ITK/10th Anniversary Activities/ITK Stories and anecdotes

From KitwarePublic
Jump to: navigation, search

ITK First Decade Stories and Anecdotes

What ?

Share with the community your stories and anecdotes of using ITK during the past ten years.

Format ?

This is an informal forum.

Short stories with funny connotation are more than welcome.

Examples

  • Tell us about
    • The oldest computer in which you have run ITK
    • The weirdest compilation option that you have tried with ITK
    • The longest you have run an ITK program

Other options:

  • Things my supervisor never knew about how I used ITK...
    • (anonymous submissions are fine in this one...)
  • Saved by an Insight Journal paper
    • Have you been saved from having to write a lot of code, thanks to an Insight Journal paper ?
    • Tell us your story.

STORIES

ITK: To Infinity and Beyond!

A highly experienced medical image processing guru told me about ITK (and VTK) in 2002.

I found myself intrigued by the promise an API which would provide a foundation to allow open source research and development developer's create cutting edge applications.

Ever since, I've been impressed by the consistency and power of this API, the strength of the open source developer community and the rapid collaborative feature implementation and bug fixing. (Itk,Vtk, Paraview, Volview and Igstk to name a few).

One particular area which I was hoping to get more insight into was the set of optimization techniques, part of ITK's libraries, which received a major upgradein ITK release 3.14.

Thanks to Kitware for providing a platform for innovation!

-Cartik S. Sharma

ITK Experiences at UNC, the CADDLab

ITK Grows Up!

  • This story illustrates the difficult path ITK followed in its early years.
  • Part 1: 2001-ish
    • Michael Bell, a research assistant in the CADDLab, walks into Stephen Aylward's office to discuss his project. He has begun using the beta version of ITK. When asked what he thinks about ITK, he responds:
      • "There are so many templates and long function names that I don't feel like I'm programming in C++ anymore" He did not mean that as a compliment!
  • Part 2: 2003-ish
    • Julien Jomier, a research assistant in the CADDLab, walks into Stephen Aylward's office to discuss his project. Stephen suggests that he consider prototyping it using Matlab. Julien responds:
      • "Actually, I'd rather use ITK. It's faster." That was one of the best compliments ITK could receive!

CMake, the only good thing to come from ITK?

  • Even the ITK development team had their doubts at the beginning!
  • George Stetten and Stephen Aylward are sitting next to one another in an ITK developers meeting in 2002-ish. Bill Hoffman is presenting CMake, which is already gaining popularity. George leans over to Stephen and says
    • "Ever worry that the only good thing that will come of ITK is CMake?"
    • There is no doubt that CMake has been a huge success, but so has ITK!!!

ITK early adopters

  • George Stetten and Stephen are talking at a conference. George has just begun converting his lab to use ITK for all of its projects. Stephen asks George how it is going, George responds:
    • "Research in my lab has ground to a complete halt."
    • Early integration of ITK into a lab was extremely challenging. It has since been greatly facilitated by its extensive software guide, CMake, and community support.

University of Iowa

Insight Journal, Success Story

From Hans Johnson at the University of Iowa:

http://www.itk.org/mailman/private/insight-developers/2006-February/007661.html

 My graduate student has been doing some atlas based segmentation were signed
 distance maps of binary objects are passively deformed.  One of the major
 bottle necks is creating the signed distance maps with the Danielson
 Distance Map, and the Future work section of his thesis stated that the work
 by Maurer should be investigated as a possible way to speed this up.

Timeline to success:

  • 9:26am e-mail arrives about a new Insight Journal submission
  • 12:30am print and read documentation while
  • 12:45am download source code, replace Dannielson Filter with new EDT filter, and compile
  • 1:00pm run on 1 dataset
  • 1:03pm curse because it could not possibly have been done in only 3 minutes.
  • 1:10pm compare new results to old results, do dance of joy because new results are nearly the same as old results
  • 4:00pm Write Insight Journal review
  Good work Penn Image Computing and Science Laboratory!
  Thanks,
  Hans

Bill Lorensen

Dashboard Horror Story

In preparation for release 1.4 (September 15, 2003), Luis started sending automated e-mails to the list reminding everyone of the date for the last checkin before freezing the repository. The freeze date was August 1, but Luis started sending out the automated e-mails July 2. Here are the subject lines:

  • July 2,2003  : FREEZING: Release 1.4: August 1
  • July 11  : FREEZING DATE: 22 Days
  • July 16  : FREEZING DATE: 16 Days
  • July 23  : FREEZING DATE: 10 Days
  • July 28  : FREEZING DATE: 5 Days
  • July 30  : FREEZING DATE: 55 Hours
  • July 31  : FREEZING DATE: 31 Hours
  • August 1, 2:40AM : FREEZING DATE: TODAY - 18:00 EST
  • August 1, 3:37PM : FREEZING DATE: - 2 Hours 14 Minutes
  • August 1, 4:49PM : FREEZING DATE: - 1 Hour 2 Minutes
  • August 1, 5:30PM : FREEZING DATE: - 31 Minutes
  • August 1, 5:52PM : FREEZING DATE: - 9 Minutes
  • August 1, 6:01PM : CVS FROZEN!!

