Kill It With Fire Book Review (Marianne Bellotti)
14 Jul 2021
From Author:- Marianne Bellotti
This is a reference to the first printed edition 2021 (227 pages)
The printed book also contains index pages
Subtitled with “Manage aging computer systems and future proof modern ones”
There are ten chapters broken down into useful detailed sections by page number
How to survive a legacy apocalypse?
The book doesn’t actually adovocate for destroying legacy systems, implementing a scorched earth approach as the title states - rather it’s about identifying and adding value to them over time. This is a book, mainly, about large legacy projects and about cultivating mature systems that can be maintained.
To survive the apocalypse “If you don’t cannibalize yourself, someone else will.” - Steve Jobs
Legacy systems exist in Organisations because they were successful in their era of inception and contemporary technology has moved on leaving these older systems marooned. Where as older systems that are not used can simply be turned off or cease being manufactured (e.g the demise of Apple’s iPod)
The book looks at examples of first contact with a legacy system (e.g a mainframe or database) and ways of incrementally gaining “wins” back to a maintainable state. There are no Silver Bullets.
The main contents of the book are engineering transformation strategies of legacy systems that can be applied to Organisations with these questions:
What the stages are?
What the problems and messes are?
What the solutions are?
What the costs are?
The book provides thought exercises around the above questions
This short book doesn’t contain code snippets or techniques for specific programming languages - such a book, for example would be Working Effectively with Legacy Code
In the excellent conclusion the book states the hardest parts with turning legacy systems into maintained systems are interconnected with the Organisational structures and can be addressed with the correct feedback loops that re-inforces the value the system has for the owners. The flaw with legacy systems is they are often built with the assumption that it will be the last and ultimate system. Legacy systems, that are so old, have been disrupted by whole new economics and newer solutions will have appeared.
Perhaps We are headed toward Vernor Vinge’s far future programmer-archeologist of the mature programming environment - the-arrival-of-the-programmer-archeologist