What visual novel development taught me

Today, I want to come back to the differents projects the team made in order to comment our progression.

Indeed, even if things are slowly changing with the increase of early access games, one often only get a hold of a game through its finale version, in other words, it’s most complete and most accomplished version. An average player may get under the impression that the differents pieces are coming together thanks to some kind of evolution stone to change into a piece of art. Reality is less glamorous but way more reassuring : creating is something you have to learn. The bad news is that it involves a lot of work, the good news is that anyone can try it out. And the Träumendes Mädchen team is no exception, especially me who has the job of being the “leader”. It’s been almost three years now and I’ve learned so much thanks to my teammates, particularly with programming. So I want to draw up some kind of report.


June 2012 : Being Beauteous

Some of the first drafts by Morsy. She improved a lot since then!

Back then, the team had already gathered around Milk and we floundered so hard. I was aware we were jumping the gun when we started with a project this big and I really wanted to make up for that with something short. The first version of Being Beauteous was born like that and we planned to make it during some French contest that happened in a small convention. Even if it didn’t looked much, BB hold a special place in my heart because it was our first completed project. It was really simple (text and some CGs) but we did have troubles during the competition because of the GUI. So it’s only with the 2nd version (the first available one actually) that we managed to fix those bugs. Also, we slowly realized that having a translation was really important to share our works to folks around the world.

What I learned : to finish a project, to make a bug-less GUI, to understand the importance of an English translation.


February 2013 : 1st Episode of Milk

Another draft by Morsy. In the end, those variations never got into the game.

The deliivery of Milk’s 1st Espidode was a real ordeal because it took us like a whole year to see the end of the tunnel. It must be said that there was a lot of content to deal with for such a young team and I immediately understood that we couldn’t go on like that. After all, we were supposed to release the whole game in a row after that! I realized it was impossible and change the plan to an episodic publication. Hence why the demo became “1st episode”. It was also a chance to get to know Renpy a bit better (choosing a custom font, juggle with all the numerous sprites and numerous expressions). Milk’s presentation was really simple and we made a lot of mistakes but I think it was a prerequisite for what came after. Most of the mistakes were corrected (the GUI) or are pending for correction since then (*cough* the scenario). It should be noted that we released the visual novel both in French and English at that time, not without pain.

What I learned : to write a visual novel, to finish a bigger project (without dying), to draw a GUI, to make a siultaneous translation, to manage sprites and expressions.


April 2013 : Ambre

I think we all understand why Melow did the game illustrations and not me 8)

Once Milk’s Episode was behind us, I had to tackle Nanoreno! It wasn’t a technical challenge (because the story was linear)  but it sure was a litterary one. Granted the deadline was a pretty difficult thing to handle too but I planned the game not to need too much assets. But really, my personal challenge was to write a short story that would inlude some kind of memorable ending, something striking. Even thought the result wasn’t perfect, I’m relatively proud of the outcome, I think the intended result is here. Ambre was also a good chance to make progress on GUI and animation. We included some light animation for the first time with the petals on the main scene and we tried something with the leafs on the city background. Some elements you would find again in the next release.

What I learned : to write a visual novel while paying attention to the assets needed, to finish a project quickly, to write a story with a memorable ending, to join an event.


March 2014 : 2nd Episode of Milk

A pretty sketch by Elairin

We took almost a year on this one again but it was mostly because of the turn-over inside the team. This second episode was mostly a validation of what we made with the first one. Indeed, many project don’t get past the demo stade, so releasing a new part of the story was a symbolic move. We wanted to show we were still game and that we wouldn’t give up so easily. We took that opportunity to change the GUI (and it was way more difficult than you would imagine!) and include animated backgrounds for the first time. Well, there are only 2 or 3 of them but it was a way to slowly get accustomed to the process.

What I learned : to carry on an under way project, to include animated backgrounds, to renew a team.


April 2014 : HVNCML

First GUI tests by Xian

Once again, I wanted to tackle Nanoreno, and once again Milk’s production was late. In 2013 we managed to finish two weeks before the beginning of the contest, this time we finished in the middle of it! Under these circumstances, it was rather difficult to deliver. HVNCML was born a whole week after the deadline. I didn’t write any postmortem because the whole game can be considered one (with a comic point of view). Back then, it was our first technical challenge because I only wanted to try my hand at a IRC-line GUI and didn’t have much planned to use it afterwards. Well, you can consider that making some part of your life public was also a challenge. In a way…it is. This little visual novel was supposed to be a funny extra to the fans that were wondering what we were doing everyday and how our meetings looked like. Yet several people who didn’t know us played it too. It was kind of embarrasing, a bit like getting naked in front of strangers. The whole approach does suppose some trust with the audience.

What I learned : to juggle between several projects at the same time, to try an IRC-like GUI, to reveal my stupidity to the world.