That night, as the nightly builds started to appear, the dashboard was littered with compilation errors. Here is the e-mail I sent to the developers at 10:32 PM.

From: Bill Lorensen [mailto:wlorens1@nycap.rr.com]
Sent: Friday, August 01, 2003 10:32 PM
To: insight-developers@public.kitware.com
Subject: [Insight-developers] Red (black) Friday for itk

Folks,
This was not a good day for itk. The flurry of checkins wreaked havoc on the
system. I'm afraid it will take several days to recover. In the meantime,
everyone's productivity will suffer.

It is apparent that people are checking in code that has never been
compiled. This is inexcusable. We have a system in place that supports
professionalism, not anarchy.

Please check the repository and see if someone else had to fix your code so
that it would compile.

Bill

I also sent a personal e-mail to the fellow that checked in the code that "wreaked the havoc." My tone was a bit stronger. That was the first and last day he checked in code. To this day, I still feel bad about that e-mail.

Bill

The next giraffe is on me

A Beer Giraffe

The ITK contractors met at least once a year at the National Library of Medicine in Bethesda. Jim Miller and I usually had dinner the night before in a pub called the Rock Bottom Brewery. The food was pretty good, but the beer was the highlight. They brew their own and had a nice selection of ales and lagers. One night Jim and I both happened to order the same beer before dinner.

The waiter asked, "Since you are both having the same beer why not get a 'carafe'? It is quite a bit cheaper." Sounded good so we did.

He came back shortly with a tall cylinder filled with 2.5 liters of beer and a tap on the bottom.

The waiter said, "Here's your giraffe gentlemen. Enjoy"

And we did.

Let it snow! Let it snow! Let it snow!

January 4, 2003. MAJOR NOR'EASTER, day 1: This was a long duration storm beginning in the morning with periods of light to moderate snow through the day, snow became moderate to heavy late in the day and continued that way through midnight, snowfall rates were highly variable through the storm, at times reaching 2" per hour, wind was not strong, the 12" of snow at Albany was enough to set a 24 hour snowfall record for this date. MAJOR NO'REASTER, day 2: Moderate to heavy snow became more light to moderate and patchy after 3:00am, but continued off an on through the entire day, wind remained light, the 1.51" of melted precipitation was good for a new 24 hour record for this date, the 8.8" of snow was also good for a new 24 hour snowfall record for this date at Albany.

The ITK contractors met several times each year for status/working meetings. Occasionally we had snow. This happened three times: Niskayuna, Salt Lake City and Philadelphia.

In early February 2003 we met at the University of Pennsylvania. Jim Gee was the host. The evening before the meeting, Philadelphia received so much snow (1.2 inches according to PhillyWeather.net) that the city came to a screeching halt. The University pretty much shut down. Fortunately, we were able to use the building that hosted the meeting. I recall the meeting being successful and we all enjoyed our stay in the City of Brotherly Love.

After the two day meeting, we returned home without incident. I sent Jim Gee the following e-mail with an attached image:

To: James Gee
Subject: Snow
Date: Sat, 08 Feb 2003 16:21:14

Jim,
First of all, thanks for being a great host. I think the 
meeting was very productive.
 
Also, I'm sorry that you folks are having a tough winter.
 
I thought you might like to see my house on January 4, 2003.

Bill

Here is Jim's response:

From: James Gee
Subject: RE: Snow 
To: Bill Lorensen
Sent: Saturday, February 08, 2003 3:24 PM

Dear Bill:
It was wonderful to have everyone here and experience a bit of
Philadelphia, including our *vicious* winter snow storms.

Jim

p.s. Why do you live on a glacier?

NOTE: The week after our meeting, the northeast experienced The Blizzard of 2003. I assume that our little e-mail exchange had no influence on that event. The NOAA site reports that:

At Philadelphia, PA 18.7 inches of snow fell during the President's
Day snow storm February 16-17, 2003. This is the third highest total
for a February storm event. The second highest total is 18.9 inches
that fell February 12-14, 1899. The highest total for a February storm
event is 21.3 inches which fell on February 11-12, 1983.

