1. 32
  1. 29

    I too was disappointed by this book. I bought it after reading laudatory comments, probably on Hacker News.

    It is a very gentle introduction to modelling things with differential equations, but there is little to no maths in it, so it is very hard to extract anything actionable from it.

    The system’s bible (https://www.goodreads.com/book/show/583785.The_Systems_Bible) was surprisingly more insightful, but it is trying too hard to be funny, the gold nuggets are hidden below a crust of bullshit.

    Google’s books on the topic were the most useful (https://landing.google.com/sre/books/) but lacked the generality I was hoping for with the first two.

    ESR’s TAOP (http://www.catb.org/~esr/writings/taoup/html/) is hinting a what good design looks like, but agin is too narrow.

    The funniest I’ve read is The Night Watch, it is only a few pages, go read it now https://www.usenix.org/system/files/1311_05-08_mickens.pdf

    I have yet to find a book that explains systems thinking the way I want it, in a general yet actionable way.

    1. 10

      The system’s bible (https://www.goodreads.com/book/show/583785.The_Systems_Bible) was surprisingly more insightful, but it is trying too hard to be funny, the gold nuggets are hidden below a crust of bullshit.

      The Systems Bible was originally called Systemantics and was supposed to be a comedy book. The fact it’s now seen as a serious book shows just how poor the state of systems education is these days.

      1. 4

        This comment is about systems programming while the link is about systems theory, the ‘system’ in those two phrases might be the same word but the fields of study are unrelated. Systems theory is just spicy differential equations.

        1. 2

          I regret I have but one internet point to give you for that last link. Most entertaining thing I’ve read in months.

          1. 1

            Thank you. I bought my copy off the same laudatory comments on HN and got that feeling that I’m reading something not quite real quite early. Now I can save some time.

            Top comment.

            1. 1

              I wish lobste.rs allowed favouriting comments

              1. 1

                I thought you could at first, but then I realized it was story submissions only. So a ridiculous workaround would be to submit the comment as a story and then favorite that. However a better alternative might be to suggest it as a feature 😂

              2. 1

                Back when I looked into it, the book folks said was the best with most industrial use was Yourdon’s Just Enough Structured Analysis (big pdf). I had to put it on back burner. Introductory chapters looked good at showing pieces of realistic systems. I also know his diagrams, which predate UML, were used for a high-assurance device. Take a look at it to either find something you like or further narrow down what you want.

                @hwayne, did I ever send you that one or you ever review it?

                1. 1

                  The system’s bible (https://www.goodreads.com/book/show/583785.The_Systems_Bible) was surprisingly more insightful, but it is trying too hard to be funny, the gold nuggets are hidden below a crust of bullshit.

                  This is one of the best books I’ve ever read, and wholly recommend it to everyone.

                2. 14

                  I’ve read a few “introduction to systems” books and I find they all follow the same pattern:

                  1. Systems thinking will change EVERYTHING!
                  2. What is a system
                  3. One ad-hoc method of studying a system (stocks and flows, state machines, etc)
                  4. “LOOK LOOK HOW ALL THESE PROBLEMS HAVE OBVIOUS SOLUTIONS “, most of which are actually terrible ideas in practice

                  I lost my patience for Thinking in Systems book right around the point she suggested storing nuclear waste on the lawns of nuclear power advocates. It’s all using the language of systems to avoid thinking about systems.

                  The best systems books IMO are ones laser-focus systems thinking on a specific problem domain. Stuff like The New Economy, Engineering a Safer World, Drift into Failure, much of Matt Levine’s Money Stuff. Maybe Cows, Pigs, Wars, and Witches? Stuff where the domain expert is trying to use systems thinking, not stuff where the system thinker is trying to sound like an expert.

                  1. 5

                    I hadn’t thought of Matt Levine’s columns as systems thinking before, but that’s very true. He’s always asking what we’re really trying to accomplish with particular actions and whether it works out or not.

                    1. 3

                      Every time I see these simplistic feedback diagrams I tell the person presenting them to choose a number, any number….

                      Call that number X_0

                      Let X_(i+1) = log(abs(X_i))

                      What a marvellously simple feedback loop. Clean, easy, obvious….

                      Tell me about X_i …

                      1. 2

                        Oh that’s neat. I thought it was going to converge to 1, but the quick Python simulation I wrote has it jumping around quite a bit.

                        1. 3

                          You don’t even need to use nonlinear functions! Even something as simple as x_{i+1} = 4 - 4*x_i is chaotic.

                          1. 1

                            The fun thing with these chaotic systems is start with two different numbers chosen very very close together and then watch the the two series walk away from each other until very soon they’re completely unrelated.