Travelogue 12/12, The End of the Road

There is no end

Yes, I know, I’m very late. One year late, in fact. Does it matter, though?

This project has ended, and yet, it continues. It was the point all along: to make use of cycles. The development has stopped, the days on the calendar have been checked, the daily updates have vanished with 2017, and the whole experience seems even more distant at the end of 2018. But the result is still here, as a digital artifact, as a playable abstract portion of my life, and as a blur of my own memories.

While very personal, the game doesn’t need me to keep existing. It now lives in the minds and hearts of everyone who has played it, no matter how tiny the space it occupies. It needed me to be complete, but only to validate its own concept. Would it be a failure if I had missed one day, or if I had given up? Would it still have any value? How many days in a row would have made it a success? Half a year? More? Less? Did the whole game actually need to be developed to be relevant, or was its concept enough?

Sooner or later, working with time leads to absurdity. Roman Opalka carried on until the end, at 79. Had he died earlier, would his work be less meaningful? How meaningful is it now?

Roman Opalka OPALKA 1965 / 1 – ∞ Détail 1-35327 Tempera sur toile

Opalka counted from 1 to 5,607,249 over a period of 46 years. He painted each number in white on black, and added 1% more white to the background with each new painting. He didn’t reach his objective of 7,777,777, where he would have painted with white on white. But he sure proved an incredible commitment to his art, literally turning his lifespan, and more than that, his whole life, into his medium. I only counted from 1 to 365, for only one year.

I think A Road to Awe has proven itself as a conceptual work of art more than a game. Sure, it can be categorized as a walking simulator, which apparently means anything using a first person perspective without relying on mass murder as its core mechanics. But I don’t know if it’s a good walking simulator. One year later,  I still haven’t played it from start to end in a straight session. I know many levels are bland and uninspired, while a few others are genuinely interesting. It’s quite a good metaphor for daily life, or at least, for my life as a creative person.

At the beginning of 2017, Owen Ketillson mentioned in this piece that the game doesn’t really function like a diary. It wasn’t obvious yet that each day of the week had its specific color. And the (kind of) secret hub to jump between weeks was still to be discovered. So yes, the game is inscribed in time, once you know what to look for. You can actually browse what my mind produced every day of 2017. But I agree the level of abstraction doesn’t really do justice to the expected intimate nature of a diary. Very few levels are visually meaningful in the sense you can actually decrypt facts, thoughts or feelings.

This one was made the day Donald Trump became president of the USA. Kind of fitting, I suppose.

The mechanical part of the concept overrides the possible meanings of individual levels. It blends them within time and space constraints and negates their unique features. Sure, some stand out. But many look very similar, especially when you didn’t build them. This being said, 3D spaces bathed in the joyful PICO-8 palette still have more personality than numbers painted on a fading background. Brutally bright colors, and yes, even the dreadful Wednesday Yellow, probably helped me carrying on with the project. Building black and white only structures for one year would probably have been depressing.

I wonder what sense of humor Roman Opalka had.

What went right

The engine

This project would never have been possible without two amazing tools: RPG in a Box, the engine I’ve used to create the game, and Itch, the website that allowed me to distribute and update it daily.

RPG in a Box was kind of a wild bet, since the engine was, and still is, in alpha stage. Would using it for something very different from its original purpose be a problem? Would the engine handle 365 levels? Would it still be around in one year? Knowing the dedication and reactivity of Justin Arnold, its developer, I was certain of my choice, and it turned out great. Justin even added some pretty cool features along the way, including one-click export to Windows, macOS and Linux, gamepad support, smooth stairs movement and other rad stuff.

On the other hand, this project was designed with RPG in a Box in mind. Its voxel-based nature and its ease of use were major criteria. With only 15 minutes a day to make a level (at least when the project started, ahem…) my workflow had to be very fast. Having followed its development closely since one year for Nékromegà, I knew how to use it, and even if I hadn’t, I think it would still be the best choice for this project. Once my base blocks and scripts and settings were ready, all I had to do was to stick to my plan. Big props to Justin for his fantastic software!

The distribution platform

Now, if I can say the project was designed to be made with RPG in a Box, I can also say it was designed to be distributed with Itch’s Refinery. With my awful internet connection, it would have been a huge pain to upload an ever-growing archive every day from scratch, since every level added about 1 Mo to the total. And every player would have had to download the entire game every day. With the Butler app, all I had to do was to set up a tiny batch script and click my shortcut to upload and update the game. There’s been a few issues, slight delays, and one major fuck up due to me messing with a dual boot setup on my machine (I wasn’t idiot enough to not make backups, though), but overall it worked flawlessly.

Press, YouTubers and players

I was very lucky to attract several nice people who took interest in the game right from the start. YouTuber Petjeh decided to play it every month while talking about what mattered to him. To see my “diary” being actually used as a diary by someone else, even if it was in a loose sense, kind of blew my mind. His first video ended up as the game’s trailer on Itch. He lost interest after six months, and dedicated his channel to the AAA multiplayer action fighting game For Honor, but it was a huge motivation factor. The account where he started his A Road to Awe videos has been closed since, but you can still watch the last of them here, here and here.

Here’s a neat video by SirTapTap.

A few others joined along, though not with regular updates, but it was fantastic to see that such an experimental game could have an audience nonetheless. I didn’t expect this at all.