July 2014/September 2014 : 3rd Episode of Milk

Orties trademark: writing nonsense on her drawings :p

Milk’s third episode is really seen as an exception among our works because it’s the only damn game that went “easily”. Meaning that we didn’t enconter any particular issue (except a small false note with the translation that wasn’t immediately ready). If anything could happen like that, visual novel dev would be so easy XD. It’s all the more surprising considering we had a lot of pressure due to the fact we promised this part would be ready for Japan Expo, a French convention. The third episode also marked the spread of backgrounds animation.

What I learned : to meet a non-negotiable deadline, to spread backgrounds animation, to release a game in a relaxed way.


August 2014 : Garden of Oblivion

A relatively accurate Ludum Dare metaphor…

In the same vein as HVNCML, Garden of Oblivion was planned as a technical challenge: I wanted to put tradionnal visual novel aside to include some point & click elements instead. With an incredibly stressful deadline (Nanoreno is a stroll in a park in omparison). If the graphic part was irreproachable, the coding part was rather…wild. The first version we returned didn’t match my expectations at all and we had to add a whole week of work to get to the result with the second version. The other challenge was for me to write directly in English with the bare minimum of preparation to fit the gamejam spirit. And I’m reserved about the final result, I could have done better. At least, I managed to include more diversity among the characters. The fact remains that the community experience was on par! Ludum Dare really is a big gathering of people and that forms bonds. If you wanna know more, read the postmortem I made right after the competition.

What I learned : to write directly in English, to jon a bigger event, to learn programming in the field, to make an hybrid VN/P&C, to discover the joy of polishing, to include a bit more diversity in a cast.


Spring 2015 (?) : 4th Episode of Milk

Orties really likes to write on her drawings…

Milk’s forth episode isn’t out yet but I feel I learned so much already. I think it’s starting from this point that I really felt confortable with RenPy. To the point that I now like to rummage through the script to improve what I can improve. It wouldn’t have happened without slowly getting familiar with the engine, project after project, and the help of Keul. The direct consequence is that this episode is full of small alterations: adjusting the brightness of sprites during night scenes, adjusting the height of non-human sprites, adjusting the font to be more readable, adjusting some transitions to be more dynamic, adjusting sound effect to fit the mood better, light animations on the illustrations. Currently, that’s the work I’m the most proud about and I hope those little things will make the read more interesting, even if they’re not that useful. I also started asking the artists to try different presentation styles (object on top of the background, insert chibi, CG cut in half, CG with many variations).

What I learned : to have a good command of RenPy, to use all my time doing alterations nobody will ever notice, to use sound effects effectively, to experiment different presentations.


April 2015 (?) : Ludum Dare 2

Since the game’s production hasn’t started, for logic reason, here’s a scene from Milk Episode 4.

Since we’re planning to enter Ludum Dare again, this part will mostly be anticipation considering we didn’t even begin the project! My main goal is to break the linearity of all our projects so far. Making choices is quite a complex thing to incorporate into a story and as a litterary person, I have trouble with that and I need to learn to write differently. It’s pretty ambitious considering the short deadline but I’ll try to stick to that goal, even if it means releasing a polished version later on. The other goal is to go out of my comfort zone with writing and talk about what I want to talk in a more…frontal manner. Indeed, many of my works address our relation to the norm and try to challenge that notion but it’s mostly told in a tacit fashion. I want to go further and stop censoring myself so much. In this perspective, I want to stage characters different from usual, under-represented ones. Like disable people. Furthermore, I plan to rest on what I learned with all the projects we made up until this point to make a game that’ll make me proud. We’ll see if we’ll manage.

What I want to learn : to make th most of what I learned so far, to write an actual interactive visual novel with branching, to make diversity an essential component of the cast, to prevent myself from self-censorship, to master Renpy enough to not have a hard time…to make a good game?



I just wanted to share another scene from Milk Episode 4.

This little overview shows that an individual (just like a team) need time to gain self-confidence and skills. Time to make mistakes, time to correct the mistakes, time to get numerous habits, time to be able to experiment and choose what one wants to do and how. I couldn’t imagine planning the Ludum Dare project years ago, I wouldn’t have felt able to! With this approach in mind (though iterations), the advices from the pros take on their full meaning: always try to make short projects before even tackling the “game of your dreams”.

Everyone need its own pace to learn but nobody can make miracles without any experience whatsoever. The issues of some visual novel Kickstarter are really enlightening in this respect: almost all the projects made by beginners (such as the ones who bragged about the campaign being their first visual novel) are currently running into difficulties and are having troubles keeping their promises. Likewise, the Katawa Shoujo spiritual heirs have disappeared and the few survivors are stuck on demo phase.

Moral : Don’t jump the gun, start small. Wait until you feel ready before shifting up a gear, it can only do you good ;).