Jason Mundok's DevCon 2009 Recap

I had an exceptionally great time at DevCon this year. Approximately 1,100 FileMaker enthusiasts gathered at the San Francisco Hilton, including Marc Adler and me representing IT Solutions. I stayed busy chatting with some of the top speakers for the new video interview series “Five Minutes with Philly FileMaker” and networking with as many new friends as possible, but I did manage to attend a handful of hand-picked sessions. Here’s a breakdown of what I witnessed!

John Sindelar – Year in Review

John Sindelar, SeedCode, gathered a whole host of goodies from around the FileMaker community and presented them in one full-on recap session to kick off Devcon Friday morning. One of the highlights for me was John’s demo of a database by Apphia Solutions that allows users to access and write to FileMaker text fields through the web viewer. The solution utilizes some PHP and takes advantage of a free tool called the fckeditor. John briefly touched on the complimentary web services plug-ins from Beezwax and 360 Works, both of which need investigated further.

Things really started heating up when John showed off the drag and drop functionality made famous by Andrew Persons of Excelisys and a technique where portals display records from the contents of an array stored in a global variable (huh?). That’s what I thought until John showed an amazing drag and drop demo put together by Jason Young of MightyData. The demo showed an extremely elegant layout where resources (from two different source tables) were dragged and dropped into a set of portals representing days and shifts. It brought together several pieces of functionality and really showed some “game changing” interface design and functionality.

Matt Navarre – Extend and Optimize FileMaker Search

Matt Navarre, MSN Media & FileMaker Talk, broke down searching in FileMaker and shared a bunch of metrics from his own personal performance testing, such as the difference between searching for data in a local table versus searching for the same data in a table one relationship away, or the difference between using “go to related records” instead of a find.

Matt then went under the hood and explained his new fmSearchResults product, which is available from SeedCode or MSN Media. fmSearchResults is a FileMaker tool that offers a Google-like search from a field that is added to the top of each layout in your solution. When a term is searched using that field, results for matches (and similar matches) from any field in any table in the solution will be returned in a single view. This technique can be easily integrated into any existing solution. The most amazing part of fmSearchResults for me was the amount of built in logic to anticipate users’ intentions, such as name and address substitution.

As a grand finale, Matt showed us a database where he downloaded every article in Wikipedia (over 3.2 million) as of May 2009. One table exceeded 40 million records. Using fmSearchResults, Matt did a search for “Eric Clapton” and results were returned in a split second. It was a mind-blowing display of the results of his work and studies over that past several years that have been realized in the form of fmSearchResults. By the way, you can download Wikipedia as text files at dbpedia.org.

Albert Harum-Alvarez – The Idiom of FileMaker

Once again, Albert Harum-Alvarez, SmallCo, took us on a mind bending ride through the concepts of design. His overarching theme was about working within the FileMaker idiom, or in a way that takes advantage of FileMaker’s native behavior, a concept that I think the FileMaker 10 release supports well.

He encouraged us to keep solutions simple and suggested that complex functionality can still be simple. How do we keep applications easy to understand for the next developer? We are often our own “next developer”. Working in the FileMaker idiom can help keep things simple. Here are some characteristics of the new idiom:

• List views are back – now that we have script triggers & dynamic reports our exclusive love affair with the portal starts to break down and we can return to the list view as an extremely valuable tool.
• Modify tables – extended power in the hands of users, again with dynamic reports and script triggers.
• Reusable scripts – we’re gaining more and more ability to abstract our scripts and reuse them. Do it!
• Factoring the relationship graph – use custom functions and get Nth record as query functions and depend less on the graph. Keep your graph lean.

Unconference – Secrets of a Successful Consultancy

I think that’s what the session was called, but don’t remember for sure. In any event, it was an all-star panel moderated by Matt Navarre, MSN Media & FileMaker Talk with Kirk Bowman, MightyData, Jesse Barnum, 360 Works, Molly Connolly, Thorsen Consulting, and Jonathan Stark, independent. Each panelist had an opportunity to explain how they run their own business. There was a nice balance of situations and style represented: independent vs. employers, custom development vs. product development, etc.

Two take away points for me were Molly’s explanation of how she deals with change orders and Jesse’s approach on estimates. Molly factors a percentage of the project budget and adds it to the bottom line to cover changes and enhancements that will come up throughout the process. The client may not end up spending all of it, but by arranging for it up front, the money is approved and can be used as needed without much hassle. Jesse explained how he keeps scope very small and estimates each piece of functionality. I’m assuming he works more in small phases rather than big projects.

Overall, this was an excellent discussion and even though it wasn’t an official session, it was attended by some of the top names in the industry. This home grown approach to sharing ideas is becoming more and more evident as this session was created by the panelists and promoted through podcasts and social media in the weeks that led up to Devcon. The room was packed and the discussion rich. Kudos to all involved.