Another important event that probably drove YouTubers to the game in the first place was a PC Gamer “Game of the week” feature right at the beginning of the project. Almost two years later, I still get traffic from this article, even if it’s now drown under months of weekly updates. Other articles popped here and there (yes, I should update this page.) Last but not least, the game has been featured on Kotaku in early January 2018, which provided a huge boost in visibility. I had totally forgotten I had answered a call for indie games submissions from the author on Twitter. Oddly enough, I noticed this sudden peak while waiting at the hospital for the birth of my son, randomly browsing my phone during this incredible night.

Timing-wise, it was absolutely perfect to become a parent a few days after the project ended. What if the baby had been born in 2017? Well, the project would have been delayed, of course. A lot. Creating an actual human being is so much more important than building abstract voxel structures that it wouldn’t have mattered at all. But everything went according to the plan. Retrospectively, a Road to Awe feels a bit like a countdown to parenthood. One last big silly project before getting serious. Or maybe not serious at all, maybe, in fact, a lot less serious, because earning a smile from your toddler requires other skills than game development, and feels much more rewarding. But that’s another story…

In December, as a side effect of the pressure to soon have to manage my time much more efficiently, I started to work on a Twine-based time tracker. Yes, you read that well. I ended up forgetting about every time-tracking tool I ever tried (including the one I used for A Road to Awe during the first months). So I though, hey, why not develop my own with my very limited coding skills? And to my surprise, I came up with a rather satisfying result. It was really refreshing to work on an actual tool, with a rather different mindset compared to game design. I honed my JavaScript a bit on this one for sure. Ironically enough, after a few weeks, I forgot about it (but a successor might appear in the next few months…)

One last thing to mention regarding what went right: some lovely people made donations. I earned about 40€. Which is, now I think about it, a lot. Most of the donations occurred at the very beginning and at the very end of the project, thanks to PC Gamer and Kotaku coverage. Maybe doing this for a few decades would earn me a living?

What went wrong

Time management

There is an old saying in video game development about project planning:

Any task will take at least twice the time you were expecting.

I’ve experienced this a lot. Unsurprisingly, it turned out to be true once again. I had planned to spend 15 minutes a day for building the Road to Awe. Slowly but surely, I ended up working on it 1 hour a day, and even more. The idea was to spend 10 minutes creating the level then 5 minutes uploading it and posting a screenshot on Twitter. But this estimation didn’t take a few things in account:

  • Working on level design for only 10 minutes was fun in the beginning. But weeks after weeks, it became both frustrating and stressful. I had to take it easy to keep enjoying the experience, so I dumped the timer.
  • On the other hand, this also meant I could now waste a lot of time being a hopeless perfectionist, tweaking every little detail, sometimes even remaking a whole level I wasn’t satisfied with. I still hate 90% of them, by the way.
  • Posting images on social media took much more time than expected, for a simple reason: I had to take a decent screenshot before sharing it. Finding the right spot and the right angle was almost as time-consuming than building the level itself.
  • I hadn’t planned to write any travelogue at all. I still did, and while these blog posts weren’t especially deep or lengthy, they took time. Day after day, month after month, nibbling on my already limited time.

All of this time wasting could have been easily avoided if I were a bot, and did everything mechanically without any afterthought, but sadly, I’m a human being. Why not train a neural network to build levels and update everything in my place? Because I lack the technical skills. And also, of course, because I’ve learned to enjoy the needless artistic suffering.


I have mixed feelings about symmetry. From a level building perspective, I love it. Symmetry allowed me to build complex structures rather quickly, since in many cases, I only had to think about one half of the level, or even one quarter, then to mirror it to create the rest. Even if it still was a manual process, it was much faster that building more organic or realistic structures. It became a kind of mental automation that could even be relaxing.

I think this symmetry-as-a-shortcut approach allowed me to output voxels on demand, like an algorithm I’d use to hide my lack of creativity. It was probably the smart thing to do given the time constraints, but I can’t help thinking it was a bit like cheating. And it wasn’t always visually pleasing, far from it. On the other hand, it’s still more creative than incrementing an absurdly big number (sorry, Roman). Maybe I should have defined a no-symmetry rule. But then, maybe many levels would have been worse, or wouldn’t exist at all.

The Maze of Stairs to Awe

This was supposed to be a road, but for most parts, it’s just a terrible brutalist mess of colorful floating stairs. Before I started working on the project, I had this mental picture of a small road meandering between trees, in a somewhat organic way. I never managed to build it. I didn’t even get close to it. Yes, it was an unrealistic goal to aim for, especially with low-resolution cubic assets… Though it wasn’t really a goal, more some kind of vision. It still haunts me, so I guess I’ll have to do it justice in a future project.

It still feels incomplete

One year after this project’s completion, even if I’m relieved it has ended, I’m also strangely frustrated. The road is still being built in my own mental space. There are many things I would rebuild entirely, colors I would change, music I would regenerate, rules I would rewrite. But I suppose it would betray the original purpose of the project, which was to turn the past into something playable.

I had to bear with unexpected road building withdrawal. For a while in early 2018 I posted some tiny pixel art on Twitter with the #lectropix hashtag. It was so strange to not create something new every day that I kept going on, but with a smaller, less time-consuming scale. It also used the PICO-8 palette, but with a 48×16 resolution, and no specific theme or constraint.

