Back in June, Team Gu released our very first visual novel. It was exhilarating to bring ‘Love & Country’ to life, but absolutely nerve wracking to release it into the wild. Now that it’s lived and breathed on the internet for three months, it seemed high time for us to conduct a post-mortem.

There were many decisions throughout the creative process that had further reaching consequences than we realized. This project required us to experiment and to learn brand new skills. Sometimes it paid off. Sometimes it derailed everything. But we muddled through.

If you’ve read our prior devlogs, you may already know about some of the obstacles we encountered. Here we’ll try to focus on the bumps and obstacles that we haven’t written about before. If you are a dev or or an aspiring dev looking to learn from our mistakes, or just curious about what we are thinking about doing differently next time, read on: 

Scope

Scope issues were at the root of so many subsequent hurdles.

The original conception for Love & Country was very, VERY different from what we ultimately created. In the earliest stages, we had discussed a user-created MC who could be from any number of different countries with half a dozen love interests, each with their own preferences. Obviously, that is nothing like our final product.

We backed away from the user-created MC idea pretty quickly in favor of three well-defined character options: a soldier on the Western front, a nurse on the Eastern front, and a spy on the homefront. That last one eventually became Lillian. In the process of fleshing out rough outlines for each character’s story, we realized that each story was a game unto itself. Narrowing the scope was necessary. 

Although we had four missions planned for Lillian, it took us about a year to write the script for the first one (There were many, many rewrites, significant changes to characters, and entire plot lines discarded. Many darlings to be killed). Furthermore, we had taken a story-first approach to script, meaning we hadn’t set any limitations to our writing or plot choices based on the number of art assets that would be required. That resulted in 47 background images and a dozen character sprites with multiple outfits and expressions each. If we had outsourced any of the artwork, this would have caused serious budgeting problems. Since Juliana created all of the artwork herself in-house, we luckily avoided that pitfall. Still, it cost us in time.  

Ultimately, we decided that Lillian’s trip to Lӧrrach was as much content as we could handle for our debut game. It was the correct choice in terms of making the project manageable, taking care of ourselves mentally and physically while still completing the project in a reasonable amount of time. 

But this decision came with some costs. Many players (correctly) sensed that while the mission was a contained and complete story, it did not complete Lillian, Émile, or Anton’s character arcs. It didn’t make narrative sense for Lillian to consummate her relationship with either love interest on this first mission. We have ideas for how the story continues, but we wanted to write the story in a way where it could stand on its own. And considering the story of Love & Country in isolation, we understand why some players were disappointed in the lack of a Happily Ever After.  

Another creative decision worth discussing in the context of scope is the limitation of choices within the story. It was very important to us that choices MATTER. Rather than writing choices that only added flavor, we created choices that sent the story branching in completely different directions. This does maximize replayability in that the player gets completely new content with each choice, but we had to limit the number of choices we had to keep the scope from spiraling. Three choices gave the player four very different routes. This resulted in Love & Country being more “novel” than “game”. 

What will we do differently next time? We’re looking at how to incorporate more choices into future projects while still keeping those choices meaningful to the plot. Also, since we are a little more experienced now, committing to a longer script feels daunting, but a little more realistic than it would have been the first time around.    

Budget

As you may know from our previous posts, Team Gu is actually very small. This is mostly a function of our lack of budget. Katie writes and programs, Juliana writes and draws, and Zachary composes the music. We were lucky enough to have friends who were willing to read our script and give us feedback. Anything we couldn’t do ourselves we either had to learn to do or do without. 

The one exception to this DIY mentality was the translations. There is a lot of French and German in Love & Country. While Team Gu speaks a little of both, we are by no means fluent. We started by checking ourselves against Google, but ultimately paid translators to review our work. AI can do a lot, but it doesn’t replace people, and it doesn’t replicate natural sounding speech with slang. After researching what it would cost, we opted to pay for this service out of pocket. 

We also spent a small amount of money on software and business expenses. We purchased a domain for our website, registered our business with the state, etc. Steam charges a fee to sell via their platform. All of these costs are relatively small, but add up. Unfortunately, no amount of DIY can replace some of these costs. All of these were covered out of pocket, too.

One feature we wish we could have added but couldn’t afford was voice acting. We did hire actors to read a handful of lines for our game trailer. It was such an amazing experience hearing the characters brought to life in this way. The cost of fully voicing the game far exceeded what we could afford without investors or doing a Kickstarter. As you may know, Kickstarter projects are a whole THING.

The one silver lining of the “no budget” approach is that it did grant us some creative freedoms that we might not have otherwise had. There were times when we’d look at something we’d done and realize it wasn’t working. If we’d been on a budget, we’d have needed to weigh the cost of trying alternatives or re-commissioning specific assets. Instead, we were able to weigh the decisions only against our own time and effort.   

Art

Hi, Juliana here with some insight on the art of Love & Country! 

If you've been paying attention to us on the internet (even before the demo was out–if you’re reading this and that’s you, that’s wild, thank you for being here) you’ll have seen that we posted some character sprites from the early days of development. I started off with the idea of a muted, almost sepia-toned color palette, to give it that old-timey feel, but after a while I found it visually boring. The style was also more cartoony, and would sort of continue to be so as I went on…until I started working on backgrounds for the game.

Sigh. Here it goes. Ahem. 

