I tweeted this, but thought I'd put it here too.
The JRuby guys have been busy. They released JRuby 1.3.3.
We upgraded to it from 1.1.1. Our app runs JRuby as does our build.
Check out the performance difference in the last few builds!
Mike Herrick's Blog.
I tweeted this, but thought I'd put it here too.
The JRuby guys have been busy. They released JRuby 1.3.3.
We upgraded to it from 1.1.1. Our app runs JRuby as does our build.
Check out the performance difference in the last few builds!
Check out the slides Fighting Disease with PostgreSQL Full Text Search and JRuby on Rails
OSCON is here!
It may be the most epic OSCON ever. OSCON + Collaborative Software Initiative (CSI) co. meeting with open source luminaries in attendance + CSI nerd night - lots of stuff to do this week.
Two things happened:
It's nice to have a head of steam!
Stay tuned - its going to be a newsworthy summer.
And now I will partake in a hike in Forest Park because I'm spent.
See Core Developer Job Description.
See InfoQ Article on Tech we are using.
Update 25-JUN Note, we filled this position yesterday. We found an ideal person right in the JRuby community - JRuby experience, testing passion, and years of medical software dev.
Here is the intro:
JRuby and JRuby on Rails continue to being adopted by many projects. After Mingle, Oracle Mix or Sun's rewrite of mediacast.sun.com, a new project using JRuby on Rails has surfaced.
From the press release Collaborative Software Initiative (CSI), the company that brings like-minded organizations together to work on collaborative software at a fraction of the cost, today announced the release of the first open source, web-based infectious disease reporting and management system.
The disease reporting and management system, which is being piloted in Utah, will be adaptable in all 50 states and available under an open source license later this year. It is designed to support local health departments in the early detection and investigation of individual cases and local clusters of communicable diseases, while simultaneously meeting the state and federal needs of outbreak control, disease surveillance and epidemiologic research.
One thing to note further down in the press release: the project uses JRuby. We talked to Mike Herrick, Program Manager at the Collaborative Software Initiative, to talk about the experience of using JRuby for the project.
How often do you get to say this about the impact of your project?
When deployed, UT-NEDSS project will directly contribute to the prevention of sickness and death by effectively collecting, identifying, tracking and trending information gathered about infectious diseases and bioterrorism attacks. Additionally, this unique partnership is providing lessons for public health informatics on:
We have a great team of folks working their hearts out on this. It truly is an honor to be part of the team.
I love moments like this - those strange joyous moments of clarity where anything seems possible in software.
I picked up a new commuter bike today to get me to work on my fairly incredible commute I used to call river walk. Its "river ride" now I suppose.
I got a 3 speed, pedal break simple old school euro bike. It's an Electra Amsterdam Classic.
I love my other bike, but it isn't practical for commuting.
Looking forward to a groovy summer on my new granny bike!
My techniques are not magic, but they seem to be quite different than how I see lots of managers/companies operating.
If you have determined that you truly need to hire someone, that generally means that your team is hurting - it needs resources that it doesn't have - at some level you are already failing. Your #1 mission as a Lean manager is to protect the team and to get it the stuff it needs to be successful. Once you are in this position, barring a crisis, you have no task that is more important than filling the position with the most qualified person you can find as quickly as possible.
So my co-worker Ed termed what I do yesterday, "the blitz". Here is generally how it works:
I think the only major difference compared to typical hiring practices is in how I do this. I go all in. It's a blitz. When I'm in this mode, I generally do it 8-12 hours a day until I'm done. I can generally wrap up hiring someone in <= 2 weeks from start to finish. And the results generally speak for themselves. This technique works great for me as the "hiring manager" and for job candidates.
Hiring people can be quick and painless or can drag on and on for weeks and months if you don't make it priority #1. While you have unfilled positions, you are continuing to fail (at some level). Your team is continuing to feel pain. Distraction grows and grows. Before you know it, your project is failing.
Anyway, this way works for me. Your mileage may vary.
Now next week I can get back to job #2 of a Lean manager: empowering the team/sharing leadership, staying out of the way, & adding real value off the critical path.
19-APR 9:30 PT: Updated with links to Lean SW Dev.
A little bit about you:
19-APR Update: this position has been filled via the blitz. Don't worry, we'll be back.
We are still learning as we grow on the health care related project I am focused on. Our Core Team has had some success. The work load and what the team is capable of are both growing. We are becoming advanced enough together to take on more work and more people bit by bit.
We are currently experimenting with adding "Contributors" to the project. For any number of reasons, we are not able to add these people to the Core Team. But we have certain tasks that are well suited for part time contributors. It works well for us and the contributor. In some cases, the contributor just has an itch they want to scratch (e.g., want to learn Ruby/JRuby). In others we have a need we *need* scratched and we pay them to do the work.
I realize none of this is earth shattering - its common with open source projects. I am just saying that its exciting to be a part of it and see it working.
I am of course biased, but I continue to think that the CSI model is game changing. It certainly doesn't apply to everything, but it applies to an awful lot. Time will tell.
We shipped our first release of an application Monday with the pre-release.