And yes, I’m itching to launch a new daily updated project. Maybe in a few years…

William Basinski hasn’t played it yet

A few months after starting the project, I ended up buying a couple of albums on Basinski’s website, and he sent me a thank you email from some personal Yahoo address. At first I thought it was spam, but no… It seems he sometimes does this, which is freaking awesome. It was very short, and it blew my mind. I’m a big fan of his work, and he’s been a real inspiration over the years. So I though, why not say hi and tell him about A Road to Awe, since it deals with the passing of time, the core theme of his work? And he answered! “Please send me the link.” So I did, but he didn’t answer, and I didn’t insist.

I doubt he’d find it very interesting, since A Road to Awe is visually and musically rather crude and pretty far from his aesthetics, but I still find amazing he showed a bit of interested. I plan to keep buying more albums and remind him I’m waiting for his feedback if he thanks me again (insert evil/distressed laugh here).

What I’ve learned

If this project has taught me anything, it is that discipline and stubbornness can do wonders when you write down your goals and stick to your plan. Imagination, expertise, technical prowess, even motivation come second when you’ve clearly defined your creative process and its inherent constraints, and are ready to follow your fixed path.

I knew precisely what to do, when, and for how long. Yes, it sounds dull, and sometimes it was. But it also was the first time any of my projects went so smoothly. I never envisaged it as something carefully planned, especially since I was in a hurry during the last couple of weeks of 2016 to set up everything when I got the idea. And yet, the few simple rules I elaborated back then ensured the success of a level design marathon that took epic proportions.

It was a stupidly ambitious project that seemed much smaller than it actually was, like most of my projects. But unlike most of my projects, I finished it. All steps were planned as the identical parts of a linear system. Had I missed one day, the road section would have been replaced by its default state, a simple straight line, and I would have carried on the next day, or the next. Unless I died.

Of course, this kind of planning could only be applied to a fully time-based project. In fact, the planning and the project became the exact same thing. But it worked.

Now I know my next projects will have to be planned in a similar fashion. Break everything down into simple, independent tasks. Yes, it’s very hard to do, especially with video games, which are usually made of awfully interdependent and complex systems. That’s why I’ll also have to be smart with my game design. Think modular, iterate, rinse and repeat, but only within tiny time frames.

That’s also why my first goal for 2019 will be to resume the development of a time management tool that will allow me to use my limited time more efficiently. And by “tool”, more than a software, I mean a mindset.

The future

Will the Road get some updates? Probably at least one, with the upcoming RPG in a Box 0.5 release, since this version will offer numerous improvements. I don’t think, however, it will be a direct update of the current build, but mostly a separate edition with a bit more polish and quality of life tweaks. Because of the very minimalist way everything is set up, the slightest change might turn the whole visual experience into something very different. I want to keep the 2017 original edition in its current state, with all its flaws and imperfections.

I initially had plans to release some kind of soundtrack, but at this point, I don’t know how nor when it will happen. I have about 400 loops to put into order. More than 3 hours of procedural music, most of it sounding like cheap ambient garbage, but with the possibility of being used as raw material for something much more elaborate. Maybe I’ll start another daily semi-automated process to remix these loops, maybe I’ll hand-pick a few of them and try to make an actual album, maybe I’ll put together a giant mix of everything… I don’t know yet. It sounds like a good excuse to experiment a bit more with ORCΛ. Ideally, I’d love to do Weekly Beats, but I know I won’t have enough time.

I have numerous other projects made with RPG in a Box to work on, Nékromegà and Phalanstery being the ones I’m especially eager to finish. I’m also toying with the concept of a rule-free game that would work similarly to A Road to Awe, but with absolutely no constraint except updating it as often as possible. Chaining little worlds and voxel experiments together, see which recurring patterns emerge, explore something more intimate, maybe a bit autobiographical. But it’s just an idea at this point. I doubt it will become anything else before a few years.

Beside the road, there is Opuscule, the writing tool and mini-website generator I managed to develop with Godot in 2018. It has become something I use almost every day for my own needs, and that I also work on every day. In a way, it has replaced A Road to Awe in my daily routine.

And, of course, I have another shiny new game idea which has been slowly taking more and more of my mental space over the last few months, involving glyphs, vectors, roguelike and card-based mechanics, but shhh… I’m forcing myself to postpone it for now.

But for how long?

You can still explore the road here:

Travelogue 11/12, November Thoughts

November vanished in the blink of an eye. Honestly, I’m not even sure what happened, except the road carried on, and still does. November ended with 26184 downloads, which means the game was downloaded 729 times during the month. It’s slightly better than October. At this point I can’t say I care, I’m just posting these numbers here because that’s what I did during the previous months.

Nothing noticeable happened on the road. The countdown before the end now feels so real it doesn’t actually does,  if that makes sense. The last branch of the levels hub is now being lit, and it brings me both a sense of loss and achievement. Frankly, I’m a bit amazed to have carried on with this project this far. It’s the smoothest game development that I’ve ever experienced. When you sum it up, the reasons are obvious:

  • The deadline is every day. So the planning isn’t abstract. Do it today: success. Don’t do it: fail.
  • No new features were added during the production of the game. There’s a few exceptions: minor tweaks, and small engine-related features that weren’t implemented by me. But overall, everything had been defined before production. There has been no technical challenge to overcome, except managing my time.
  • Posting every level’s screenshot helps me to keep track of my own progress. It is both a motivational tool and part of the development discipline. I’m not sure I would have kept going on without an audience.
  • Once the ritual has started, it’s easy. It’s mechanical, even if level design needs creativity. Sure, I may hesitate for a few seconds, or a few minutes, before coming up with a design idea. Sometimes I have a clear idea and stick to it. Sometimes the level builds itself. Sometimes I’m painfully uninspired. But I always come up with something.
  • I follow the rules. I occasionally bend them, but never spontaneously. It’s more a slow evolution of the constraints, a refinement of my creative process over the months. I allow myself a little freedom, but always stick to the core principles and the initial vision of the project. Rules are tools that can be adapted, not commands carved in stone.

