I found part two exceptionally enlightening – this is the first book/article that I've read which deals with the matter of managing software developers directly, from the point of view of a former/current programmer.
Part one (The Practice of Programming) contains a very useful series of four chapters entitled “Painless Functional Specifications” - again, fantastically simple ideas and solid reasoning as to why it's critical that (almost) any piece of software development should be preceded by the thought process and documentation phase that amounts to a functional specification. I would just about say that this series of articles (or something resembling it) is a must read for any developer. Too many developers (myself included at certain times in the past) in my opinion are all to willing to dive into coding ahead of putting the simplest of designs on paper – this process alone can reveal fundamental flaws in a plan. Joel goes so far as to recommend that developers take a simple course on practical writing, totally agree. Almost all of the other chapters in this section are brilliant and enlightening – certainly however “Painless Functional Specifications” stands out as gold IMO.
Parts three, four and five are not so well organised and start to devolve somewhat into (loose?) assemblies of blog postings; many of which are revealing and fascinating (especially IMO, where Joel makes comparisons to his experiences as a Program Manager at Microsoft). There's a few posting about .NET that are particularly intriguing.
Many of the articles in the book were written in the years ranging 1999-though-2004; a period of time that the .NET Framework was in it's beta manifestation or in v1.0 and v1.1. I find this section fascinating for two reasons. Firstly, Joel has recorded, like a time capsule, some of the ridiculous announcements and summaries of .NET that Microsoft first put out there. I remember at the time, as a student and emerging professional developer being simultaneously intrigued and baffled as to what the heck .NET actually was. Over a period of a couple of years the hype boiled down to an apps development framework that was more-or-less similar to Java. Some of the fantastically broad and meaningless spiels that Microsoft were using to describe .NET at the time made it seem of equivalent magnitude and significance to a manned mission to Mars or Skynet becoming self aware, without providing any detail whatsoever about what .NET actually was. I guess however a campaign that marketed .NET as “Microsoft's equivalent to Java” wouldn't go down to well with non-technical folk. Joel saw though the bollocks back then with crystal clarity. The other element of Joel's commentary of .NET that rang true was his questioning of the need to engage .NET seriously as a development platform, and the cost of rewriting existing code-bases to integrate with .NET. To do this is a costly venture that many businesses are still struggling with the practicality of today. I remember at the time, the organisation that I was working with was wary of .NET and the cost that it represented from a business perspective. Joel's writings from this era reflect the exact same sentiment. For myself, at this time I was a fledgling developer who has just realised what .NET was all about and was keen as to get stuck into it – this didn't help things, and eventually led to my departure from a good company (voluntarily!). I have gone on to become a strong and accomplished .NET developer, at the cost of an opportunity to be part of a great NZ success story. Them's the breaks I guess – I enjoy what I do regardless and am still on good terms with the company that provided me with my first break into the software development industry (they are my Aunty and Uncle – the company is Abtrac), and that's what counts I think. Besides, I have moved on to another NZ success story! In summary, Joel's critique on the turmoil that the introduction of the .NET Framework brought to the Microsoft software development community is in depth and, I think, heartfelt. I enjoyed it.
In summary, this is an extremely useful book that is written in a lively and humorous manner. Well worth a read for anyone who has anything to do with software development. And in my opinion, a must read for Microsoft software development leadership/management people.