maps from Lost Blast door
Here are two links that have the map that was shown on the blast door in the last Lost episode. Map 1
and an Map 2
Of course, I have no idea what is going on still.
Enterprise Software pays off in decades not years
Everyone wants immediate ROI for everything they buy. Who doesn't want immediate gratification? I know I do, why shouldn't large corporations want it. Truth be told is Enterprise Software pays off over decades. Take online bill pay. How long did it take to deliver on that promise? I would guess at least 10 years, maybe more. Of course, now large banks enjoy significant savings and consumers have been freed from the misery of check writing. It took awhile but it paid off.
If the customer really wants immediate ROI they should do the following:
1. Immediately stop buying all new software.
2. Keep every process the same.
3. Stop delivering new stuff.
In the short-term they will save, if that is what they really want. Otherwise, except the fact that what is deployed today will payoff down the road. That is the reality.
Congrats, to George Mason for making the Final Four. It's been fun watching their tournament run. March Madness is simply awesome, so much excitement. I only wish we could have this for college football. Can you imagine filling out an 8-team NCAA Football bracket? Now, that would be madness!
Ridding the Pine
Assume for the moment you are Marissa Meyer the V.P. of Search Products and User Interface at Google. You have brought the world phenomenal applications such as Gmail and Google Search. What would use as your business email client? Probably Gmail? Maybe Thunderbird? No, you would use the ancient email program Pine:
"I use an e-mail application called Pine, a Linux-based utility I started using in college. It's a very simple text-based mailer in a crunchy little terminal window with Courier fonts. I do marathon e-mail catch-up sessions, sometimes on a Saturday or Sunday. I'll just sit down and do e-mail for ten to 14 hours straight."
I know Google believes in minimalist UI design but wow is this extreme!
With all these Web 2.0 companies launching, I am surprised no one has created a NCAA Bracket Pool Aggreator. This way I could do my picks once and then enter all the brackets on ESPN, CBS Sportsline, Yahoo, etc... What we really need is somekind of Gartner Magic Quadrant for NCAA Men's Basketball bracket sites. Then we would all just use the one that is in the top right that has both vision and the ability to execute.
No back to watching the games. So far so good, my final four is still alive.
Ranch 616 Restaurant Review
I tend to eat the same places over and over. I been trying to venture out more and eat some of the smaller local places here in Austin. This weekend I went to dinner with some friends at Ranch 616
. It was very good and I would recommend it. It is mostly southwestern style dishes. Some of the choices included: fish tacos, duck, and steaks.
Chip had one of the largest meals I have ever seen served. He had a rib-eye that came with a friend egg and a side of enchalados. He only finished half, which was still impressive. I had steak, mashed patotoes, and some shrimp. It was all very good. Check it out if you get the chance.
Session 3: Postgres
Next up. Infor about Postgres.
- Missed speakers name... Will post later
- New features in postgres
- Native windows support,
- point in-time recovery
- Two-phase commits
- In 8.1 all you have is roles. When you create a user you give them a role. No more unix style rights and permissions.
- Performance improvements include:
- In-memory index bitmap scan
- Improved butter management
- Vacuum cost delay
- Background writer
- Advanced features:
- Table partitioning
- Inheritance -- Take a parent table with common information. The child table then extend attributes that are specific to a specify application.
- Partitioning -- In 8.1 you can use subtables for breaking a big table into smaller based on things like date.
Austin Bar Camp Session 2: Marketing Cluetrain Style
- Tara Hunt
is giving a presentation on Cluetrain Marketing. She is going over what cluetrain is and how to apply it.
- Dude, Doc Searls
is in the house sitting two seats in front of me here at Austin Bar Camp. We are officially cool.
- Don't target the current thing or idea. Understand what needs will be in the community tomorrow.
- Reward your superfans. Gave examples about Apple and the "Spread Firefox" movement. Does not mean money, more about attention.
- Get involved in the community. Hire someone from the community.
- Be your own customer. What you buy it? Would you use it?
- When seeding a community be ready to back off. If your idea doesn't take then stop look at it. Pushing to hard my create enemies.
- Tools to use tagging, blogging, bookmarking, and social networking.
- Don't take yourself too seriously.
- Where to begin:
- Who to you server?
- What is your core strength?
- Wht is your core score? Number of users, number of photos (flickr)
- What action can you take today?