And, well, that’s all for this month. My free time is decreasing. Don’t forget to play the game if you haven’t yet, you can still be part of the journey:


Travelogue 10/12, October Thoughts

We’re now at 300 levels and beyond, with exactly 25558 downloads since January 1st (without counting November). That’s not even 600 downloads for the month, even less than the 700 of September. Not that I care at this point, it’s still an amazing performance, but it’s decreasing, and it’s nothing compared to the almost 300 downloads per day the game had reached  a few months ago.

Regarding October, two major events deserve to be mentioned: one major fuck-up due to the unlikely combination of Linux, Nvidia and Itch’s Refinery, and a new building block.

In the first case, it all started with me wanting to install a new Windows 10/Manjaro dual boot on my development laptop. Which was pretty stupid, but where would be the fun otherwise? In fact, I have a bit of experience with Linux, so it wasn’t totally reckless. Except I didn’t know Nvidia GPU drivers are a nasty mess to install on some configurations. And my Asus laptop, well… installing these drivers on it caused a complete boot failure and the inability to use either my Linux or Windows partition. Not to mention the EFI boot partition I erased right before that, but that was another problem. Anyway, the end result was I had to use my old laptop for a few days. My backups worked perfectly but the problem, because of course, I encountered another problem, was that for some reason, Butler (Itch’s developer tool to update games) refused to authenticate said old laptop.

This means I had to to upload manually the whole game to put new levels online in time. I had to upload the whole 350 Mo archive instead of the less than 1 Mo daily patch. And of course, all players had to re-download the whole game instead of the patch. Uploading manually also broke the versioning system in place since the beginning of the project. It was reset on a different development channel. So, if you’re using Itch’s Refinery, be careful to never upload manually a build with the same name than the channel you’re using… Just saying.

The irony is I only wanted to install Linux in order to test Rotonde, a very promising decentralised social network, which requires the Beaker browser, hence either MacOS or Linux. Imagine how I laughed when the first Beaker Windows build was released a couple of days later.

You can find me on Rotonde here, by the way:


It’s really worth a try if you’re interested in the future of the free web.

And now, for something entirely different. A new building block. One more broken rule, since I didn’t intend to create any new building block during the whole project. But this one was sorely missing since a pretty long time, while a few others were proven rather useless.


Here are the main characters of the game.


This deserves a bit of explanation. So:

  • corner5 is the new kid on the block (quality pun, I know). Especially useful to join other stairs.
  • cube is, you guessed it, a cube. Probably the most used of all blocks.
  • corner3_b is a purely decorative block that still looks cool.
  • corner3_t is the same thing, but upside-down.
  • cube_nw is exactly the same thing than the regular cube, except it’s non-walkable.
  • light is not really a block. It’s an empty block that emits light.
  • corner4 is a kind of corner stairs that is useful for pyramids.
  • corner4_r is similarly useful for, well, upside-down pyramids.
  • stairs is about as obvious as the cube.
  • stairs_r is in fact the lower half of stairs.
  • corner4_s_b seemed like a good idea but is in fact mostly useless.
  • corner4_s_b_m is equally useless.
  • stairs_s isn’t even walkable, but it can be great to soften angles and create intricate shapes.
  • windows is a fancy version of the non-walkable cube.
  • corner4_s_t is as crappy as the others 4_s.
  • corner4_s_t_m is… well, you get it.

To sum up, half of these blocks are rarely used. The ones from the corner4_s series are interesting for decorative purposes, sure. But they’re not walkable and only look good when used together, so… Retrospectively, I think some slabs, or “half-cubes” would have been way more useful. But that will be for another project, maybe.


This new block is so badass.


And that’s all for now. There’s two months left before the end of the road, and frankly, most of the time, I wish it was already finished. But about one level in ten looks and feels great, and this is what makes the whole experience valuable. Carry on through the boring, uninspiring parts, and dig the little gems along the way. Life in a nutshell.

Get the game here:

Travelogue 9/12, September Thoughts

As the number of remaining levels switched to a two-digits number, I have a strange feeling I can hardly describe. Even when I reached the “more than half of the road” milestone in July, the end of this project still felt very abstract. Now it’s getting serious. There’s something I’m the only one to see: the ID of each map I create in the editor, which is decrementing. The countdown is now under 100, and it feels real. Every day brings it closer to zero.

The game has been downloaded about 25,000 times. That’s barely 700 downloads for September, so its seems Itch’s algorithm is burying it even deeper. Not that I care at this point. I never expected it to reach such crazy numbers in the first place.

