Dependency Management
Pluck It
Edit
#rubygems
#gem-development
#dependency-management
#microservices
#ruby
#ruby-on-rails
#code-reusability
Summarized using AI

Pluck It

Adam Cuppy • November 15, 2015 • San Antonio, TX

The video titled 'Pluck It' features Adam Cuppy discussing the process of extracting micro-libraries and encapsulating them into RubyGems for efficient code management. Key points of the presentation include:

  • Expectation Setting: Cuppy clarifies that the talk will be less interactive than his previous ones and focuses on gem publishing, aiming at those new to the topic.
  • Introduction: He introduces himself as a principal at a web consultancy, highlighting the topic's relevance to improving development processes in Ruby applications.
  • Defining Micro-libraries: Cuppy explains the concept of micro-libraries and how they can improve projects by encapsulating small pieces of functionality, particularly emphasizing the Single Responsibility Principle to avoid bloated code.
  • Identifying Opportunities: He suggests looking into Rails app components, like models and controllers, for duplicated or application-agnostic code, which can be effectively transformed into gems.
  • Creating a RubyGem: The speaker guides listeners through the gem creation process using the command line, starting with bundle gem, which sets up necessary file structures. He discusses the importance of including tests and proper structuring in the gem’s code.
  • Publishing and Managing Dependencies: Cuppy discusses publishing gems to RubyGems.org and the role of Bundler in managing dependencies, contrasting it with RubyGems' purpose as a package manager.
  • Advantages and Disadvantages: He outlines benefits such as easier maintenance and better reuse of code, as well as potential challenges, including managing updates and additional overhead.
  • Illustrative Examples: Cuppy provides examples of successful small gems, such as Truman and Gravis, demonstrating the practical value of creating even simple libraries.
  • Conclusion and Resources: The presentation concludes with reminders to utilize community resources for support with RubyGems and an invitation for questions, promoting further engagement beyond the talk.

The key takeaways include the importance of extracting functionality into micro-libraries when sustainable, the straightforward process of gem creation and publishing, and the benefits of community engagement to aid in Ruby development.

Pluck It
Adam Cuppy • November 15, 2015 • San Antonio, TX

Pluck It by Adam Cuppy

How many times have you written the same bits of code, over and over, and thought, “You know, if only this was big enough to be a gem, I would pluck it out." Often, we think of a RubyGem as a larger library of code that we “bolt on” to an app. And, these smaller code blobs become a hassle to distribute to the multiple apps that use them.

A small micro-library, done the right way, at the right time, can greatly improve an app.

But, when can you benefit from extracting a micro-library? And, how do you build and publish that code into a RubyGem? I'll go through the process, from A to Z.

