The evolution of C++: Why pointer to data members

The anser: I dont know… But I needed to find out.

It all started when I was reading my copy of Design and Evolution of C++ by Stroustrup himself. It is a superlative work on the design process behind (arguably) the most popular language on the planet.

In one of the topics [13.11] I was confused by Stroustrup’s statement:

pointers to data members have proven a useful way of expressing the layout of a C++ class in an implementation [Hübel, 1992]

Now that particular reference goes to a university research paper :

An implementation of a persistent store for C++ - Aarhus Universistet

With nothing else to go on, I searched on the university site (yes. I dont speak danish. No - I’m not being stupid). When I could find nothing, I tried very hard to track the author down.

Very, very hard.

No trace whatsoever - except on LinkedIn (where I dont have an InMail account). Finally I found a posting in Danish on a .Net site. I registered myself (that is a long, long tale… and we shall not speak of it) and managed to send a message to Peter. Of course I had no hope of getting a reply.

And then today, Peter replies back to me and sends me a link to the Gzipped postscript of the manuscript. Cool!

Thanks Peter… I hope u dont mind me linking to a PDF version of the same [An implementation of a persistent store for C++]

P.S.: Why did’nt Google index this PostScript? does it have a problem with gzipped postscript?


The evolution of C++: Why pointer to data members


July 11, 2008

Find me on Twitter @sandeepssrin

Did i make any mistake? Please consider sending a pull request.