When I look back at the first levels, I can’t help but notice the road has been slowly evolving, and started to follow its own rules. Over the last couple of months, I’ve added new sound packs to Mixtkl (well, old, kind of deprecated packs, in fact). Which means it’s been a while since I don’t simply take the first generated ambient track that comes up, but make several tries before finding something that sounds satisfying. To my surprise (and my delight) there are quite a few instruments with darker, more aggressive vibes, including beats. While I like most of them, they rarely fit with A Road to Awe’s initial ambiance, so I have to tweak a bit my generative process, which takes longer. Good thing I don’t use a time limit any more, eh? I guess over the months, making a level went from fifteen minutes to one hour…

On the other hand, there are plenty of new drones and synths that add some variety to the mix. Maybe you’ve noticed slight changes, a small evolution in the game’s soundtrack. That’s the reason why. While two tracks can never be the same, after eight months, they started to sound very similar, so adding new samples kind of feels like a breeze of fresh air.

30 seconds of procedurally generated ambient every day.

A while ago on the RPG in a Box Discord server, Tchey, who runs the French indie games blog Jeux 1d100, asked me a few questions about A Road to Awe.

I just realized you were releasing one level everyday for one year, for your game to awe. That’s awe-some. I don’t like much the game or the levels by themselves, but the performance is quite something. Not sure what to think about it to be honest, but surely “awe” is one word for it. Did you really start 235 days ago ? Did you miss one day for some reason ? What’s your hidden goal with this plan ?

I came up with answers that I believe would be interesting to repost here.

It’s an experiment inspired by a few artists using time as their main material. A way to explore creativity through constraints. An absurd challenge that gives meaning to the passing of time. An exploration of self-discipline, boredom and repetitiveness. An auto memento memori.

I guess yes, this game, from my point of view as a developer, is a journey through daily monotony and the means to transcend it. It’s a Sisyphean project. Let’s quote a bit of Albert Camus’ thoughts on this subject, shall we?

The struggle itself […] is enough to fill a man’s heart. One must imagine Sisyphus happy.

And I think that sums up my feelings regarding this project. I’ve learned to love it, despite, or because of, its absurd repetitiveness. Not that I made huge efforts to start it. I had to do it. But to adhere to a strict daily discipline hasn’t been easy in the beginning. Now, it has become natural. I won’t lie, making my daily level isn’t a pure moment of joy. It takes efforts, especially between the moment I rendered the ambient loop and the moment I put the first block in the map.

These are seconds of existential torment. What will I build today? How can I make it different from yesterday? How comes I always end up with variations of the same structures? How comes it’s so hard to create something new? Have I exhausted my creative juice yet?

And then a pattern of some sort forms into my mind. I visualize where I should click to claim the emptiness of my computer screen and turn it into architecture. Then all bricks fall into place and a new road section is built.

I actually started more than 265 days ago, because I tested the process before the year started. I never missed a day so far.

I don’t know what would happen if I missed a day. Would it make the whole process meaningless? I had a few moments of panic, where my new level was uploaded only a few minutes before midnight. But does it makes sense in other time zones than mine? The game is available worldwide. You can get it anywhere, anytime, if you have an internet connection. It’s not like every player is waiting for the new version at a precise hour.

One light equals one week. We’re getting closer.

Initially, I decided that if I lacked the time to make a level, I’d simply upload its basic template. But now I know I wouldn’t be happy with that. I’d still make a level for every day of the year, even if I’m late. I’ve broken my limited building time rule months ago because I found it too frustrating, so there’s no way I’d skip a day or upload a “permanent placeholder” section.

I suppose I should mention that due to personal reasons, there are increasingly more chances that future levels will be delayed at some point. Maybe the game will be complete on December 31st. Maybe it won’t. But in any case, I intend it to have 365 levels.

My hidden goal is to laugh at the whole thing once it’s finished. And probably to make a bunch of ambient records using the loops generated every day, following a new set of rules I still have to define. Maybe I’ll also release some enhanced version with a few tweaks and different ways to experience the game, I’m not sure yet.

And that’s for what will happen in 2018. There will probably be some kind of epic ambient album recycling the loops of 2017. And possibly an enhanced version of the game, whether if offers more ways of experiencing it, or better visuals, since RPG in a Box will be updated to Godot 3, which offers many improvements regarding lighting and shaders.

I don’t exclude more time-based projects like this one in the future. Maybe I’ll even reboot it at some point. But I still have Nékromegà, Phalanstery and other projects to work on, so we’ll see…

Until then, you can play the game here:

Travelogue 7-8/12, July + August thoughts

Yes, I’m terribly late, but I was on vacations. Which means I’ve started spending a lot of time on a crazy cyberpunk adventure game project instead of getting some rest like human beings are supposed to do. I’m having a lot of fun working on it, but this is not the place to discuss it, so if you’re interested, follow me on Twitter.



That’s why this travelogue issue covers two months. Both were, download-wise, a disappointment. The reasons are unclear, though very probably due to a change in Itch’s discovery algorithm. Downloads have suddenly, drastically decreased. Let’s do the maths. 21,270 at the beginning of July, vs. 23000 downloads at the end. And 24,290 on the 1st of September. That’s not even 2000 downloads for July, while June had more than 5000 downloads. 56 players on average per day instead of more than 150. And it’s even worse in August, with about 42 daily downloads. What the hell happened?