The JRuby guys are great. Its becoming a great little community. I'm happy to be a small part of it. There seem to be big plans for the future after 1.1.
After working with it for the past 3 months in anger, I continue to believe that its a very compelling platform.
Pete and I worked on our deployment script last night. In the early days of our project, we thought we'd use Capistrano. Make no mistake, it is a great tool (I took it for a spin).
But when you are using JRuby, its a bit overkill.
I'm fairly new to the Ruby world, but I know the other deployment options for Ruby. They aren't awful, but they take a bit of work & have some moving parts to them.
A lot more work than handing over a .war file with a configured # of run-times to run within the JVM, a database script, and a runbook anyway. That in and of itself is compelling.
Oh yeah, you have to also run:
$apache-tomcat-6.0.14$ bin/startup.sh
(that's the runbook)
Anyway, our final iteration of our initial release is primarily focused on authentication, authorization, and entitlement.
I have a feeling we'll end up rolling our own, but there are a number of Rails plug-ins out there for this purpose that may have some useful bits. Also, a number of Ruby based rules engines.
So far, my real tired google search turned up:
I'll dig into these tomorrow. Let me know if you have had success with these or any other plug-ins/gems that may be of use for a fairly sophisticated user authorization/entitlement impl.I'll try to update this post with our findings.
19-MAR-2008 Update
We are still digging into this, but so far we haven't used any above given our situation, but Bill Katz Rails Authorization Plug-in looks somewhat promising so far. Seems easy enough to make work with SiteMinder in front of you.
27-MAR-2008 Update In case anyone stumbles upon this later, we ended up writing our own. Nothing against the other's we just didn't see enough compelling in them to take them on. Rolling our own seemed easiest.
In the meantime I have to say that I'm having a great time. Long hours to be sure, but the work is very rewarding.
For me, the most joyous part (other than working with exciting all FOSS technology, and being a member of a budding high performing team) is working with a business partner that truly seems to understand and embrace Lean Software Development. The proof will be in the eventual pudding, but its amazing what can be accomplished with a willing business partner that understands the fundamentals of software.
Or perhaps they just trust us, because we generally know what we are talking about. ;)
Or maybe we really are onto something and our business model is a better way to develop software.
Or maybe its all three :)
She used to make fun of me for blogging.
Now she makes fun of me for using twitter.
Anyway, you will quickly see that I married up if you read her blog. She talks about her more than full time job of being a mom.
I couldn't agree more with:
If you've made the transition from a hierarchical environment to an agile, self-organizing team, you know what I'm saying. You won't ever want to go back..
I've been called various things. My last title was "Architect". Before that "Technologist". And now "Program Manager". I used to be a pretty decent developer, but never a great one. I'm starting to get ok again. I'll be pretty good again soon. I've worked with a lot of great developers so I know the difference between good and great. I figured out somewhere along the way that I'm fairly good at sensing what teams need and sucking it up and doing the often crucial not-so-fun stuff. I like doing this stuff because it is rewarding knowing that I contributed to a team. As time allows I make technical contributions too (more and more now). There is nothing like being on a high performing team and carrying your own weight.
Seagull Architects are an abomination. I had a realization as I left my last job; for a variety of reasons I never want to have the title "Architect" again. In talking to other "Architects" and former "Architects" I was fairly surprised how common this sentiment is. I'd be happy with what I have now or "Developer". I also like Steve Vinoski's "Member of the Technical Staff". I don't really care what you call me, but I do care that I carry my own weight. More than anything I care that I'm on a whole team that is already or has the potential to become a high performing team.
Let's just all just stop using the term "Architect" in software. How about we get rid of "Engineer" too while we are at it.
Oh dear - 2008 is going to be busy.
CSI launched in early 2007, the brainchild of Stuart Cohen (formerly CEO of the Open Source Development Labs, which merged into the Linux Foundation in 2006). Its business model is to use the open development model to create necessary software for niche markets. For example, CSI's most recent project is creating software to automate the BITS Shared Assessment Program for companies in the financial services industry.
While it's unlikely that a community would spring up around this sort of software naturally, CSI is using the open development model to create software needed by industries and letting several companies split the development costs. Its role is to do the collaboration and community building, project management, tech support, and foundation building necessary to shepherd the software through its life cycle.
This is an idea that should have been obvious years ago, and we expect to see the CSI do well in 2008-- and attract copycat initiatives as well.
Update I liked the way it read so much that I put it in the post.
I keep running into Nick Sieger in my wiki work. He wrote a couple of very helpful things:
Warbler (replacing Goldspike Rake soon) & ci_reporter mentioned on the JRuby and Continuous Integration page I started in the JRuby wiki today).
My other favorite JRuby guy is Robert Egglestone. He fixed the Rails 2 Goldspike Servlet bug I found Sunday. He has recently updated the Goldspike Roadmap. A release of the Goldspike component is coming this month to follow up on the JRuby 1.1RC1 release the other day.
The discussion on CI caught Mike McKinney's eye. He put together a great how to on JRuby and Hudson.
I love open source and the internet machine.