- Good prez. It was fun to watch Tara present Cluetrain in front of Doc. She seemed nervous and did a great job.
First session at BarCampAustin -- getting Shit and Getting Shit Done with Blogs
Presentation by our favorite Redmonk Analyst Michael Cote
- He is talking about to use blogs, wiki's, search appliance, etc within the Enterprise.
- Cote's requirements for "Enterprise Blogging Systems":
- Self registration
- Easy to setup and run: you'll do all the work
- Multi-platform (for scaling)
- His advice is to do all this yourself as a skunkworks internal project. Don't wait for IT, they won't like it at first...
- The system he used included: Roller, Lyceam, WordPress, and Hosted....
- Other stuff you will want to have behind the firewall: aggregators, search, wiki's, and RSS. You need to create your own blogsphere within the Enterprise. Recommends using Newsgator Outlook plugins for new tech people.
- Best bullet point of presentation: "But I got shit and got shit done."
- There was some discussion about Stock (Wiki's) vs. Flow (blogs).
- Lots of discussion about how to sell blogs in an enterprise.
- Excellent session. Lots of discussion. Could not blog it all since I needed to chime in.
Attending Bar Camp Austin
I am too cheap to get a pass to SXSW interactive
but Bar Camp Austin
is just the right price -- free. This is my first bar camp and I hoping it's good. I will try to post about the session throughout the day.
The photo stream is here.
Re: Cote on Enterprise Agile
Cote has a good post on Enterprise Agile
. All too often the consultants and academics of the world punt on providing advice on how to apply an idea to your work environment. It's great to see Cote taking this issue head on. That is how can enterprise software vendors use and apply agile. Here are some of my ideas building from Cote.Re:Scrums-of-Scrums needs work
This is all about fighting the fiefdom mentality. If you have multiple teams, then at some point infighting will start. Each team will think the other team is a bunch of idiots. If they are geographically separated, then this divide will start faster and grow deeper over time. To combat this you need to make sure at least some people are rotating amongst the various teams. This helps minimize the amount of animosity. Also, you have to find first-line managers who are willing to reassign resources without making it an emotional endurance test. If employee A has the skills to something on team 1, then let him move and encourage it. The single most important thing though is to make sure the teams understand they are working on one collective product. Instilling a common vision in everyone helps keep everyone on the same page
Re: Sales and Marketing
The truth about enterprise software is it's more like consulting then consumer software. A consultant engaged on a long-term project needs to cultivate a good client relationship. Most enterprise software revenue actually comes from maintenance
. So what should the software sales rep be selling? First, he has to sell only the features in the product. Now, in return for towing the line on features, he should be able to sell predictability and reliability. That is the software is going to release on time every time. Everyone knows software releases are late and sometimes moslty lacking. Agile fixes this if nothing else. You should be able to easily predict when your going to release. As far as presenting the road map to customers, limit your discussions to only the story cards done in the first few iterations. This way you have them in the bag by the time customers are hearing about them. Enterprise software is more like a marathon then a sprint. You will never have all the features done for all the customers. However, if you deliver on time, then you are way ahead of the crowd. More importantly, customers will start to appreciate this thus preserving your maintenance revenue.Re: Support:
Make support its own agile team. This group should be responsible for creating all hot fixes. All hot fixes should be rolled into quarterly service packs. Service packs should only contain bug fixes, no new features. Most importantly, make sure this team is staffed with people who like the firefighting that comes with the job. Support can't be a dumping ground for low performers. Think about it, support is one of the most important parts of the company. They are chartered with working with the customers at the most difficult times. If you don't staff it with good people, say goodbye to the maintenance revenue.Re: Upgrades
If you are delivering package software then it's impossible to give customer new releases more than a few times a year. Ideally, I think you have one major and one minor release every year with service packs every quarter. The upgrade rarely works and mostly fails because it has been tested in a controlled environment. The best option is to replicate a real customers environment and test the upgrade on that. If that is not possible, then find the largest customer you have. Chances are they were already planning on testing your software before it goes into their production environment. Commit yourself to working with them to test the upgrade. Both sides win here, your largest customer gets preferred treatment and you get to really test the upgrade.
Okay, that's it from me. I think this is a great topic and would be very interested in how others are doing Enterprise Agile.