Author Archives: Kyle Tucker

Finding A Way

Comparing software development to an apprenticeship like blacksmithing makes a lot of sense. The analogy can work for other professions but I think it works especially well with software development. The reason being that software development compared to other fields offers a lot of freedom and different ways of doing things. In other fields, you often only have a few ways of doing things. For programming there are often many ways to solve problems. Not every solution is good, but it can be used. I think it is a lot black smithing in that way. Many techniques are capable of getting the job done. You can go your whole life without knowing about them, but still be component enough to do good work. In the beginning, you are capable of doing very little or bare bones work. But as you increase your knowledge and experience you become more capable of doing much more. And maybe most importantly create work that can last in the most efficient way. Smithing a sword and building good software are not so different.

One of the most profound patterns to me was the “Exposing Your Ignorance”. Accepting that you are not perfect and there is no way to know everything. Especially, as a beginner it feels like a need to master so many things in such a small time. I feel like I never know enough. I often feel like “how can I get a job like this, I know so little”. Everyone seems to know more than me so it can get discouraging. But I like what this pattern is trying to instill. Accept that you are a beginner and that you will need help. This also goes into the finding mentors pattern. Having someone who can guide or help out of a tight spot. 

I think I like all the patterns about finding your own way. Like the long road pattern. Accepting that maybe you learn in a different way or approach something in a way that best suits you. No one person floors the same path. The software field is so diverse with some many different areas of specialization. There’s no one path that suits the goals you have in mind. Overall the biggest lesson I took away was not criticizing myself for not being a master. Accept that proficiency comes with time and no road is without bumps.

From the blog CS@Worcester – Code Craft by Kyle Tucker and used with permission of the author. All other rights reserved by the author.

This is my intro blog for CS-443

hello

From the blog CS@Worcester – Code Craft by Kyle Tucker and used with permission of the author. All other rights reserved by the author.

Technical Debt As Maintenance

In another podcast from Stack Overflow again by Ben Popper with special guest Johnathan Schneider, discusses the nature of technical debt surrounding open-source projects. The main example used throughout the podcast is refactoring code, switching from java 7 to java 8. And in later examples sometimes java 8 to like java 17. As they explain the difficulty when switching, especially when more manual work is needed, talk about the ways technical debt is created. After more discussion around the software he made to help refactoring, goes into more detail about the nature of technical debt. We think of technical debt as something that we impose on ourselves. Like taking shortcuts or code that needs to be updated. One of the main points of this podcast is to think about technical debt more as maintenance. Obviously sometimes it is self-inflicted, but often it just comes with the territory. As new things come out, things will need to be updated. It’s easier to think of it as something that needs to be done for the sake of the health of the program. Rather than as something I did that I need to fix because of my own short sightedness. 

I chose this podcast because it is related to our discussions of technical debt and refactoring code. In class, we talked about the nature of technical debt and how often it is a code structure thing or maybe a breaking change. It is more thought of as something that comes as a result of not thinking ahead. While I think that is true in many cases. I think a lot of the time it can’t be avoided, like in the case of code versions. Thus it’s better to think of it as maintenance rather than something that could have been avoided. For me, mentally saying that this problem is something that couldn’t have been avoided is better. If I think it’s my fault then I feel like an idiot and get down on myself. Maybe it is just a mental thing, but it goes a long way for the mentality of a person or team. Another aspect of the podcast I think it is important to think about is how the open-source nature plays a part in this. Code is being worked on at all times, different people. So when the code is updated and your code no longer works. It’s important to realize that this isn’t your fault and just is the nature of things. The code not working is not your fault and just needs to be maintained so it works again.

Citations

https://the-stack-overflow-podcast.simplecast.com/episodes/open-source-ecosystem-reduce-tech-debt/transcript

From the blog CS@Worcester – Code Craft by Kyle Tucker and used with permission of the author. All other rights reserved by the author.

The Future Of Technical Writing

In this podcast episode from Stack Overflow by Ben Popper with special guest Ryan Donovan, discusses the evolution of documentation and the future of technical writing. The whole episode is about the nature of writing documentation is changing fast. With the advent of generative AI and other tools like linters that can automatically write or run tests make technical writing more automated now more than ever. One of the first talking points is why documentation is being pushed a lot more now. Increasingly because programmers don’t like writing documentation this job is delegated to technical writers. Who will go into code repositories and write documentation.These technical writers end up being some of the most important people when it comes to code structure. Since they write the documentation that defines how the program is specified and built. 

After moving through some background information the next talking point is the future of technical writing. This was of the most interesting parts to me because it discusses not only just the AI aspect, but things like how other tools play a role. The question then becomes will AI replace technical writers? The answer they concluded was no. Although it is able to generate and automate documentation. Often that documentation needs to be looked over. For the moment AI is not able to create fully functioning documentation to that level yet.

I chose this podcast because it directly relates to what we have been doing in class lately. Looking at the Open API yaml and documentation related files. I don’t think there is a specific lesson to be learned from this podcast. More about just thinking about the future of technical writing. As programmers, we will all have to interact with documentation at some point. Whether that be something like java docs or yaml files. I think it’s important to know what goes into making documentation and the people who write. Because it really is a career all in of itself. Like I mentioned before, one of the most interesting aspects of technical writing is how AI is changing the landscape. Like many areas right now, people are worried if their jobs are in danger. I think with a lot of other cases human supervision is still needed. Another point that I didn’t really think about before now is the importance of documentation other than just information. Like in the case of yaml files that define structure. Is how important it is because it literally specifies how software operates. Like the definitions of endpoints that must be fulfilled and other specifications.

Citations

https://the-stack-overflow-podcast.simplecast.com/episodes/your-docs-are-your-infrastructure/transcript

From the blog CS@Worcester – Code Craft by Kyle Tucker and used with permission of the author. All other rights reserved by the author.

A Simple API

RESTful APIs are a common and standard structure used to develop software applications. Lately in class we have been delving into how these APIs work using Thea’s Pantry as an example. I found this blog that gives a concise and easy to understand explanation of REST works. He gives a brief overview of what REST stands for and then explains its purpose. It used to make data transfer between client and server easier and more encapsulated. One of the points that was interesting to me and made a connection to what we learned in class about coding principles. The principle of encapsulation and keeping things as independent as possible. Trying to keep the server and client as independent as possible to make scalability and maintenance more efficient. The REST API serves as an interface between client servers that serves to communicate for them so they don’t have to.

It then goes into more detail about resources and URLs and return formats. It was also interesting to me that return formats could be more than JSON, the example they used was XML. You just have to specify, but I don’t know a reason why you wouldn’t use JSON. The Reason I chose this blog is because it is a simple explanation of what REST API is. For me personally, APIs can be kind of hard to understand, but this blog explains in a way that makes it easy to understand. The picture was a big help. Other than that I don’t really have anything to say. 

https://medium.com/@MakeComputerScienceGreatAgain/understanding-rest-api-a-comprehensive-guide-52fc10f6c9ed

View at Medium.com
View at Medium.com
View at Medium.com

From the blog CS@Worcester – Code Craft by Kyle Tucker and used with permission of the author. All other rights reserved by the author.

By Any Other Metric

Estimation is a huge part of software development. It exists in many aspects of our work. The area where it is most important is estimation of development time. Many software methodologies are reliant on breaking down work into smaller parts and creating workflows. Like the famous Sprint from Agile. A common theme among all of these is estimating the time it will take to complete these tasks. And on a larger scale, how long it will take to finish a feature or product.

In this blog by Yaakov Ellis, discusses the importance of estimation. Moreover, the skills needed to make a good estimate. He first discusses the usefulness of estimation. How good estimates increase efficiency and reduce crunch time. One of the most important points of the article is the prerequisites needed to make good estimates. Such things like good and detailed specifications and understanding functionality. A point that stuck out to me is that he says that the people who do the work should do the estimates. Seems like something obvious but needed to be brought up nonetheless. There were many other points like not giving in to outside pressure to give an estimate that was unrealistic.

I chose this blog because a lot of what we talked about in class was teams deciding how much time to spend on certain things. Not only sprint lengths, but every part of Agile. Estimating the length of the project to even just a simple bug fix could take. I think it is a skill that is often overlooked. As someone who is bad at giving estimates and gives in to pressure to please people. I think that is something I will think about when making these decisions more. Some last points that I liked was the idea that you should estimate the public so everyone on the team can see them. I think that would decrease the risk of bad deadlines that force crunch on developers. 

Citation

https://stackoverflow.blog/2019/10/23/why-devs-should-like-estimates/ 

From the blog CS@Worcester – Code Craft by Kyle Tucker and used with permission of the author. All other rights reserved by the author.

AI Is Not A Software Engineer

In this blog, the author discusses how much the times have changed for new CS graduates. Reminiscing about how little they knew and how easily they got a job. Then talks about how much more prerequisite knowledge is needed to even sniff a job. The topic of the article is how now more than ever it is easier to get code that works. Thanks to AI, code is now more plentiful than it ever was before. However, all code is not good code. This leads to them discussing how despite how much code there is these days. Having people capable of understanding and able to build software are still very necessary. 

Although AI can now code for us, the coding wasn’t the hard part in the first place. The hard part was building software, and making good software. It’s easy to throw a bunch of code snippets together that accomplish something. But it is something entirely different to build specialized software that fills certain functions and meets certain criteria. AI cannot replace people, even though it may take away some jobs. At its heart, AI cannot build unique software. Teams of capable developers are still needed. The nature of how people code is changing. It’s becoming more important to be able to harness AI, but still oversee and build functional software.

I chose this article because I think it relates to team building. Like the article said, you need people who can understand code, not so much write it. Writing code is easier than ever, but finding people who understand how to build software is harder than ever. When using these tools it’s important not to rely on them too much. Discerning who can actually code these days is probably one of the most important skills for employers these days.  I think it’s important for me and everyone to keep in mind that AI is a tool. Tools dont make up for lack of knowledge. Tools are used best by people who know how to use them and maximize their use. One tool can’t solve every single problem. At the end of the day, knowledge is the most important part of being a software developer. 

Citations

https://stackoverflow.blog/2024/06/10/generative-ai-is-not-going-to-build-your-engineering-team-for-you/

By Charity Majors

From the blog CS@Worcester – Code Craft by Kyle Tucker and used with permission of the author. All other rights reserved by the author.

How AI Tools Separate Us From Information

It is no secret that ChatGPT has blown up recently. It is not just used by CS people, but everyone from all walks of life. It has become a common tool used to help people with a wide range of problems. Offering a quick way to get answers without needing to look for answers by yourself. However, these AI tools are not just a catch all solution for every problem. In this blog from Stack Overflow called “Knowledge-as-a-service: The Future of Community Business Models” discusses how these recent developments have affected how we access information.

In just the last twenty years alone, the way of searching for knowledge has changed. Going from books, to search engines, and cloud technology allowing for farther reach. In recent times we have seen the rise of AI tools that help guide us to the answers we seek. These AI tools however, create a separation between knowledge and the people who make it. AI does the searching and synthesizing for us. Although convenient, it raises the question if that is the best way for people to learn.

Some common concerns held by people are that ChatGPT offers answers. It often does provide context as to why solutions work. What works for one dev environment might not work in another. AI is also reliant on humans for new consumption knowledge. If humans are not creating new knowledge, AI cannot create new information. The credibility of these tools often comes under scrutiny as well. Many developers mention how much variance there is to answers. Although these are certainly draw-backs, developers are learning that community created content is more needed than ever.

I choose this topic because I believe that most students use ChatGPT or some other tool to help us. I myself use it often to help with pretty much every single class I take. But I definitely rely on it the most for CS. I ask how something works or what is the best course of action. I think it is a common concern for many employers cause many don’t know how to actually code. Many people just copy and paste without learning. I am guilty of this myself. But I have been working on trying to actually understand every bit of code. And learning of where and when to apply these code snippets I use. I believe it is still very important to learn from sources outside of chatGPT. Like from classes or other websites composed of trustworthy data. It’s good to learn how to do things yourself without relying on outside sources.

Citations

https://stackoverflow.blog/2024/09/30/knowledge-as-a-service-the-future-of-community-business-models/

By Ryan Polk and Ellen Bradenberger

From the blog CS@Worcester – Code Craft by Kyle Tucker and used with permission of the author. All other rights reserved by the author.

The Troubles Of Learning Codebases

Often when joining a new job, you won’t be writing a new codebase from the ground up. Often we’ll have to familiarize ourselves with the preexisting codebase. Sometimes this codebase is old and needs to be updated or refactored. But how should we go about doing this? It is challenging enough to try to learn a new codebase, let alone trying to change or add to it. That’s what the question the article “Reducing The Learning Curve For Supporting Aging Codebases” by Scott Fitzpatrick aims to answer. The articles offer some dos and don’ts when it comes to trying to make it easier to learn a code base.

One of the first don’ts is to not rely on face-to face conversations and messages/email conversations. While these conversations can be helpful for developers in the moment, they offer little for someone trying to learn the codebase. Plus they can be hard to find and subject to deletion. I think the main point to take away from this article is the importance of good documentation and other resources. When it comes to making a codebase easier to understand, not just for newcomers. Having multiple avenues to try to learn from is important. Beyond just looking at the code, things like data flow and entity-relationship graphs. 

Other aspects beyond code are important too. Like good documentation on local environment set-up and system requirements/dependencies. Making these clear not only makes it easier from the newcomer, but also it helps learn about the frameworks and why they were chosen for the project. Knowing why an application uses certain frameworks and projects can help a new developer understand the reasoning and choices of the team.

Learning a new codebase is very challenging for anyone. I think this article offers good insights as to why certain practices are good and why some are bad. This also once again highlights why I think communication is the most important skill for a software developer. Being able to communicate with other developers not only through code, but through documentation is a very important skill. Making the lives of fellow developers easier by finding as many ways to communicate design and ideas. Although most developers hate writing documentation, it is a necessary evil that must be done.

I will definitely keep this in mind when writing code. Of course writing code that is readable takes precedence. But keeping up to date documentation is also vital to keep projects on track. I personally haven’t worked on a large project involving a large codebase before. So I haven’t had to write documentation for something that I didn’t write. Most of the time I don’t feel the need to write what the function or class does because I made it, so I already know. I think that’s a habit a lot of programmers have, that leads to a lot of lack or bad documentation. It’s something that I don’t have a lot of practice in, so its a skill I’ll have to start honing.

From the blog CS@Worcester – Code Craft by Kyle Tucker and used with permission of the author. All other rights reserved by the author.

Speed Over Design

The following blog I would like to talk about is called “The Hidden Cost of Speed” by Brayden H. Hord. He begins with a quick story about a project he worked on. In this story he describes how he, in an attempt to impress his bosses and meet his co-workers needs, pushed out a product as fast as possible. This worked for the moment. His bosses were happy and he continued his work. However months later, bugs and issues are arriving daily. The software he quickly developed was being used on a daily basis, something he had not anticipated. Now all the shortcuts he had taken earlier had come back to bite him. Now he had something that was being used extensively that was built poorly. 

Now he and his team had to work laboriously to try to fix these fundamental issues. Fixing the problems but also trying to interface between management and stakeholders. The truth is that these problems could have been avoided. If he had better planned and took more time to access the needs and requirements of the project. The moral of the story is that taking time to build right saves headaches down the line. The rest of the blog goes into more detail about why planning and communication are fundamental for all software developers. 

The reason I choose this blog is because I think it highlights one of the most important factors when it comes to software development, Communication. Most software is not built by one person, but rather a team of people. What makes a good team is communication, making sure everyone is on the same page. I think this is important to remember because building without a plan is a recipe for failure. It’s easy to get excited and try to push something that works. But something built on shoddy foundations is always destined to fall. 

Sure your code may work at the moment, but somewhere down the line issues will arise. As needs and more complex architecture is needed, the holes in the code will rear their ugly head. That’s why building code that takes into account not only the needs of now, but also the needs of the future. Building architecture that makes life easier in the future, not harder. I think that this is an important lesson for any software developer to know. Because building something right not only makes your life easier, but everyone else on the team’s lives easier as well.

From the blog CS@Worcester – Code Craft by Kyle Tucker and used with permission of the author. All other rights reserved by the author.