I am in no way a background artist. In fact, as an illustrator, I actively try to avoid them. I find that one of my greatest pitfalls as an artist is that my will to create a composed, scenic, full illustration is almost none. I draw people, that’s my strength. So imagine my frustration trying to draw backgrounds for the first time, having no idea what the rules are or what my vision should be. Because of this, it led me in a direction that felt sort of utilitarian. You need a car? Blam: car. Need a bedroom? There ya go. And it all turned more painterly than I had intended. Which…clashed with the style of the sprites. Thus, with the characters, we ended up with what I hope is a happy medium–same slight cartoonish flair, but fully rendered. All in all, I’m proud of the backgrounds I cooked up (even though some are not perfect DON’T LOOK TOO HARD), because they led me to the eventual aesthetic feel of the game.

Another thing I want to share, related to what Katie mentioned about L&C’s beginnings, is the doodles I did that inspired the idea of a WW1 spy romance story to begin with.

I drew a duo–a Russian spy and a British soldier. I had no plans or story, just characters. You can see the elements of Émile in “Alfie” here. As it turns out, I just like tired-eyed smoker boys. And as for “Vita”...well, you might see her in a Guhuhu future project. Who knows!

Anyway, that’s about it. I’m excited for more Gu work and I thank you all for your wonderful comments on the art. It really warms my heart when I see that someone resonated with the visual world I created, because oftentimes you feel like what you’re creating won’t be appreciated. But that’s just the plight of being an artist! Onwards!

Music

From Zach– Now for the music’s turn! I’m generally pretty happy with how the music in the game came out, but there was one big mistake I made that I’m going to rectify for all games I work on going forward. There’s a lot of music in this game that I’m very proud of. I believe the sonic palette is effective and fairly well explored, and I’m happy with the way ideas flow from one to the next; less so at the beginning of the game where I was still figuring out the system, but over time I feel I found a satisfying way of tying together all of the themes in a cohesive way. The mixing can be a little rough, but I’ll give myself some grace that it’s totally listenable and I improved a remarkable amount over the course of the game’s development. The real problem is: there is way too much damn music in this thing. We became very attached to the idea that every single route would be entirely unique as far as music goes. Certain things may be similar, sure, but any like material would be adapted to fit that specific situation. This comes from a good place, but it was entirely impractical and altogether unnecessary. What I should’ve done was employ a sort of hybrid system: have some pieces of music that follow the action very closely as needed, but also keep a few tracks in reserve that can loop ad nauseam for sections that are similar. Does every little conversation between Lillian and Emile need bespoke music if they’re mostly just trading barbs? Does every single spy sequence need to move at its own pace if characters are sneaking around the same place under the same circumstances? Probably not. 

I put in a remarkable amount of sweat work when what was really called for was efficiency. I am going to be far more economic with all future projects, as writing this much is unsustainable. The gains are meager enough that time would be better spent refining those recurring tracks, rather than making variations so slight that I’m the only one who notices the difference. Also, no more prepared piano (outside of Love & Country). I can’t begin to express to you how much of a pain in the ass it is.

Technical Execution

Katie again– Like everything else about Love & Country, the technical execution started in one place and ended in another. The very first test we did was in TyranoBuilder. While Tyrano is an excellent tool for those without coding experience, it’s not sophisticated enough to accomplish what we had in mind. 

We ended up switching to Renpy, which required that I learn to code but had the benefit of robust documentation and a very active online community, making it more beginner friendly than other engines. However, there was still a learning curve to contend with. A lack of familiarity with Renpy as a tool also made it more difficult to plan scenes. Every effect and feature required research.

It’s also worth noting how difficult it was to implement audio in Renpy. Renpy doesn’t support the use of middleware and the commands for audio are limited. Timing audio across various channels was often cumbersome. For this reason, I will likely switch to another engine at some point. However, for the next project, my intention is to stick with Renpy. Rather than focusing on learning another coding language and interface, I want to build upon what I learned in the last project.  Hopefully by choreographing some of the more cinematic moments, I’ll be able to pull off more interesting special effects and camera work.

Marketing

Marketing was the greatest challenge of all for us. Part of the problem was our own knowledge gap. No one in Team Gu has expertise in marketing. As with many aspects of the project, we had to teach ourselves as we went. Even with the best of intentions though, our execution was far from perfect. We were not nearly as active on socials as we should have been in the months leading up to release. We also did not correctly estimate development time for the game, meaning we were still coding when we should have been done and focused on marketing.   

One of our biggest obstacles though was determining how to pitch the game. Beyond “visual novel” how else do you describe this thing we made? It’s otome adjacent. There is romance, but it is not really a dating sim. There is a potential love triangle between Lillian, Anton, and Émile, but it doesn’t play out as a binary choice for the player. How do we classify the genre? Is there enough focus on the mission to truly call it a “spy thriller”?  Is it fair to call it a “romance” without a traditionally happy ending? Were there enough choices to consider our story to be “interactive fiction”?

Honestly, we’re still asking ourselves those questions. For any label you choose, there will always be reasons why it doesn’t fit perfectly. We’re definitely going to focus on posting consistently across social and doing a better job will be the marketing timeline for our next project. And as to the philosophical debate around how to best describe Love & Country, we’ll probably still be having it.   

Conclusion

As we move into our next project, we’ll definitely carry some of these lessons forward. Planning will be a vital component in avoiding many of the issues that we encountered. We’ll be actively looking for ways to simplify and streamline as much as we can.   

We hope you will continue to support us by supporting Love & Country. Reviews and comments matter a lot to small outfits like ours. And we do read them all! If you have any questions for us, don’t hesitate to drop them in the comments.

Sincerely,

Team Gu <3