Software Engineering discussion
date
newest »
newest »
Yes, several interesting anecdotes.I liked the areas of focus on people and international culture in this chapter. Even small companies often deal with international cultures or developers. Alot of "learning by google search" results in reading code samples by people in other countries too. This added welcome variety from previous chapters.
I found the "specialization is for insects" and "using UML to generate implementation code is a terrible idea" sections suprising.
There is amazing agreement amongst the Three Amigos (and me) that UML 2.0 is an example of design by committee, run amuck. A big part of the clutter added to UML from version 1 comes from an attempt to use it beyond its sketch and blueprint application, reaching for a full blown graphical programming language. Interestingly, Ivar seems to advocate this direction while at the same time mourning the unrestrained growth... you can't have it both ways. I think that Jim and Grady have perspectives more in line with my own on this (and most other) topics. In fact, Ivar's views that everything new is really old were both repetitive and overstated. And, he seems unfamiliar with the growth of software engineering programs in academia that address many of his concerns about university-level education.
All three seem to miss leading the UML effort, and stand by in dismay as they see what the standards body did to their baby. While books like "UML Distilled" (my favorite) can help cut through the clutter, it is very difficult to hide in design tools, which must conform to the standards. It was fascinating to hear that Rational (where the three came together to define UML and the associated tools and which is now owned by IBM) did not use the tools they built. Overall, I think that UML is a good thing if (as the originators seem to do) you only focus on the essential core.