Well, there were a couple of people who found genuine gaps (like XML literals in VB or
yield
keyword in C#). There were a couple of trolls, and a couple of people just having a good laugh ("I prefer C# over VB because I am an American!"). But the majority of comments were pure opinion. "Code is cleaner", "more readable", "I hate semicolons", "I love curly braces", "too verbose", "closest to plain English" were some of the statements repeated over and over. IMHO, this entire discussion sheds more light on the .NET development community than on programming languages themselves.It's no secret that people come to software development using [at least] two separate routes. Some study Computer Science in college (even if it's not their major or they never graduate). They are probably taught programming courses in Java or C++, so C# comes naturally to this group. Second category of developers started out in a different line of work and discovered Office automation with VBA somewhere along the way. Or perhaps they learned VBScript in order to maintain their department's ASP page on the intranet. When .NET came along, this group made a transition to VB.NET.
Now, I'm not trying to argue which group has better programmers - I've seen extremely bright engineers without CS degree, as well as some dim bulbs who turned out to have a Master of Science in CS. But it's a common knowledge that C# was designed from the ground up as a managed object-oriented language, while VB.NET is essentially the outcome of multiple cosmetic surgeries made to an aging body. First change happened when original BASIC - Beginner's All-purpose Symbolic Instruction Code - was updated to support structural programming. It has acquired the "Visual" prefix, but didn't become fully object-oriented until its VB.NET incarnation. Nowadays, Microsoft works diligently to keep the language on par with C#, adding constructs like generics, lambda expressions, closures, and so on.
However, the efforts to modernize VB have little impact on most VB programmers, who probably just aren't familiar enough with contemporary design and programming patterns. So, it's no surprise they tend to get a little bit defensive...
Interestingly, I myself managed to travel both paths to software development. My college major was Applied Mathematics and Cybernetics, and I had plenty of instruction on typical CS subjects. We used Turbo Pascal in the classroom, and by the end of school I transitioned to Borland C++. Incidentally, Soviet Union imploded at about the same time, and in the chaos that followed, my aspirations to find a job in IT became laughable (people were lucky if they had any job at all - it was not unusual in those days for a doctor to work as a taxi driver). So, I ended up doing bookkeeping, accounting and then business planning for a big multinational corporation.
Before long, I was dabbling in Microsoft Access and creating automated databases and spreadsheets for my team. VB was easy and forgiving, and, more importantly, it was ubiquitous. When I finally managed to switch my career back to IT, I didn't feel comfortable with latest C++ tools and frameworks, so I stuck with VBScript and VB6. When .NET was introduced, my first instinct was to transition to VB.NET. However, I decided that it was time to re-educate myself. I started reading about design patterns (which weren't even on the radar when I was in college), test-driven development and extreme programming. I studied source code and tackled new classes of problems, like multi-threaded services development.
Eventually, I realized that C# was a better choice for me, made a switch, and never looked back. This was around 2005, when gap between the two languages was fairly big. Five years later, it is almost gone. But like I said earlier, it's easier to update a compiler than to change people's mindset. Both VB and C# are here to stay, I'm just waiting for someone to port another of my college-era languages, Prolog, to .NET framework...
3 comments:
yes, they already transitioned Lisp into LINQ to SQL with Lambda. They can cause .Net is interpreted and Lisp is functional language.
Prolog, however, is a complete different story, not because of implementation but because they need to marry logic semantics with back-to-earth abstraction of OO. These are very different paradigms.
I just wrote a comment that was completely erased...
In essence, it is easier to marry OO and functional ( C# and Lisp with its Lambda expression staff) especially having interpreter than it is to marry OO and logic ( C# and prolog). Later is completely different paradigm... There is the same impedance in a way as there is between OO and databases...
Workflow Foundation has a built-in rules engine that is a step towards Prolog-like inference engine. And I believe F# already has list-matching capabilities. Come on, Microsoft Research, mix them together and give us Prolog.NET! Or call it P#, I don't care.
Post a Comment