Preface p. ix
List of figures p. xi
Acknowledgments p. xiii
Introduction p. 1
What You Need to Know p. 3
The software development process: what is it you do all day? p. 3
The application lifecycle: from concept to construct p. 4
How to deliver effective bioinformatics applications p. 6
What Is Software Engineering? p. 9
Software engineering: applying formal rules to creativity p. 9
The benefits of discipline: reproducibility, stability, solidity, reliability p. 11
A limitation of engineering p. 11
On the edge: agile computing and evolutionary development p. 12
Before Beginning p. 15
Project Definition p. 17
What you have got to do, in one sentence? p. 17
When have you got to deliver? p. 18
How will success be measured? p. 18
Who will measure success? p. 19
Overall vision? Expected results? Required functionality? You need them all p. 19
Work with what you have p. 20
Requirements Capture p. 21
When your customers are talking to you about what they want, stay focused p. 21
User stories p. 22
Avoid 'blue-skying' in meetings - summarize, don't improvise p. 24
Pencil > keyboard: screen drawings p. 24
You will need to know what your sample datasets and expected results will be p. 29
Yes, now is the time to start thinking about how to test for success p. 30
Documents describing staged deliverables need to be agreed to, if not formally signed off p. 30
Get requirements capture right at this stage, so that toppriority requirements can go into the functional specification p. 31
Separating Function, Interface and Implementation p. 33
User stories will tell you most of the functionality that is required p. 33
From screen drawings to screens p. 37
Your implementation has to create and connect the interface and the functions p. 37
Implementation Considerations p. 39
Languages p. 39
Platforms p. 40
Operating systems p. 40
Prototyping tools p. 42
Debugging tools p. 42
Proof of Concept, Prototyping and Buy-in p. 45
Explaining the development process with prototypes and proofs of concept p. 45
'Too much information! Too much information!' p. 46
Markets: how to sell yourself and your work, and why you need to p. 47
HyperCard, genetic algorithms, evolutionary development and 'doing science' p. 48
Using adaptive development to encourage customer buy-in p. 50
Getting it Done p. 53
Data in, Data out and Data Transformation p. 55
Begin, process, end: process flow diagrams p. 55
Boxes in boxes: data structure diagrams p. 59
Where to Start? p. 63
You will need to get in data p. 64
You will need to cope with errors, because they will happen p. 64
You will need to report results p. 65
Known and unknown coding p. 67
From designs to pseudocode p. 67
From pseudocode to code p. 71
Functional, then Optimized p. 73
Get it out of the door, and in front of your customers, as soon and as often as safely possible p. 73
Planning shows p. 74
When not to display work in progress - getting them used to having it p. 75
Customer retention and repeat buyers p. 76
Optimization: benchmarking, assessment, refinement, hardware p. 77
Focus on essentials, because your customers will p. 78
Coding Style p. 79
Don't write it if you don't have to p. 79
KISS - Keep It Simple, Stupid p. 80
When you look at your code, does it make your eyes hurt, or does it look like poetry? p. 80
Naming of parts p. 81
Going gracefully into the darkness p. 82
Prevention is better than cure p. 83
Flexibility is vital p. 83
Keeping track of what you are doing p. 84
For Some Values of Done... p. 87
Writing the Friendly Manual p. 89
Start as you mean to go on: make your code clean and clear p. 89
Using what you have done so far: it's half-written already p. 90
Writing documentation from functional specifications, and vice versa p. 91
Ask yourself, and others, who still needs to know what? p. 91
Testing - What and When p. 93
Test plans p. 93
Writing a test plan p. 94
Expect the unexpected p. 95
Why developers shouldn't be their own testers p. 95
Involving users in the testing process p. 96
Some things to watch out for p. 97
When to stop testing p. 99
Rollout and Delivery p. 101
You will need separate development, test and production environments p. 102
Delivery notes p. 104
Verbal handovers p. 105
Building the installation package p. 105
Instant gratification mode p. 106
The useful/stable balance p. 107
Support and Feedback p. 109
When do you start? p. 109
Pre-emptive support p. 109
What are your local customs? p. 110
Development costs v. support costs p. 110
Watch out for unanticipated feedback requiring urgent priority readjustment p. 111
Planned and Unplanned Enhancements p. 113
Good applications are never finally finished p. 113
Things not to say in meetings, No. 94: 'Oh, you found that one...' p. 114
'It's not a bug, it's an opportunity to further enhance the user experience' p. 114
Big shoes: managing change in the workplace p. 114
Priority 2: It'll be along real soon now, when we're all less busy p. 115
Slippage p. 115
Project Signoff p. 117
Dealing with bad stuff: focusing on the successfully delivered objectives p. 118
Identifying potential endpoints and agreeing an exit strategy p. 119
Where now? p. 119
Index p. 121