Simon Murphy of Codematic had lined up an array of speakers and topics that drew a packed crowd. Morten Siersted’s company F1F9 sponsored the lunch for all the delegates. Simon opened the conference with a summary of the evolution of Excel development and closed it with a personal view of project management.
Stephen Allen of ACBA spoke on his auditor’s Electronic Working Papers (EWP) application, which he developed from the need to demonstrate a chain of evidence from the system being audited through the risk to the audit finding, conclusions, and recommendations. He describes this as “a recursive programming approach required to rearrange branched thought processes into linear output for reporting”. A trial copy of ACBA-EWP can be requested from ACBA (UK) where you can also download a demonstration workbook.
Bob Phillips spoke on Personal BI (Business Intelligence) made possible by the PowerPivot addin for Excel 2010. He covered the DAX language and mentioned the work of Tom Gleeson, the datasmith. There were some questions from the Access developers in the audience, as Bob indicated that PowerPivot is the future of data access and reporting by Excel.
Dr. Mike Staunton of the London Business School spoke on his approach to coding style to permit portable code from VBA to VB.Net and C#. He likes the free tool ExcelDNA and TangibleSoftwareSolutions.com code translators, who feature his Wilmott article on VBA to C#.Net.
Mathias Brandewinder spoke on VSTO, showing that version 2 was not the bugbear that version 1 had been. He demonstrated several simple applications including Ribbon and Custom Task Panes; see VSTOStocks.codeplex.com. For UDFs he recommends ExcelDNA and XLL-plus. The audience were with him right up to the point he mentioned that VSTO needs a Visual Studio 2010 Professional licence at $800. However, ask your national MS developers’ contact who may know of special offers with eg Technet.
Charles Williams recounted his experience of migrating 8000+ lines of VBA code in his FastExcel addin to C++. This was necessitated by the introduction of the 64-bit version of Excel. To save him time, he used the Plantech XLL+ addin for Visual Studio and the STL/Boost library. His blog at decisionmodels.com is worth following for tips on improving Excel performance which can be better done by more efficient algorithms than language tweaks.
Morten Siersted of F1F9 spoke of the FAST modelling standard. Aimed at the general spreadsheet financial modeller, it stresses virtues such as transparency (comprehensibility) and being “beautifully simple”. This is achieved through a few rules for organising code in consistent formula blocks, grouped detail rows. He enlivened the rules by comparing their recommendations of staged inputs to a TV cook and font color coding to plumbing diagrams.
Simon provided the usual swag that conference attendees love to collect and also gave away draw prizes – I had contributed my book on Spreadsheet Check and Control, and the Q-Validus Spreadsheet Safe training manual. Some of us,because of our connection with EuSpRIG, were interviewed by the BBC for their upcoming programme on spreadsheet risks in “More or Less” (about statistics and economics) to go out on the World Service. It was good to meet so many faces belonging to people I had hitherto known by email or newsgroups, such as Simon Hurst, Anthony Prince, Andy Pope, Russell Proctor of BetterSolutions, Philip O’Brien of Pasigraphy, and colleagues from previous work. I’ve taken away some New Year resolutions for things to work on!