5 ms·

> Even Fred Brooks' Mythical Man-Month - informed by his leadership of the massive bet-the-company "360" project at IBM - lacks data. It's still great though (i

> Even Fred Brooks' Mythical Man-Month - informed by his leadership of the massive bet-the-company "360" project at IBM - lacks data. It's still great though (in, uh... my subjective opinion, I guess). I have just read that book. It is the computer book with most tables pr page of any I have seen. How you come to this conclusion is beyond me.

hyperpallium 7y ago
The "data" in this context is empirical data in support of the hypothesis that the man-month is mythical (aka adding people to a late project makes it later). You also need a control. Similar for other hypotheses, such as a "programming product" and "programming systems" take three times longer than a "program" (written by e.g. a garage duo). BTW Your "beyond me" was unnecessary. https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html
- hyperpallium 7y ago
  Thought I'd better check this. I went through the book's eponymous chapter 2 The Mythical Man-Month (p.14, 1975 ed), and it has not a single table; not a single datum. It has many graphs, without data points, and the text and their smoothness indicate they are graphs of functions, not data. I have a disturbing feeling that the downvoters in this thread simply don't know what it means to validate a mathematical model with data. BTW It's not a criticism of the field that it's not yet an engineering. It's just where it is at the moment.
  - blub 7y ago
    A philosophical book of essays about software from 1975 is not relevant for a set of recommendations on architecture 44 years later. You are correct that chapter two lacks supporting evidence. And that's a problem for the book, it doesn't mean it's ok to omit data everywhere now.
    - eitland 7y ago
      I seems to me that you and hyperpallium agree. I still hold that 1.) That book is still one of the most relevant books I've seen lately (which might be scaring, I haven't made up my mind yet) 2.) While he might not have (or at least not include) data, he argues hos case well, and while I wouldn't bet money on experiments to follow his smooth curve I would be willing to bet (a small amount of) money that for non-trivial software his ideas would still result in better estimates than one that accepts "man-months" as a unit of progress. And for those that haven't read the book yet it might be worth noting that while some people might find the book quotable more or less from cover to cover there is also a number of observations to be made that takes a full reading to realize. Here is one I'd like to point out: Some quotes from the book might leave one thinking that he says programming cannot be scaled. I'd say the opposite is true, he just argues that it must be done in a thoughtful way instead of by mindlessly adding people to the project.
      - hyperpallium 7y ago
        Let's be clear: I think (i) the book is great and (ii) your rude and inaccurate comment was not. (i) I said: > It's still great though (ii) I said: > lacks data You replied: > It is the computer book with most tables pr page of any I have seen. > How you come to this conclusion is beyond me. In case you were wondering, this is the part where you apologize handsomely.
        eitland 7y ago
        Can you explain what is rude? I am not a native speaker so I just looked up the phrase you seem to point to and the two first definitions says it means: - Very difficult for one to comprehend or do. - not something that I can understand Am I missing something here? Besides, I feel you moved the goal posts early on. I replied to you talking about the whole book, you replied with a single example. Please let me know if I've missed something.
        hyperpallium 7y ago

hyperpallium 7y ago

The "data" in this context is empirical data in support of the hypothesis that the man-month is mythical (aka adding people to a late project makes it later). You also need a control. Similar for other hypotheses, such as a "programming product" and "programming systems" take three times longer than a "program" (written by e.g. a garage duo). BTW Your "beyond me" was unnecessary. https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html

hyperpallium 7y ago
Thought I'd better check this. I went through the book's eponymous chapter 2 The Mythical Man-Month (p.14, 1975 ed), and it has not a single table; not a single datum. It has many graphs, without data points, and the text and their smoothness indicate they are graphs of functions, not data. I have a disturbing feeling that the downvoters in this thread simply don't know what it means to validate a mathematical model with data. BTW It's not a criticism of the field that it's not yet an engineering. It's just where it is at the moment.
- blub 7y ago
  A philosophical book of essays about software from 1975 is not relevant for a set of recommendations on architecture 44 years later. You are correct that chapter two lacks supporting evidence. And that's a problem for the book, it doesn't mean it's ok to omit data everywhere now.
  - eitland 7y ago
    I seems to me that you and hyperpallium agree. I still hold that 1.) That book is still one of the most relevant books I've seen lately (which might be scaring, I haven't made up my mind yet) 2.) While he might not have (or at least not include) data, he argues hos case well, and while I wouldn't bet money on experiments to follow his smooth curve I would be willing to bet (a small amount of) money that for non-trivial software his ideas would still result in better estimates than one that accepts "man-months" as a unit of progress. And for those that haven't read the book yet it might be worth noting that while some people might find the book quotable more or less from cover to cover there is also a number of observations to be made that takes a full reading to realize. Here is one I'd like to point out: Some quotes from the book might leave one thinking that he says programming cannot be scaled. I'd say the opposite is true, he just argues that it must be done in a thoughtful way instead of by mindlessly adding people to the project.
    - hyperpallium 7y ago
      Let's be clear: I think (i) the book is great and (ii) your rude and inaccurate comment was not. (i) I said: > It's still great though (ii) I said: > lacks data You replied: > It is the computer book with most tables pr page of any I have seen. > How you come to this conclusion is beyond me. In case you were wondering, this is the part where you apologize handsomely.
      - eitland 7y ago
        Can you explain what is rude? I am not a native speaker so I just looked up the phrase you seem to point to and the two first definitions says it means: - Very difficult for one to comprehend or do. - not something that I can understand Am I missing something here? Besides, I feel you moved the goal posts early on. I replied to you talking about the whole book, you replied with a single example. Please let me know if I've missed something.
        hyperpallium 7y ago