Summer, of course, and vacations. Tedium, possibly. However, that doesn’t explain such a sudden drop. Most of the traffic was coming from, bringing around 70 to 100 visitors a month. But in July, I’ve had only 10 visits. It seems like my game has been pushed waaaaaaay down the page, which is quite unfortunate. My guess is, all kinds of game downloads were counted as “equal”, but a distinction has now been made between newly uploaded games downloaded for the first time and updated games which are downloaded regularly. New releases obviously seem to have more weight than updates. Does that make sense? Probably. Will it encourage frequent updates for existing games? Definitely not.

I’m pretty sure this was done for good reasons, i.e. preventing developers from abusing the system with very minor updates to stay higher in the list. Don’t get me wrong, technically, it’s what I’ve been doing since the beginning of this project. Except I have no idea how the algorithm work, nor ever intended to exploit it at all. And my daily updates aren’t fake or minor: they stand as the very core of the game.


Five weeks on the Road to Awe.

What bugs me here is how dependent developers are, one more time, of any distribution platform’s obscure decisions. I don’t think Itch is better than Steam at this. We know absolutely nothing about their ranking algorithms. Not to mention games featured on the home page are curated manually. If you’re best pal with the Itch team, (or if you happen to be Double Fine), good for you. If you’re not, well, you’re no one. In a system actually built for indie developers, shouldn’t ranking algorithms be publicly explained, or even open-sourced, and refined by the community to avoid exploits?

Itch doesn’t owe me anything. They’re a wonderful indie publishing platform for indie and hobbyist developers, and I love the unique service they provide. This project would simply not exist without Itch’s Refinery: it was built around it. But in the end, the rules that define how games are displayed or not on certain pages is as opaque as Steam’s system. I have no power at all over my game’s visibility, and that pains me. Oh sure, there’s that thing called “marketing” which takes roughly half of a game’s budget when done seriously. But I’m doing this as a creative hobby, remember? I don’t have the time nor the energy to promote my work in a professional way.


Five other weeks on the Road to Awe.

And now for the plus side…

This changes nothing to fact some people enjoyed the game enough to send me a few bucks. This is, in my opinion, and for my ego, a major accomplishment.

YouTuber Petjeh continues to make his monthly recap with A Road to Awe! I’m still amazed my game can be experienced as an almost passive, ambient medium. This “non-game” approach is definitely one of the most interesting highlights of this project. Watch it here. (for some reason, this doesn’t seem to be embeddable).

The awesome Justin Arnold has added game controller support for RPG in a Box, which means you can now use a Xbox compatible controller (and possibly other controllers as well) to play A Road to Awe. As I mentioned in this dev log, using a game controller offers a new way to experience the game.

He also enhanced how stairs collisions are handled. You can now follow precisely the slope’s shape, or smooth it with a linear or curved style (not to mention you can assign a different tile for the collision to tweak things even more precisely). Maybe you were used to that shaky ride when climbing up or down by now… But the player’s movement will now follow a linear path, meaning oblique trajectories are now as smooth as when walking on flat terrain.

Want to play the game for the first time or to experience a smooth gamepad ride? Try the link below!

Travelogue 6/12, June thoughts

As I mentioned on Twitter, I think I know how Sisyphus would feel once arrived at the top of the mountain. I’ve made one level every single day since January 1st. And doing so, I’ve tweeted about every new level with a screenshot and a number. We’re now beyond 182/365. While screenshots are numbered in chronological order, the internal level IDs are numbered in reverse order so the latest level stays on top of the list. Which means in the editor, it’s kind of a countdown to 1, with a value that started at 365 and now below half of this number. If we drew two lines, they would now have crossed.

It seems downloads are still increasing. On the 1st of July, we were at 21,269 downloads, vs. 16,783 on the 1st of June. Comparatively, that’s almost 500 more downloads than for May. It means an average of 150 daily players, which keeps blowing my mind. The game has got a couple of nice video reviews that may explain this (not to mention the ongoing monthly recap series by Petjeh).


A YouTuber with 12K subscribers can sure help a bit spreading the word.


This one is in Indonesian so I’ve no idea what he’s saying, but he seems to have fun.


Something unexpected also happened. The amazing Justin Arnold, who develops RPG in a Box, the engine I’m using, had the great idea to update the control scheme in first-person view. So it’s now possible to hold the left mouse button to move forward and simultaneously hold the right button to look around and change direction. Instead of clicking a distant tile and holding the right button to look around during the ride, you can now choose to control every step of your walk, without the clunkiness of the keyboard. A few people complained about the controls, so this is indeed a very welcome addition, and also a new way of experiencing the game. It somehow feels a bit like racing game, where the goal isn’t to go fast or avoid obstacles, but to make the camera turn as smoothly as possible to keep that ambient flow intact.

Regarding level design, there’s not much to say. I’m starting know what works and what doesn’t, but that doesn’t prevent me from building levels I don’t like. It’s still quite hard to design balanced levels. I’d say a good level doesn’t feel repetitive nor predictable, avoids losing the player in a massive coloured void, has interesting landmarks and points of view, and has music with a rhythm that complements its colours and structure. And yet, all these elements put together create something very subjective, that can feel satisfying or not depending on who’s playing. They can even contradict themselves. Symmetry is usually predictable, but visually pleasing. Repetitiveness can be aesthetically striking, but dull to explore. Music can be dissonant, but create a unique atmosphere. Reaching the exit within 10 seconds or 3 minutes can feel surprising or boring. A beautiful screenshot can hide a bland level, while one I hate will get more likes and retweets than any other before. I suppose it’s all in the eye of the beholder player.

