babellog

Friday, September 17, 2004

The List

88Pro once said quite correctly that what sets apart the IT industry from the others is its maturity(or rather the lack of it). But IT being in it's infantile stage is not without its benefits; the body of knowledge being quite small, it enables the bibliophiles amoung us to acquire a decent CS library. I've been wanting to compile a list of timeless CS books; the classics; the bibles. So to start things off ...

Programming

1. The Art of Computer Programming (Volumes 1,2,3) - Donald Knuth
2. Data Structures and Algorithms - Aho, Ullman, Hopcroft
3. Compilers, Principles, Techniques and Tools - Aho, Sethi, Ullman
4. The C programming Language - Kernighan & Ritchie
5. The C++ Programming Language - Bjarne Stroustrup
6. Algorithms + Data Structures = Programs - Niklaus Wirth



Databases

1. An Introduction to Database Systems - C.J Date

inzy winsy start, soooo, add away people...

Tuesday, September 14, 2004

Java morphs into C++

First we got generics and printf.
printf is a holdover from the C era. And while I won't want to get into an argument about how useful templates and metaprogramming really is, it's probably one of the most hated (or feared) elements of modern C++ (along with memory allocation).

What's next ? friend functions.

Reminds me of that old joke about English being the official language of the EU, but needing a few "minor" changes.

Will we wake up in ten years and find that the wheel has been reinvented ? Again ?

Thursday, September 09, 2004

Comments

Considering the recent problems we encoutered with comments, shall we shift to haloscan comments?

Reasons for using haloscan

1. Easy posting as other comments are visible while posting.
2. And the most important is we can publish the comments feed.

Your comments please?

No, comments will be considered as agreement with change :).

Value of babellog!

At the moment it's valued at B$5,967.83 :-)

Value of babellog

Blog Shares

Monday, September 06, 2004

Technical Gap between Generations

Already in the hacking world people strongly feel that new generation of hacker are not skilled enough. However that’s not what I want to talk about. If you notice at Universities initially they taught C and then Pascal became popular as a teaching tool even though real programmers don’t use Pascal. I definitely think there was little knowledge loss when the transition from C to Pascal happened. Now universities are moving towards Java, as a teaching tool, to teach data structures and algorithms. Although Java might be a good tool to teach algorithms and data structures, I think it hides lot of low level details (concepts) from the programmer. Recently I saw someone writing a file handling program in C without actually understanding concept of file pointer which introduced a bug which he couldn’t fix it for two days. Like languages there is a transition on the operating system too, *nix to Windows (take any company in SL and the 90% of the development platform is Windows, and I guess (not sure) Unis too mostly using Windows platform). Again developers are loosing the needed skills. I have seen people trying to use Excel macros for hrs to do a job which can be done using awk in couple of lines. May be I am just seeing a technical gap which really doesn’t exist, or do we have a real problem t address here?