Big Mac Mistake

In the summer of 2001, I was invited to give a talk at the National Library of Medicine (NLM) Board of Regents Meeting. This would be the first presentation to NLM upper management on the Insight Toolkit. The scheduled date for the meeting was September 12, 2001. My plane reservation was for 11:30AM on September 11. Of course I never made that trip because of the attack on the World Trade Center. At GE we were gathered for our morning dashboard reviews of VTK and ITK. My wife called to say that a plane had flown into one of the toweres of the World Trade Center in New York.

Needless to say, as events unfolded in the next few minutes, my flight to Washington DC was cancelled. In the notes for the NLM Board of Regents on September 12 the chilling comments:

The meeting ended early because of the attacks on the World Trade Center and Pentagon.

and

Dr. Lindberg had just begun his presentation when notification was made that the NIH was closing due to the attacks on the Trade Center and the Pentagon.

My presentation was rescheduled for February 12, 2002. At the February meeting I described the ITK software and the our successful collaborations between industry and academia. I proudly described how ITK could run on PC's and workstations. And how we tested the system nightly on Sun computers, Silicon Graphics, and a range of personal computers.

Dr. Lindberg, the Director of NLM, asked "Does ITK run on the Macintosh?"

I paused, since Terry Yoo had informed me that Dr. Lindberg was a big Mac user.

"No", I replied.

"Why not?", Dr. Lindberg asked.

"None of the contractors have Macs", I said.

Although Dr. Lindberg enjoyed the talk. And he specifically commented on how he was impressed with our multi-institution collaboration. I think he was disappointed that ITK did not run on the Mac.

Of course, today, the Mac is a well-supported platform. The initial port to the Mac was done by one of ITK's early adopters, shortly after that meeting. I believe he was Jon Harald Kaspersen from SINTEF.

Luis: What were you thinking?

GE Research hosted a Visible Human Segmentation and Registration Toolkit Architecture Meeting from January 19-21, 2000 in Niskayuna, NY. This was the first working meeting of the ITK contractors.

Stephen was the PI from UNC and Luis accompanied him.

Here is the weather report for Albany, NY on those three days. Please note that all temperatures are degrees Fahrenheit.

Day/Date High (F) Low(F) Snow Comments
Wednesday, 1/19/2000 19 -1 0.0" 100% sunshine, light winds, cold and dry
Thursday, 1/20/2000 21 7 1" Cloudy and cold, light snow through the day, minimal accumulation
Friday, 1/21/2000 20 0 1.2" Midnight high temperature of 20, daytime high temperature was 8, pre-dawn light snow in spots, 100% sunshine during the day, windy, gusts frequently to 35 mph, wind chills to -40


Luis Comments:

This was a very enlightening meeting. I learned about the dark arts of driving on the snow-covered and icy roads.

  • Ross Whitaker was very kind to gave Stephen and me a ride to the GE site in Niskayuna, NY.
    • That was the first time I saw someone use the parking break in order to stop a car.
  • On the way there I also learned that you should not pass another car on a bridge,...and finally understood a road sign that I have seen in North Carolina many times but that never made sense to me:
  • Finally I got to discover the Dantesque contraption that no man in tropical lands ever have to know about:

And it's checked in!

October 28-30, 2001 I attended the Workshop on Shape-Based Retrieval and Analysis of 3D Models. Tom Funkhouser hosted this invitation-only NSF sponsored workshop.

I recall three things from this meeting.

  1. The anthrax attacks that originated near Princeton
  2. The list of speakers and attendees was very impressive!
  3. Dimitris Metaxas gave a talk titled "Shape Models for Tracking, Recognition and Medical Applications". During the talk Dimitris described the hybrid segmentation algorithms his group was adding to a new toolkit called, the Insight Toolkit, ITK. After describing the algorithms, Dimitris stated, "And all of the code is checked in!". None of the other attendees besides me understood what this statement meant. Terry Yoo, our ITK Project Officer always told us that unless the code was "checked in", our contract obligation was not met.

Are you with the Ackerman Wedding?

At every contractor meeting, we tried to arrange an evening dinner for the whole group. This was a chance to meet and get to know each other outside of the formal meetings. Socializing is a huge part of any multi-group project. Wherever we met, we delegated the group evening mealk to the local hosts. Dr. Ackerman was our designated social chairman when we met in Bethesda. Several times we had dinner at Trattoria Sorrento. Dr. Ackerman was able to arrange a special menu for us. This was an efficient way to handle a large number of guests. At least once, as I recall, we were presented with the "Ackerman Wedding Menu". The restaurant always placed us in our own area of the restaurant. These were the best meals and social events of our many ITK evening meals.