I’d like to make it clear beforehand that the following novel is fully my opinion and not that of my employer.
Whenever I hear a developer address themselves as a “C# developer” or a “Ruby developer” or an “[Insert programming language here] developer” I really can’t help but cringe. Why? Because as an industry, we should be long past the days where you align yourself to a specific programming language. Now, I’m certainly not saying you shouldn’t pick favorites (that’s just human nature), or even lean towards one most of the time (I currently do ~90% C# development), but at the end of the day, being a great developer means learning core fundamentals which can be applied to any language. If you have a solid grasp of object-orientation, functional programming tenets, the benefits of dynamic programming, and so on, you could seamlessly pick up a new programming language that fit one of those paradigms. Most of the time, languages share ideas between each other, so keeping a keen eye to the languages landscape can also help promote fresh concepts regardless which one you’re currently using.
A better question to ask could be: which language paradigm do you prefer? Many developers are very passionate about dynamic programming languages, and so having to use a statically-typed language might feel constraining to them. But, even that debate is getting blurry with languages offering strong type-inference (e.g. F# or Scala) that can get a statically-typed language pretty close to providing a “dynamic” experience. The addition of LINQ into .NET 3.5 also proved that languages that aren’t functional in nature (e.g. C# and VB) can have a strong functional offering. Things like this lead us to believe that the best question to ask here would be: which language paradigms (note the plurality) do you prefer? Single-paradigm languages are a thing of the past, and developers have an even better opportunity to be productive by using a language that amalgamates the paradigms they prefer in a way that is as close to their preference as possible (no language is without its problems).
That said, I’m also not promoting “polyglot programming” here (which I think is mostly a load of crap). What I am saying is that it’s a waste of time to get caught up into a single programming language or even a single paradigm, when that would only be doing yourself a disservice. This is why I also don’t think that a Java vs. Clojure vs. Scala vs. F# vs. Whatever makes any sense either. Sure, each language is going to have tangible benefits and reasons you could argue that it’s better than another, but “better” becomes largely subjective based on your individual scenario and perspective.
For instance, I personally love Clojure and think it’s JVM interop is just beautiful, but would I recommend it to someone who is familiar with Java and is looking for a functional language? No, because the learning curve is too drastic (unless they’re already a Lisp nerd). Scala is just as powerful and provides a substantially lower barrier of entry for someone coming from Java. And it’s type-inference provides plenty of powerful scenarios that make up for it being a statically-typed language. But even that decision is based on my own perspective. Point being: I believe in being prescriptive, but unfortunately you need to do your own research and make your own decision because ultimately the only deciding factor is your own personal situation/preference.
Languages –> Web Frameworks
What does any of that have to do with the current state of the web? Everything. Just like programming languages, there was a time where it was common for developers to align themselves to a specific web framework. If you address yourself as an “Rails developer” or an “ASP.NET developer” or an “[Insert web framework here] developer” you’re doing yourself a disservice. Why? For the exact same reasons mentioned above. I consider myself a developer first and a web developer second. I like to think that I’m a student of all the cross-cutting concepts that make up what it means to develop for the web. Those skills can then easily translate to any framework (and language) you pick.
The real questions to ask isn’t which framework is better but which attributes of a web framework you prefer and make you more productive. Do you like component-driven development (e.g. WebForms, Wicket)? Are you more efficient using an MVC architecture (e.g. Rails, ASP.NET MVC)? Do you fancy a more “light-weight” HTTP framework (e.g. Sinatra, Compojure)? Going further, even those questions have additional sub-questions. There is no “one-size fits all” approach to MVC frameworks for instance, and in some cases the expressiveness of a framework is dependent on a specific language, so you need to pick the one that ultimately suits you best.
Just like languages, some web platforms allow you to be multi-paradigm. For instance, with ASP.NET, you can easily intermingle MVC and WebForms together into a single application. That may or may not be the right option for you, but it’s an option nonetheless. This further drives the point home that arguing WebForms vs. MVC just doesn’t make any sense. Neither Microsoft nor anyone in the community can or should make that decision for you. Why? Because both are equally as good choices, and can ultimately only be chosen based on what makes you and your team more productive. Point being: I believe in being prescriptive, but unfortunately you need to do your own research and make your own decision because ultimately the only deciding factor is your own personal situation/preference (does anyone else hear an echo?).
If you’d like someone else to make your decisions for you, that is fine as well. Just be clear that that is what you’re doing if you ask an open ended question like WebForms vs. MVC or Ajax vs. Silverlight, etc.
