Wednesday, February 21, 2007

RE: Not Another C# Versus VB Article

I was just digging through some old emails, and I came across this gem. This was a follow-up discussion I had with Nigel Shaw, author of Not Another C# Versus VB Article posted on CodeProject.

Date: May 26, 2005 1:05 PM
Subject : RE: Not Another C# Versus VB Article

Hi Nigel,

I just finished reading your article, "Not Another C# Versus VB
Article", and first of all let me say thank you for this work - I'm sure it will be very useful for consultants like myself, as a well-thought-out argument for building development teams on and switching them to C#.

As a developer with about 12 years of professional experience, my history breaks down into something like 1 year of BASIC, Pascal, Fortran, etc., 5 years of VB (vb3 through vb6), 2 years of Java, 1 year of VB.Net and C#, and then 3 years of C#, MC++, and C++. Needless to say, there have been periods in my career where I have considered myself a VB Guru, a Java Guru, and more recently, a .NET Guru. Thus, I feel particularly qualified to comment on your article, and also maybe offer some insights.

I absolutely agree with your central premise - that the fundamental difference between VB and C# is one of culture, rather than technology. Prior to the emergence of .NET, after I had a good amount of exposure to Java, I began to develop very similar conclusions about the differences between VB and Java - which were, after all, much more technically different than are VB.Net and C# - but even then, I believed the central premise applied - as different as they were technically, the really fundamental difference had a lot more to do with what kind of culture gravitated towards each technology. That said, at the time I always had the feeling that there was something fundamentally flawed with each of those 2 technologies, and the general culture that they seemed to promote. On the VB side, of course, you have the opposite 20/80 rule that you've commented on well in your article, in terms of developer capability. On the Java side, my experience at the time was that there was a strong culture in terms of technical competence, but that there was also a lack of focus on solving problems efficiently, often opting instead of the elegant or over-architected solution. This often lead me to believe that perhaps the VB culture actually was the bestwe could do, because it surely would not be susceptible to "too much" elegance!

Then when .NET emerged, my initial impressions were that the platform itself would provide a great step forward in terms of combining the benefits of these two cultures, while eliminating many of the negatives. After some experience dealing with cross-VB-C# teams and seeing how the two sub-teams interacted, I would absolutely say that I can confirm many of your assumptions about the culture breakdown being carried forward to these two technologies. I have been amazed at times to see the same exact piece of logic produced in two entirely different ways by VB.NET and C# teams - and with those two different ways demonstrating precisely the difference in cultures we are discussing. When I initially moved back from the Java world into the new .NET world, I was very gung-ho on VB.NET, first because I was sick of much of the flawed ideology of Java and saw much of that carried on to C#, and second because I was similarly enthusiastic to get back into what seemed to be the most MS-oriented technology - VB.NET. But after time, I began to pick up on these culture differences, and as I found myself gravitating towards the C# culture more and more, I found that the technology conventions supported by C#, and more importantly, those more "encouraged" by C#, were much more in-line with the type of development that I wanted to do.

Ok, so a few notes for your article:

1. Under "The Culture of C#", you write the following: "Marc Andreessen, CEO of Netscape and therefore the 'God of the Internet'" - I believe that you are referring to "god" in the pluralistic sense of the word, i.e. "the god of love", "the god of war", etc. - thus it would be not only incorrect English, but also potentially offensive to those that believe the word "God" should only be capitalized when referring to the one God.

2. The statement "Java applications...were more powerful and feature-rich than the vast majority of VB apps..." struck me as a bit off-base when I first read it. While you may be technically correct in a certain sense, I think this statement is a bit loaded and really depends on context and point of view of the reader. For example, if I am thinking from the viewpoint of a Windows GUI developer, this statement would seem absolutely false - even today it's not necessarily true, because even the most modern Java windows toolkits cannot match the power and look-and-feel of the equivalent toolset in VB - thus it is simply not possible to build Java Windows apps that are as "feature-rich" in that sense as VB apps. However, if I am developing server-side multi-threaded distributed messaging clusters, then yes, of course Java offers a lot more capability than VB has offered in the past, or even maybe today. But the point is, the "vast majority" of VB apps up until .NET were most certainly Windows GUI apps, while the vast majority of Java apps has always been server-side. So the statement as you have it is very "apples-and-oranges" to me. I think the point you are trying to get across is about how Java was technically superior on many levels at that time - I would just re-write this in a less ambiguous and debatable way than what you have.