RubyConf 2015

00:00:11.990 Before we get too far, I just want to take the opportunity to set some expectations. If you've been to one of my talks before, you know they tend to be quite interactive. However, this talk will not be as interactive. I had sent out a tweet making this clear, as I wanted to give you a primer on what to expect. This talk will essentially be a walkthrough on how to extract small libraries, encapsulate them into RubyGems, and publish them. If you feel that this topic is still suitable for you, and you want to learn about gem publishing, then this will be perfect for you. But if you've done a lot of gem publishing before, this may be a bit too cursory for you. I completely understand if you decide that this isn't what you want; I'm not offended at all.
00:01:14.100 My name is Adam Cuppy, and I am a principal at a web and mobile consultancy called Zeal, based in Oregon. We specialize in many of the topics I’ll be discussing today. If any companies are interested in training or development of some kind, we would love to work with you. If not, that’s totally fine too! You can find my Twitter profile here, and I welcome interactions there. I love working with people and helping to fill in educational gaps. Another thing I want to ask is that since we are quite spread out, please come forward. There’s plenty of space here, and I would appreciate if you all could fill in the front. Don't be shy; we are all friends here!
00:02:18.600 Everything you see today will be published online, so feel free not to take copious notes. It’s actually easier for presenters to see your faces rather than just Apple logos. Just keep in mind that everything is going to be available online, including this talk, which is recorded. If you have any questions during the talk, just keep a mental note of them. I am more than happy to answer your questions. I’m on GitHub as a Cuppy. All the code you’re about to see has been published either under my name or under coding zeal. My Twitter profile is there as well; feel free to tweet at me!
00:03:10.500 This presentation is affectionately titled 'Pluck It,' focused on extracting micro-libraries and building RubyGems. This track revolves around the theme of less code and is based on our experiences at the consultancy, which I've been part of as a Ruby engineer. We realized that there are many fantastic opportunities to create gems that are less than a hundred lines of code; they don’t need to be large at all, and doing so can be quite beneficial for various reasons, which I will discuss later.
00:03:32.699 To start, we will look at a basic Rails app. I know we’re at a Ruby conference, so please forgive me. You can throw food at me later. The fact remains, even though I’m utilizing a Rails app for this example; you could also use a Sinatra app or Lotus, which is a cool framework. In fact, it could work with any Ruby library that can leverage RubyGems. For this talk, I’m using a Rails app to illustrate the point.
00:04:01.259 Most Rails apps will have a models directory where we find the source material for extracting micro-libraries. Micro-libraries often center around services; they might be some sort of factory or builder code. Another place to look is in your controllers, especially inside the controller code where you handle interactions. There’s a good chance you’ll find a lot of duplicated code there too, providing a big opportunity to extract micro-libraries from that.
00:04:43.380 For instance, if we look at a basic User model, we see some functionality baked into it. There is an attribute on the user model that pulls in a Gravatar URL. To do this, we have a constant that specifies a URL template based on the Gravatar documentation. We also have a public method called gravatar_url that builds the URL and returns it. There’s a private method on the user object that creates a hash hex value for the user’s email address. This implementation is quite basic, but it highlights a larger issue. How many of you are familiar with the Single Responsibility Principle?
00:05:55.349 This is where we run into problems. Essentially, we are placing functionality into an object that doesn’t care about how the URL is built; it just needs to know that it exists. Therefore, it’s common to extract this code and put it into its own class. We could create a new class in the lib directory called Gravatar and move the functionality from the user object into this new class. This is a simple process and allows for additional features, such as supporting different sizes of Gravatar images.
00:06:41.190 Going back to our User model, we can now start reusing the code by calling the Gravatar library we created. This is a good and straightforward fix and is frequently done in Rails applications. When you have code or functionality that is application-agnostic or domain-agnostic, it’s a good idea to extract it so that you can manage it more effectively and create a better testing environment, free from interference with the user model.
00:07:33.240 This leads us to a vital question: How many times have you written the same small bits of code over and over, thinking, "If only this were significant enough to be a gem!"? We often find ourselves creating single-file gems. You know what I'm talking about; one file we can reference back to in different applications. Or we might just pass it around, sharing the code without telling anyone. Such practices are not sustainable, especially when it comes time for updates. For example, if you’ve written code now in production, and you lose access to that codebase, what do you do if you find a bug?
00:08:32.760 This frequent occurrence makes it essential to consider RubyGems. Most of us familiar with Ruby know that every project that uses external libraries employs RubyGems to manage those dependencies. To put it simply, a Ruby gem is a packaged collection of files, or one or more files bundled together, that can be utilized across projects.
00:09:44.010 It’s important to note that while RubyGems helps manage packages, Bundler is the tool we use to manage dependencies. Each Ruby gem has a version, and Bundler helps you handle all the intricacies of dependencies to ensure your application runs smoothly with the necessary library versions. Prior to Bundler’s widespread use, handling dependencies was an absolute nightmare. Now it simplifies things immensely.
00:10:39.180 When we run our application's Bundler command, we can see a simple output that shows which gems are being used. Bundler retrieves these gems from RubyGems.org or potentially a private gem server, if applicable. This way, we can manage our application's dependencies very effectively.
00:11:49.310 Now, it's essential to understand that while we often think of RubyGems and Bundler as interlinked, they serve different purposes. RubyGems is the package manager while Bundler focuses on managing dependencies. It is not uncommon to see variations in the version of a gem and even a lack of awareness of how things work beneath the surface.
00:12:46.760 As a quick overview, generating a gem with Bundler is straightforward. You run the command `bundle gem` followed by the name of your gem. This generates several crucial files: a `lib` directory, the main gem file, a version file, README and LICENSE files, and some useful configuration files for the build process. These files will serve as the foundation for your gem.
00:13:41.050 The generated code is quite handy; the `lib` directory includes a file to require the version and sets up a module for our gem. This is an important practice to avoid conflicts with any other gems and to maintain proper organization. Within that `lib` directory, we can create our specific classes and functionality.
00:14:38.000 After structuring our gem, the next step is to make sure we correctly implement any necessary tests. There's a generated `spec` directory as part of the bundling process, where you will create your tests to verify that everything functions as expected. Once this is done, you would run the test command to ensure that everything passes.
00:15:42.960 After your tests are in place and passing, you need to modify the `gem_spec` file to set the gem's properties. This includes the name, version, author details, and descriptions which will help others understand what your gem does. By encapsulating all essential information within the gem itself, you reduce chances of information becoming out of sync with external sources.
00:16:56.300 Finally, once you're satisfied that your gem is working, you build the gem by running `gem build` followed by your gem's specification file. This generates a `.gem` file that you can push to RubyGems, making it available for use. Once published, it becomes a resource that others can access and utilize.
00:18:08.050 There are many advantages and disadvantages to creating RubyGems. Some benefits include managing dependencies more effectively and the ability to isolate your code from your application. This last point is especially useful as it allows you to handle updates and changes independently. However, it introduces additional overhead and requires ongoing maintenance. It can complicate the process of making changes because you’ll no longer be able to directly modify the code.
00:19:12.599 In summary, extracting functionality into gems is most beneficial when the cost of maintaining internal libraries exceeds the cost of maintaining external dependencies. If your code is widely reused across different applications, it can often make sense to pull it out into a gem. This can vastly improve your team’s development workflow and the maintainability of your projects.
00:20:21.579 To illustrate this concept, consider small gems like Truman, which is a simple truth value manager created with only 30 lines of code. It's a small utility but its value is evident. Another example is Gravis, a small library that was extracted for generating Gravatar hashes, which showcases how even tiny libraries can be successfully published as gems.
00:21:19.130 As we approach the end of this talk, remember that there are vast resources available to help you with RubyGems, including a detailed guide on rubygems.org. If you have further questions, don’t hesitate to reach out through Twitter or GitHub. I truly value community engagement and would love to assist anyone interested in further exploring RubyGems.
00:22:14.910 Thank you all for attending! Are there any questions before we wrap up? Yes, I recommend various services for private gems. There are some that offer authentication, which is essential. If you need an ad-hoc solution for private gems, you might consider mounting the gem from a local directory incorporated into your application. But it is often better to use a dedicated gem server that allows for authentication. Thank you once again for being here, and I hope you have a great rest of the conference!
Adam Cuppy
@acuppy
Explore all talks recorded at RubyConf 2015
+76