Web Frameworks –> Architectural Styles
So why does the title of this post claim that debates about which language and/or framework is the best is pointless? Frankly, because that debate isn’t interesting anymore, especially in the context of the web. The current state of the web is such that it is the compelling ideas that are interesting and how they affect the way we live our lives. If you want to become the next Twitter (who doesn’t?), you should be focusing your attention on how to add value to the overall social network of applications. From a technical perspective, what makes those types of applications possible isn’t a single language or framework but rather an architectural decision that can maximize on the potential for success. Out of those decisions, new languages and frameworks will be born, that will be founded on solving actual problems rather than just academic exercises.
Am I claiming that you shouldn’t discuss how to make a language or framework better? Of course not. Any given day of the year you will find me discussing everything from MongoDB to Grails to OData and so on. Discussion is absolutely necessary, and it is the continual engagement with the different communities that keep us sharp as developers. But, MongoDB as a product isn’t itself interesting. When it becomes interesting is when MongoDB proves to be a great option for enabling a compelling idea. Hence, talking and debating over whether MongoDB is better than CloudDB or SQL Server is pointless outside of the context of actually creating something meaningful.
When we raise the level of conversation from frameworks and languages to architectural styles, does the debate over which is best become any better? Not really. If we’re talking strictly about web applications, there any number of choices here. RIA (e.g. Silverlight, Flash, JavaFx) vs. Server-centric (e.g. ASP.NET WebForms, Struts 2) vs. “pure” Ajax vs. some hybrid approach vs. Whatever is just as boring a discussion to have as languages and frameworks. Why? Because there is no right answer. The point being is exactly the same for architectural decisions as I’ve already stated twice. The best thing we can do as developers is strive to stay as relevant as possible in all areas of our industry. Am I saying you should become a jack of all trades and a master of none? Certainly not, but there is a healthy balance between betting your career on a specific technology and positioning yourself to move forward with the industry as a whole. Let’s take a look at Twitter, and see how they’ve done this beautifully.
Twitter Case Study
If I was to ask you what makes Twitter so successful from a technical perspective, what would you say? Would you talk about its use of Scala for the backend or Rails for the front-end? What about its extensive use of memcached and their rather genius caching strategy? While all of that definitely peaks my nerdometer, it doesn’t have anything to do with why they’re successful. I’d argue that what makes them successful is also what makes the debate over what language, framework, or UI technology the best is pointless: their API.
Let me ask two other questions here:
- Is Twitter a web site? A web application? A web service?
- What is your favorite Twitter client? The website? Seesmic? Tweetdeck? Tweety 2?
Answering #1 would depend entirely on how you view Twitter. You could say that it is a web site, and yes www.twitter.com does exist, but I’d call that their “canonical UI”. What I mean by that is that the website is nothing more than the default client that anyone new to Twitter might use. It is the baseline experience that Twitter personally created to make sure that anyone could access their service. Ultimately though, Twitter’s value is wrapped up in their services and ultimately their data (including users and tweets).
The answer to #1 leads directly into #2. Because Twitter isn’t just a website, and because they’re not partial to who uses their data, that allows us to actually choose a favorite client. Even further, we might have a favorite client for different platforms (e.g. Windows Mobiles, iPhone, Android, Mac). Because of the way that Twitter positioned themselves, they’ve enabled every user on every platform to choose the experience they prefer. That’s a very powerful achievement, and is exactly why it no longer makes sense to argue over what UI or web framework is the best. If you want to achieve large success, it isn’t in the client business, it’s in the services and data business. Whoever has the data has the power.
This is also why it doesn’t make sense to ask whether it’s better to use Scala on the JVM for Android or jQuery and ASP.NET or Flash and ActionScript, etc. I don’t know about you but I’d rather come up with a great idea and then make my data portable so that any of those aforementioned experience would be possible. That’s when your reach and social relevance can be truly compelling. This is also the blueprint being employed by just about every startup worth anything (e.g. Foursquare, BlockChalk). Languages, frameworks, and architectural styles are nothing more than tools in our arsenal that can enable us to create our ideas. We’ll continue creating new ones, and there will continue to never be a right answer in any of the categories.
The discussion over services and data opens a whole other can of worms that I’m not going to go into in this post. Maybe another day 
Have The Conversations
What can you do to contribute to the overall conversation of the web? Be open-minded and be solution-driven. If you need to ask a question about product comparisons, make sure to have some solution context to back it up, so that recommendations can be made based on merit and not pointless debates over which is “better”. Also, be willing to hear people out when they’re talking about some language, framework or style. Good ideas come from all directions, and you can guarantee that what you believe in today will be obsolete tomorrow (not literally, but it almost feels that way sometimes).
Cross-pollinated conferences like CodeMash and MIX are the perfect place for such conversations to occur. When the Ruby folk and Java peeps and .NET group and PHP clan (etc.) get together magic happens. If you see me at one such event, I’d love to chat about what cool stuff you’re doing. Otherwise, if you’re at a conference, talking about cool stuff, hopefully I’ll find you
