#web stack

Introduction: Web development has come a long way since its inception, and it continues to evolve at a rapid pace. As technology advances and user expectations grow, developers need to adapt and embrace new skills to stay relevant in the industry. One such skill set that has gained significant importance is full stack development.

The Rise of Full Stack Development: In the early days of web development, there was a clear divide between front-end and back-end development. However, with the increasing complexity of web applications, the need for developers who can handle both sides of the development process became evident. Full stack development emerged as a solution, combining expertise in both front-end and back-end technologies.

Seamless Integration and Improved Efficiency: Full stack developers possess the skills to seamlessly integrate different components of a web application. They can handle database management, server-side logic, and user interfaces, ensuring a smooth user experience. By having a holistic understanding of the entire development stack, full stack developers can streamline processes, reduce dependencies, and improve overall efficiency.

Adaptability in a Fast-Changing Industry: The web development industry is constantly evolving with new frameworks, technologies, and trends. Full stack developers are equipped to adapt to these changes more effectively. They have the flexibility to switch between different technologies and frameworks, allowing them to stay up to date with the latest advancements and meet the demands of modern web development.

Versatility and Collaboration: Full stack developers have the advantage of being versatile professionals. They can contribute to different aspects of a project, collaborate with team members from different disciplines, and understand the bigger picture of the application’s development. This versatility not only enhances their problem-solving skills but also fosters effective collaboration within development teams.

The Future of Web Development: As web applications become more complex and integrated, the demand for full stack developers will continue to rise. Employers are increasingly seeking professionals who can handle end-to-end development and bridge the gap between front-end and back-end teams. By acquiring full stack skills, developers can position themselves for exciting career opportunities and be well-prepared for the future of web development.

The landscape of web development is evolving, and full stack skills have become essential for developers to thrive in the industry. The ability to work on both front-end and back-end technologies, seamless integration of components, adaptability to change, versatility, and collaboration are all critical aspects that full stack developers bring to the table. By embracing full stack skills, developers can stay ahead of the curve, unlock new opportunities, and contribute to the ever-changing world of web development.

For processes not already integrated into systemd the two main alternative managers seem to be Circus and supervisord. The main advantage of circus appears to be its ability to handle sockets.

Asides for a six-month break before FIrefox 4, I've been using Firefox for over a decade now (and for six months on my phone, too). I appreciate how their breaking of IE's stranglehold make the modern web possible. I have 25 extensions running right now, at least ten depend on the ability for Firefox extensions to do whatever the heck they want.

More germane here is Mozilla, the parent organization of Firefox. I love Mozilla's commitment to an Internet free of censorship and walled gardens. Unlike Google and Chrome,  is what Mozilla is about - as the Open-Source Evangelist at Mozilla once said, "we just make browsers".

The problem is, when you only make browsers, you can easily convince yourself that they can do whatever you want it to. Mozilla is beginning to treat Javascript as a first-class language and the Web stack as a jack-of-all trades platform.

The most prominent example is with their new phone operating system, Firefox OS, which runs this stack top to bottom - including for basic things like the user interface and making phone calls. But the list doesn't end there. They're also at work on:

Writing a Flash interpreter entirely in Javascript

A video decoder written entirely in Javascript

Writing compliers and optimized versions of Javascript so that it can be treated as a bytecode language

I'm fine with this vision, but the Web stack is just not mature enough to be used like this yet, and trying to work within its current constraints is distracting Mozilla from actually making this first-class future possible.

The first problem is that, Javascript is not a good first-class language yet. I know, you can write anything in any language - heck, people are writing incredible things in Javascript now. But its idiosyncratic OOP model, the lack of typing, the inability to easily import other namespaces (resulting to incomplete injection hacks) may not be obstacles to beginners or experts, but place serious hurdles for intermediate programmers to write robust code. Help is coming with ECMAScript 6, but the flip side of consensual standard writing is that this is happening way too slowly - people were commenting on hold-ups five years ago. There's a reason some people swear by CoffeeScript and TypeScript.

Moving up, Javascript is too tied down by the browser. One example of this is language design. One simple way is that some key language things (like events) assume the existence of an actual HTML document, which you can't create within Javascript itself. Sure, once you have a blank HTML document, you can populate it with a Javascript representation of the HTML, but a) you can't start a document this way, and b) it's just so rare.

Most importantly, there's just the complete inability to access underlying hardware features. This one is slowly changing, but without larger change, underlying access will always be limited by the binary nature of access in the Web stack - every webpage has sandbox-mediated access to some feature, or it doesn't. Hence why we can't get anything cool like arbitrary file writes, even for apps written entirely client-side. 

This is also probably the reason that Javascript performance is so slow. Tens of thousands of man-hours, across several corporations and over a decade, have been pushing Javascript to its absolute limit, but try (as I have) running something like a long regular expression. C#: 5ms. Chrome: 200ms. Firefox: forget it.

Firefox is, at most, chipping away at these problems. Firefox OS needs lots of APIs to the underlying hardware, as you would expect from a phone. So they made up a few dozen, which will solve their specific problem (after these APIs spend years in standards-setting purgatory). But this system trails technological change too badly to be an all-purpose language. What happens when someone wants to use Javascript to power an robot, or if phones get new features?

This problem also hampers Firefox's aforementioned effort to tackle the speed and language-weakness problem, the Emscripten compiler and asm.js Javascript framework. Yes, getting close-to-native performance in a browser is important, but native apps are appealing for reasons besides speed. Chief among them is that you can access new and powerful hardware features at will. (And as Ars noted on a related point recently, this level of hardware access is also important for speed.)

(Not that I like asm.js so much anyways. it requires explicit support by your browser, and your resulting "Javascript" is so constrained and obfuscated that it's really its own programming language already - a virtual machine, but without the original code sitting on top of it anymore to help someone debug any problems. But that's just me hating.)

So that's why I think Mozilla's jumping the gun. Instead, Mozilla should shift resources to some important projects on their back burner that could solve all these problems.

Probably the farthest one along is Servo and Parallel Javascript, their efforts to bring parallelism (and thus also better tab/window isolation) and power efficiency to the Web stack. Right now, they're leaning on Samsung to get this done, which is a shame. Getting this done would make all computers get a 2-8x possible performance increase, make Javascript scalable if Mozilla ever does succeed in making it a first-class language, and let it retake the Web innovation crown from Chrome to boot. (While I'm at it, I'm also a fan of sweet.js - it could provide some much-needed syntactic sugar to the problems I suggested above while we wait for ECMAScript 6.)

Amazing overview of all the different pieces that might go together to build your web stack these days. Highlights what I have been saying recently that you need to plan for multiple technologies.