FLAP Phase Two began yesterday - Sunday, 23 November 2014 - with the launching of a Preparatory stage.
This stage of the work is being funded by personnel donations from EHS of Jordan, the VISTA Expertise Network, and OSEHRA, respectively. OSEHRA is continuing to host the infrastructure for us here at their community website. The Network is donating my time. EHS of Jordan is donating four team members - Tariq Hamkari, Murat Khemesh, Ahmad Besharah, and Danya Gharbieh. Danya, Ahmad, and Murat set up their new OSEHRA accounts and joined the FLAP workgroup, and Tariq will join shortly, and within the next couple weeks they'll begin posting periodically to this workgroup so you can get to know them and follow along in their work.
The five of us, collectively called the FLAP Jordan Team, met as a team for the first time at 4:00 p.m. Amman Time (5:00 a.m.Pacific Time) and again this morning at the same time. The Jordanian work week runs Sunday through Thursday, with Friday and Saturday as their weekend, so our daily internal calls will be on their weekday afternoons (their time, mornings my time). We introduced ourselves, discussed Fileman's history, its role as the architectural foundation of VISTA, and our plans.
Plan for the FLAP Phase Two Preparation Stage
Our plan in a nutshell for the FLAP Phase Two Preparation stage is to compare three versions of Fileman
A) Fileman 22.0, from the 5 November 2014 FOIA VISTA,
B) Fileman 22.2, from FLAP Phase One, and
C) MSC Fileman 1051, updated Friday by George Timson, which includes the past year and a half of bug fixes and enhancements, and formally contributed to OSEHRA by Medsphere Systems Corporation last week,
to create a fourth version
D) a draft, up-to-date MSC Fileman 22.2, compliant with OSEHRA's new software-attribution policy
that can be used as the basis for updating
E) the Fileman Development Environment
from which a new MSC File Manager version 22.2 can be bundled and released through OSEHRA. This new MSC Fileman 22.2 will be the core of OSEHRA VISTA, the foundation for OSEHRA's Code Alignment Phase One project, and the baseline for FLAP Phase Two.
We also discussed the agile methodology as described here (http://www.agilemanifesto.org) and here (http://www.agilemanifesto.org/principles.html). We did not describe the illuminating satirical counter-example here (http://www.halfarsedagilemanifesto.org), which describes how to claim to pursue the agile methodology while actually torturing it into the traditional waterfall approach, but we will be keeping that in mind as an executive summary of too many object lessons to name, whose examples we very much do not want to follow (My thanks to Dana Miller of UPMC for introducing me to this satire at the Code Convergence Workshop).
Put another way, for this project failure is an option, the only option. We intend to fail early and fail often, to learn from those failures, because the evidence-based methodology recognizes that the only path to success with highly complex ventures is to try things, make mistakes, learn from our mistakes, and so build up the necessary body of experience that lets us fail less often and succeed more often. That's the paradoxical insight that drives true science and engineering, to cast aside the puffery and braggadocio that too often characterize project planning in favor of the humility of empirical discipline and the good-natured openness and curiosity required for one's ego to survive the experience.
The OSEHRA approach to agile - documenting our iterative and continuously improving plans in Atlassian's Jira and Agile tools, capturing our growing body of knowledge in Confluence, and along with our main development continuously creating a library of unit tests with M/Unit - will ensure that neither documentation nor development nor testing nor planning is given short shrift, but that all four are allowed to benefit from agile's empiricism and rapid prototyping. This will also ensure that the FLAP Phase Two Preparation stage is an agile project, not an "agile" one.
For Fileman, the use of the true agile methodology - which emphasizes empirical, evidence-based, ongoing, iterative design instead of rational, evidence-averse, up-front, single-pass design - is the only way to safely transform the foundations of VISTA.
I. The FLAP Jordan Team will set up the first four environments (A, B, C, and D) on their servers in Amman (the fifth [E], along with a constellation of supporting environments, already exists on the FLAP team's servers in the U.S., which are managed by FLAP sysadmin Larry Landis).
II. They will take a close look at the three most recent VA Fileman patches
1) DI*22*167 SEQ #148
2) DI*22*169 SEQ #149
3) DI*22*170 SEQ #150
and compare them to these environments to build a report of the deltas, from which a plan for reconciling them will be created and the patches melded into environment D. That will complete the resolution of envronments A and B.
III. They will then begin the main body of this work, which is reconciling versions B and C to finish creating D.
As this work progresses, we will break down these epics into stories and tasks both here in this blog and in OSEHRA's agile tools.
If you have any questions, comments, or concerns, please do not hesitate to respond with them. Like most OSEHRA projects, this is an open, community-based project. We welcome your input.
If you wish to become more deeply involved by donating time or money to help with this work, please contact me or OSEHRA product manager Mike Henderson, and we'll help you figure out the details to get you started. We welcome your participation.