3. I really like the section "Propagation of the Culture in .NET" - but I think you would do better to split up your list into 2 lists or something. About half of the points are what I would consider absolutely undebatable points about what is wrong with VB.NET (1, 2, 3?, 4, 7). However other points (5 and 6, and to a smaller extent maybe 3) are very debatable in my opinion, and seem to have a lot more to do with personal preference. I personally believe VB is the superior one on points 5 and 6 for example, and I don't agree with your conclusions that they detract absolutely from the potential of VB to promote a strong "culture". I believe the opposite could be true in fact - if these features were part of C#, they would be a further benefit to the environment, and would help provide distinguishing factors that set C# ahead of Java (like properties, events, delegates, etc.). So my point here is that at least a couple
of your arguments here seem to be very much based on personal preference, and much less like "absolute truth" that should be logically accepted by anyone reading the argument - and that detracts from the rest of your points, which are much more like absolute truths.

4. Speaking of those absolute truths, I would make some further comment to expand your point #4 - I believe this is truly fundamental to the continuation of the negative aspects of the VB culture. For example, when I first moved to VB.NET, I obviously started looking for parallels and language equivalents between VB and VB.NET. When I saw two options for doing the same thing, for example String.Length and Len(String), I noticed that the first one looked a lot like the Java way of doing things (which I was trying to escape from), while the second looked like good old familiar Len() in VB6. I continued to use this function until it came time to port a section of code from a VB.NET assembly to a C# assembly a colleague was working on - and then it hit me - this non-standard, non-object-oriented VB extensions stuff is not directly portable! If I would have wrote it with the new .NET convention String.Length instead, it would have directly ported, basically with the addition of semicolons here and there. Instead, it because a non-trivial excercise to port code between the two languages, because it was actually more like porting code bewteen two cultures!

So, once I realized this, I had to go back and re-examine a lot of what I had been doing in VB.NET and try to understand the "right way" for true .NET compatibility. This excercise really helped to clarify my understanding of the distinctions between the object-oriented aspects of VB.NET and the non-object-oriented aspects. Even more interesting, when I finally realized that all of these things simply exist in that VB assembly, I thought I should be able to simply reference that assembly from C# and give myself all the same "powers" from within the C# environment. This would have been especially useful, for example, with the famous IsNumeric() function in VB, which doesn't seem to have a .NET equivalent, but which was always very useful in VB. Unfortunately, this doesn't work from C# - you can't reference the VB dll and call all the VB functions, because it just doesn't work! So this led to an ultimate rewriting of a lot of code to remove usage of Instr(), Mid(), Trim(), Len(), IsNumeric(), and countless other non-standard-.NET functions that seem only natural to a VB

As a final note, I would recommend a book to you, if you haven't already heard of it - I just finished reading it, after having it reccomended by a friend of mine, and I'm not sure if it is because the content is still so fresh in my mind or not, but while reading your article, I found myself going back to the ideas in the book many times and drawing parallels. The book is called "The Tipping Point", and discusses the ways that ideas turn into trends and then into epidemics, and how specific kinds of individuals are attracted to those ideas.


I received a gracious reply from Nigel, and wrote a follow-up:

I'd be happy to be involved in a rewrite. Let me know what you have in mind, and perhaps we could start brainstorming a new outline, new ideas, etc.

Not sure what you think about this, or if there is already significant work out there - but I would consider expanding the scope to cover the "evolution of culture" as it has gone on through the various iterations, including a bit about culture as it applied to both Delphi and Java as well, and maybe even new contrasts between "J2EE" and ".NET" culture. My predictions for the future would be an eventual "survival of the fittest" race towards a re-unified culture in which the best of .NET and the best of J2EE co-exist happily in a single culture.

However, while this topic of general evolution of cultures is kind of a pet subject area for me, it may not be the best idea to expand your overall scope, as the current main premise you have (stated loosely, "C# is better than VB") is very pointed and controversial, and no doubt has a lot to do with what has drawn such a great response for your article throughout the community.

Perhaps along these lines I would consider a separate article as a subsequent chapter in the "series", following on from the basic premises in this one...

Hmm, I'll have to put some thought into it myself...


No comments: