All the vibe coders seem to be programmed to do the same thing. Deploy React apps. Regardless of what you want to do, it seems to spin up in React as the default. You can tell it to use other languages, but I've not had much success with any other language. I'm also not versed in many languages, other than javascript.

WAY back in the day I did BASIC. Then there was that time that I got into PERL... I've also done some PHP, those were the early days of Wordpress and Drupal. Or course there was also that time that I tried to learn REBOL, but couldn't figure out what the hell I was actually going to build with it. Then of course Dartlang came along, and because I'm 100% immersed in the Google ecosystem, I thought it would be fun to learn that.

Long story, long; the only language that I've found any utility in was javascript. So, hate me all you want. I just didn't go down the same path as you.

I've been trying to vibe code a new version of VideoGameGossip(“VGG”) for a long time now... Ever since I bought the domain (along with a basic WordPress setup) off an Indian bloke living in the UK. I actually don't know if he was Indian, there's just that meme that keeps coming to mind.

The Name alone, “VideoGameGossip.com,” was worth the price I paid. So I'm certainly not mad about it. And I got some use out of the setup that he provided, but it just wasn't the site that I wanted to run.

For what it's worth, I do have a certificate in frontend web design, which is about three years old at this point...? I don't remember exactly, but I think it was the beginning of 2023. It was right before Amazon and Google decided to lay off a bunch of people, so the certificate wasn't very valuable when I got it. I was fighting over entry-level positions with senior-level programmers. In any case, I do know a bit about React, even though I'm not a huge fan of it. I also have some AWS and old school LAMP admin under my belt. Though, I prefer Digital Ocean and Google Cloud over AWS, all-day, every day. But I suppose that's the subject of another blog post.

In any case, I'm actually a fan of this vibe coding trend. I just think that it still takes SOME programming knowledge to be able to enjoy it. Also, if you've spent any time messing around with running LLMs locally, that experience helps with the process of understanding what's happening during a vibe coding session. This is true for all tech though, having a better understanding of the tech that you're working with will make you a better user of that tech.

General Vibe Coding Opinions

So, it's disappointing to look at most of the discourse surrounding the use of AI coding assisstants and tools. Unfortunately I only see two types of opinions when it comes to vibe coding:

• It's the best thing ever; OR
• AI is an abomination from hell.

Personally, I think it's in-between. You do benefit from it as long as you understand the code that it's producing. A lot of people will just copy and paste and complain that it doesn't work properly. Even back in 2022 and 23, this was even going on in my frontend bootcamp class, people were already resigned to the idea that Chat was going to take their job. Also, asking Chat to solve their homework assignments for them was much easier than actually figuring out the problems. Sometimes it worked, other times it didn't. I preferred to actually put in the work and do the coding problems. So my opinion of vibe coding is somewhat more nuanced.

No Two Services Are Quite The Same

The original “vibe coded” version of VGG was built using lovable.dev. That service is good, but there's no easy way to deploy your app to a custom domain (or at least there wasn't not that long ago), and as far as backend infrastructure is concerned, the only option is Supabase.

As I'm sure you can probably tell, currently I'm utilizing Google's “AI Studio”, which will then deploy to the Google Cloud ecosystem/infrastructure natively, and almost seamlessly. Not to mention, a simple web application is relatively cheap, the only real expense is storage. The application is so light that I'm only expecting to pay a dollar at the end of the month. Yes, you read that correctly, $1. If for some reason a bunch of people visit the site, I would probably need several thousand visits for it to spin up enough instances to really cause any problems with the site. Most likely what would happen is that it would stop pulling news articles before it cost me any real money.

If I have to build a custom back-end, I could... Golang would probably be my choice. I'd have to spend more time with it, but for what I want to do, I don't think it would be too complex.

Hopefully I'll be back with my post about our time in Mexico... I think it's a decent story. I just felt the need to talk about the stuff I've actively been spending my time on—of the things that aren't work-related.

Also, it's been a while since I've run or hosted a website, so I'm using this opportunity to get up to speed on the current tech. Let me know what you think! I would be happy to hear your thoughts.

• Use the Application
• Look at the Code

Two years ago I signed up for a front-end web development class. Granted, I've had plenty of experience over the years with building websites and whatnot, I've been playing around with that kind of stuff since I was in grammar school. Nonetheless I'm still interested in learning more, especially when it comes advanced programming techniques, new technologies, and different methods for building websites. React was the one popular method that kept coming up. So that's what I decided I should learn. Yet, in over two years, I haven't built a functional website using React (other than the demo project we put together in the boot camp I did). That being said, the project just before my final project was a virtual Tarot deck. This post is both the story of, and the reflection on, that project.

Keep in mind though, it's still a work in progress. I'm still developing it and adding features to it as well as improving the interface. But the current version accomplishes what I wanted; it's a mobile-centric web app that anyone can access.

I originally built Tarot.js with vanilla ES5/6, and I was able to nail down some crucial concepts and core mechanics during that period.

In the past I would have just gone through it myself without the boot camp, but by learning alone, you don't get the feedback needed to REALLY improve.

You may have noticed that the project does not use a React workflow. I understand what React has done for Facebook and other larger enterprises, but it's a very unique use case when you consider the majority of websites out there. MOST websites do not need to use React to function properly. And there is quite a bit that you can do on the web that doesn't require a database or someway for people to login to your service, or even query an API. All of those are useful skills, definitely, but the parameters are always changing. At the end of the day, all modern JavaScript projects are just an abstraction and they generally compile down into vanilla HTML/CSS/JS.

As it turns out, whatever is popular or is currently being used is bound to be replaced by something else even better, sooner rather than later. SO, the important skills that I learned for when it comes to web development (specifically for front-end work) is to show that you can adapt, continually learn, and make definitive decisions about how to proceed with a project. As long as your decisions provide tangible results for you and your team, then that's what truly matter, not the framework or the ideology you're ascribing to. More importantly though, it has been in pursuit of simply completing a project that I've decided to go the development route that I've gone down. It's a mishmash of all the techniques that have caught my attention and interest over the years. All the stuff that I find fun to use and do. It's also a fairly simple workflow.

Although the app is developed solely for mobile use right now, I believe that it can be tailored for a desktop experience in the near future. And I may look into that sooner rather than later, but I was thinking of the TikTok/Shorts/Reels format, and tarot readers on those platforms using the virtual deck... I do ask for a donation for the compiled downloads, but if you're savvy enough, feel free to download the source and compile you're own version. It'll be interesting if I ever find any other people who are interested in a Tweego workflow as much as I am.

[Twine]() was a very deliberate choice... Traditionally, Twine has been used for building interactive fiction. I think a digital book format that is both web-based and has interactive qualities to it will be the format that wins out on a long enough timeline for what it is that I want to create and share with people. So long as the source can be rendered by most web browsers, I think I'm safe.

When I took the time to learn how to publish a physical book and I released my version of Ayn Rand's Anthem, the one thing that stuck with me was that epub3 format is just specialized HTML... As such, the Twine Story format serves the needs of humanity in much the same way, in the short term at least. It is a very robust format for sharing information in general, as we continue to transition to fully digital lifestyles.

Ghetto Classy Swordfish...

Now that the basic functionality of the deck works, I've been focusing on the little white booklet that always comes with your tarot deck. It gives you insight into the card and explains how to use them. The largest portion of the text is essentially done... All of the individual card explanations. However, I am going back through them and editing where I see fit. All of the definitions were taken directly from A.E. Waite's The Pictorial Key to the Tarot (1911), so there is some improvement that can be made to the text for a modern audience. Additionally, I'm incorporating the essay portions from that book as well as portions of Papus' Tarot of the Bohemians (1896) and S.L. MacGregor Mathers' The Tarot (1888).

I think Crowley's magick is still copyrighted, so if you're so inclined to learn that stuff, I would recommend Lon Milo DuQuette's work on the subject of the Thoth Tarot. It's top notch scholarship, and he is an author for Weiser Books, who is the company that holds the copyright to a lot of Crowley's stuff. Nonetheless, I'll endeavor to incorporate Astrological and Kabbalistic attributions where I'm able.

While working on the booklet portion of the app, I'm also going to work out the best way to take screenshots... I'd like a button in the app itself to make it easy for people to snap a picture of their spread and share it online. Additionally, I'd like to overlay some of the spread information on top of the cards at the time the screenshot is taken, or perhaps add another layer of information to the main interface that can be toggled on or off.

My wife said that I shouldn't worry about screenshots because most users will just use the function built into their phones. Regardless, having it built into the app encourages it.

In any case, I've much interested in your opinions. Should you have any trouble accessing the deck online to check it out, or if you have any comments on any part of the program and text, I'm more than happy to hear what you have to say. If it's something that I can fix or improve on, I'll definitely endeavor to do so if I think it will improve the product.