During these six months, I feel I’ve learned a lot. And I also feel I’ve learned nothing.


Travelogue 5/12, May Thoughts

So, it’s that time of the month again. We’re now at 16,783 downloads, which means roughly 4000 more than in April. It’s about 15 more downloads per day. Maybe more people discovered the game, or more players updated it daily, or both. In any case, it seems A Road to Awe is slowly getting more popular. I have no idea what one can find in this game. A colourful routine discovery? A daily one-minute long chill-out session? I’m not even sure what I find in this game.

After five months, it has become a habit, maybe even a reflex. I doesn’t feel like a process or a ritual, but mostly like a part of me. It’s like eating or getting to bed. Something I have to do daily, that I sometimes think about, but merely notice. This possibly explains why I gave up with tracking the duration of my level design sessions. Stacking voxels has become natural. Definitely not in 15 minutes though. You can easily double or even triple that time. But doing so feels much more satisfying than building something as fast as possible.

However, it doesn’t mean I’m happy with every level. One of the strangest, hardest thing to accept is that I can like a design but hate every screenshot of it, because there’s simply no good angle to make an interesting composition, while the layout itself feels actually good. And the reverse can be true. Pictures of an awfully basic level with a dumb straight path may look stunning in some cases. At this point, I’m not even sure that the game itself is more important than the screenshot. It’s a different way to experience the same material. Many of my followers don’t play the game, or don’t care about it at all, but every day they see a new screenshot in their timelines, and even without playing, they are involved in the process.

An interesting example of a level that could only be truly appreciated in 3rd person view.

So, what is a “good” level made of? I still don’t know. Especially when the screenshots I don’t like are retweeted way more times than the ones I actually like. I’ve noticed a few patterns that people seem to enjoy more than others : massive symmetric buildings, mask-shaped structures, pseudo-figurative landscapes… But a cool-looking screenshot doesn’t always mean a pleasant level to explore. And yet, screenshots have aesthetic values of their own. Are they a by-product of the game, or the end product?

More importantly, where is the game? In the editor, when I build new levels? In the “playable” interactive sequences? On social networks? In a video streamed by a YouTuber? In the mind of my friends and relatives who understand the game (or don’t)? There are so many ways of experiencing the concept, so many interlinked processes, that I’m not sure where it starts and where it ends.

But that’s probably the point. I’ve always envisioned A Road to Awe as a self-incrementing loop fuelled by the many layers of my daily experiences.

Speaking of which, deciding to drop the 7 blocks width rule was, in my opinion, an excellent idea. It gives me much more space and freedom for interesting structures, while being barely noticed by players (except when they say levels look better). I’m currently sticking with a maximum width equal to the level’s length, which gives a maximum surface of 19×19 blocks, and no vertical limit. So I can finally build cubic levels, something that definitely makes sense in a world made of cubic elements. Why didn’t I think about this earlier, being a cube lover? Because I focused on the “road” part, assuming it had to be more long than wide. That was a logical choice, but a choice implying a figurative, material, not-so abstract approach to what a road can be. Except most of my levels barely look like a road, so what would be the point to follow a pseudo-realistic rule? I say, long live the cubic chaos that link a point to another!

Until next time, you can play the game here:

Travelogue 4/12, April Thoughts

Every day I expect to be bored and lack creativity, yet every day, after a few minutes, I get a fresh idea, or an iteration of a past idea, and create a new level. Sometimes it sucks. Sometimes it’s good. But the road keeps growing, and so far, it has been an amazing ride. We’re one third done. Four months passed, eight remain. Time fucking flies, kids.

This travelogue issue is a bit late, due to real life events, like family and friends visits and sick pets. But I managed to stuck to my daily schedule so far, and I don’t know if I should be surprised or not.

We’re now around 12800 downloads, and I still don’t have a clue about what’s going on with these numbers. Downloads seem to want to stay totally unrelated to the number of page views. However, the average activity is very stable. The game has been downloaded around 3500 times in April, which means an average of 115 people download the game every day. Maybe they just update it and never play it. But they still download it, and it makes me happy.

One source of frustration is how awesome some levels look in the editor compared to how lame they look in the game. I sometimes design complex geometric patterns, which are totally invisible in first-person, or worse, look like flat, boring crap. When the road will be complete, it could be funny to add a third-person mode. That would make the game totally different…

How could you imagine such a shape while playing in first-person view?

There’s another thing worth mentioning: I finally broke a rule. Limiting the road’s width to 7 blocks has become increasingly frustrating lately, so I’ve decided it’s time for a change. The core idea behind this rule was a that road section should be longer than wide, but it doesn’t really make sense any more given the bunch of totally crazy levels I’ve designed. A square section now seems to be a more interesting area to work with, especially since the spontaneous “visible exit from start” rule already keeps levels coherent. Probably too coherent, since it has proven annoying many times. But this is probably the most important rule, the one that avoid the anguish of getting lost in a colourful void.

Last but not least, I gradually forgot to use a timer. It seems like I don’t even bother any more. I probably should, in order to stick as close as possible to my original concept, but on the other hand, wouldn’t it be absurd to limit my creativity when I have an idea that would require a bit longer than 15 minutes to take shape? There are a few people who really enjoy the game, and it wouldn’t feel right to offer them rushed, half-finished levels. Even if I still try to be as fast as possible, I think the players deserve more than an arbitrary set of rules. I guess that makes two broken rules, right? Oh, well…

To see the results, if you haven’t already, join the journey here:

Travelogue 3/12, March thoughts

Today is the 91st day of the year. Meaning, it’s also the 91st level I’ve designed for this project. If there’s one thing I’ve learned, it’s to number the days. Not that it’s especially useful, but it gives me the illusion of a better understanding of time. And I can brag about it. Every single day.

I could also identify days by their internal level name. Today is _17_275_04-01_6.

Wondering why April Fools’ level includes a fish? It’s an April 1st custom in some French-speaking countries. People stick paper fishes to each other’s back (kids, mostly).

Let’s break this down.

The underscore is for sorting purpose. In RPG in a Box, it’s sorted after numbers, so it makes sure levels are added after my daily templates. Because yes, I have one template per day, with a default entrance and a default exit, and a specific colored lighting setting.

17 is for the year, obviously. I don’t intend to increase this number, or at least, not within the current application, since it could go on until I die and clutter the project’s tree accordingly, knowing it already takes quite a while to navigate. But it allows me to sort my “official” levels before the ones I made in 2016 for testing purposes. And it reminds me which year it is.

275 is a countdown. It decrements every day, and allows me to keep the newest level on top of the others. I’ve never told anyone “today is 275th remaining day before the end of my project”, but I could try, just for fun.

04 is the month, I guess you’ve figured this out. And 01 is the day, of course. Note the 0 before the 1, to sort this after 10, 20 and so on.

And 6 is the day of the week, which matches the number of its template. 6 is for Saturday. 6 is for purple, with a gradient to Sunday’s pink. Yes, in case you didn’t notice, I use 7 color templates that never change, and the levels cycle between them. So you can tell the day a level was made by its color.

Currently, the game has been downloaded 9300 times, which seems insanely huge to me. At the end of February, we were at around 5800, which means it was downloaded 3500 times in March. Not bad.

If you haven’t played the game yet, it’s here:

Travelogue 2/12, February thoughts

Two months and 59 levels later, the show goes on. The thing I absolutely didn’t imagine when I started this so-called “abstract diary” is that players could use it as a kind of canvas to project their own thoughts and feelings, and actually make it their own diary. This is, as you can see below, what Petjeh’s doing, reflecting on the past month and the important things that happened to him during this time, while simultaneously playing the game.

To think I’m creating a game someone else can rely upon to express themselves is fantastic. And the most interesting thing is, in this way, A Road to Awe literally is an ambient game.

Ambient music must be able to accommodate many levels of listening attention without enforcing one in particular; it must be as ignorable as it is interesting.

Let’s misquote Brian Eno.

Ambient games must be able to accommodate many levels of attention without enforcing one in particular; they must be as ignorable as they are interesting.

That would hardly be possible with a gameplay more complex than just walking to the exit. What I find especially interesting is you can explore levels as a player, tile after tile, looking around, slowly, or click on the exit and become a spectator.

I know people who let the TV or the radio on while doing something else, not paying attention at all, yet having this need for a kind of background noise. Here, it seems A Road to Awe can fulfill this role, as something interactive enough to require a few clicks here and there, but not demanding enough to require full attention like most games do. This is a way of “gaming” I would have never thought about before starting this project, one that I didn’t even think possible. Now I’ll definitely have to push these interactive ambient boundaries further for future projects.

Sometimes I spin some ambient record, totally forget about it for hours, then suddenly realize there’s music playing, and that it has been good music because I forgot it. Ambient has the unique quality of being forgettable in a positive way, of stretching time, becoming time itself, at an unconscious level. I’d sure like to explore this kind of quality further in video games.

And now for something completely different: A Road to Awe’s statistics. They simply don’t make sense, or at least, not yet. There are days when the game is downloaded maybe 20 times. And sometimes, all of a sudden, it gets nearly 300 downloads. Except there’s no new traffic source. And no special day of the week. For now, I’m unable to spot any pattern. There are even days when the page views decrease but the downloads increase almost proportionally. Maybe some analytics guru would understand this better, but to me, it seems perfectly random. But hey, I like random.

I can still say there are more players now than in January. On January 31st, the game had been downloaded about 2400 times. Now, it has been downloaded 5800 times. Which mean 1000 times more in February than in January. Not bad at all! I have zero understanding about how the Itch page ranking works, but my guess is, the more downloaded a game gets, the higher it appears on the “top games” page.

Since my game is updated daily, it means naturally way more downloads than for most other games, so maybe that’s why it’s not buried in the bottomless pit of never played games yet. Yes, algorithms are our invisible masters, in case you doubted it.

I rarely take only one screenshot for a single section.

Let’s finish this monthly post by mentioning that I spend more than 15 minutes a day on A Road to Awe. It would mostly be 30 minutes. Why? Because stupid me didn’t count the time needed to take screenshots and post them on social networks. Taking a good screenshot isn’t that simple.

I could have decided that each screenshot would simply be the default starting position of every road section, but in many cases, that would have just resulted in an uninteresting colourful mess. So every day, I find myself playing the new road section long enough to find that perfect (or mostly, not too bad) shot. I could say, somehow, that marketing takes me as much time as developing. But you wouldn’t be reading this I weren’t taking this time.

See you next month!