next »

[nathy after dark]

Progress Update #7 (October 21-25)

I've uploaded version 3.5 of the trailer to YouTube and I think it's a vast improvement over all previous versions. The first trailer I made was highly story-focused, but the game's story is so dense that it felt incomplete despite the unwieldy amount of spoken dialog. My second trailer dispensed all written words in favor of a series of pans and zooms across the game's striking visuals. In the end, the ideal solution was to add minimal silent dialog to the second version to succinctly express the game's tone and themes rather than introduce the story.





In addition to writing press emails, I'm starting to make outreach to Lovecraft and interactive fiction forums. At first I thought I could approach forum outreach as a simple extension of my email press, but after talking to some local devs about it, I realized it demanded a technique of its own. Shotgunning self-promotional posts with the game's trailer and website to every Lovecraft forum would earn me multiple bans and a bad reputation, for good reason. My game lives or dies based on whether my niche is on my side, so I'm going to be extremely careful to respect every community.

Much like games writers, every forum has its own culture and specializations, so I'll be posting about my game in one forum at a time, asking for focused feedback instead of trying to gather attention. For instance, I'll ask literature-focused communities for writing feedback, horror-focused communities for feedback on the tone and immersion, etc. Hopefully this will also earn my game a small group of heavily-invested fans, a win-win.

[nathy after dark]

The game is on Steam Greenlight now! Last night I pushed really hard and finished the game's press kit, the last thing I strictly needed before beginning the campaign. I decided to make the launch this afternoon, and hit "Publish" as soon as I got home. This was my reasoning:

October horror hype is coming to an end. To start marketing a horror game a few days after Halloween would just miss the boat, and my announcement would fall right in the time where most people are bored of horror. Early/mid November is packed with high-profile game releases. My next best option would have been November 17th, and I don't want to wait that long, so I'm starting before I'm ready. I can crunch to get more press emails out in the next couple days and hope for the best.

Here I go to update the first post with links and green coloration and stuff.

[Greipur]

Good luck!

[nathy after dark]

Progress Report #8 (October 26-29)

After day one on Steam Greenlight:



In the first 24 hours we made an excellent start. Comments were overwhelmingly positive. Another dev who's been through Greenlight told me it took their game three weeks to get so many comments. So already, Greenlight has been a really validating experience for me and the team.

After day two on Steam Greenlight:



Local devs warned me after the first day or two, the game would see a big visibility drop-off. This was absolutely correct, and it happened a little bit sooner than I had hoped. At this point, it's become obvious to me that Valve has really opened the floodgates (385 votes converting to 43%!) so I'm not as worried about whether the game will be greenlit. However, I do want the process to be quick so I can put my full attention back to the gamedev for a while. At this pace I might get my way, and I might not.

I'm thinking if things get much slower, I can write and produce an audio-heavy trailer for targeted marketing at blind player communities. I don't know if I've said this before, but I've heard blind communities are super friendly and supportive to developers, so I imagine that could be a real boost to the game on Greenlight. (I'll make the audio trailer before release either way.)

If you'd like to support the project, now is a good time to vote.

Programming Stuff

I figure I'll just stay on the programming aspect of things until I see where the Greenlight trend goes. If it grinds to a halt, I'll put myself back on marketing to push through.

Today I resumed working on my new ContentManager class. I finished implementing it, and I know I'll really appreciate the future simplicity it gives me, but oh man, did it turn into a mess along the way. I've essentially bumped the complexity of content management to the back end. The upside is that now I have a mental image of the ideal flexible content management system, which I'm sure will come in handy for my next big project. I won't bore you with the details right now.

Later I'll put up corresponding scripts from the prototype and the new build, for before/after comparison. Scripts are much, much nicer now.

[NinthPower]

Oh hey Nathaniel, I voted for this game VERY YES.

[nathy after dark]

Thanks Austin and Greipur for the votes and supportive words. They're very much appreciated! 

[foreignsubstances]

Very cool! I like where you're going with this. I'll have to read more of the posts when I get home.

Edit: Gave you a yes on Greenlight. Will be following the game - glad to see another text-heavy game on TIG!

[Keops]

I voted YES to help a fellow TIGer and also because this does look interesting indeed. Wish you the best with your Greenlight campaign! I'll be keeping my eye on this devlog too. Good luck!

[Vivid Foundry]

Congrats on releasing it onto Greenlight - right in time for Halloween too - and best of luck! You've got my vote! That's really exciting that you're including voice-over so that more people get to enjoy the game.

[nathy after dark]

Progress Update #9 (October 30-November 15)

I just did the math on how long it's taken me to reach what I claim to be 30% completion on this project. Now I want to cry what a terrible mistake. I started this log on July 9 of this year, which makes 120 days my rough estimate on how long a third of this game takes to make. 120 + 120 + 120 = 360; 360 - 120 = 240 more days working on this project?! And July 9 isn't even the true date I started working. Good thing I'm just a committed hobbyist, so my only worry over a year-long production span is my boredom and frustration.

Just had to vent that. I'm sorry.   Video games are the best thing in the world.  

Of course there are some flaws in the 360-day estimation.

• I've been working at a sporadic pace since the summer, but some external commitments should be winding down to level the playing field.
• Up to this point I've been largely focused on game engine features and platform compatibility. These things are much more time-consuming than implementing the game's content. Since all the art and audio is done, re-implementing the chapters should go very fast.
• The to-do list I wrote up in my original post barely even resembles the reality of what I need to get done. My task management has migrated from TIG to GitHub to, most recently, sticky notes I plaster on my monitor frame (so cathartic). I'm going to revisit the original post soon and make sure it reflects the full range of my goals at this point.
• I've listed my release window as "Early 2016" on Greenlight, so it must be so.

The first point is moot because whenever a gamedev says regular life is about to quiet down, they're lying to themselves. The third point probably means the project will take even longer. What was I thinking when I estimated January for completion? Wubba lubba dub dub.

Anyway. Progress! I've given an interview with a local games publication which should be hitting the web early this week. I'm hoping it brings a significant boost to Greenlight progress, because while we're still steadily (and oh-so-slowly) climbing in votes, we're actually dropping a bit in % to the top 100 rank. Which is just sad to watch.





Our game's curve looks stunted compared to the current leaders... But we did reach 50 comments today, which is really cool, and not one of the comments has been negative! Except that odd fellow who asked me to stop referring to it as a game.

On the technical side of things: tonight I finished implementing the basis of my game's menu system. Menus are parsed from JSON files such as this one, and can be controlled with both mouse and keyboard. They currently look horrendous:



The important part is that they work. And since I've now completed 15/32 of the items on my outdated to-do list, I'm awarding myself 40% completion.

[nathy after dark]

Progress Update #10 (November 16-23)

I got an interview published with Enemy Slime, a local gaming news outlet. I talk about my creative inspirations and stuff, so if you enjoy this log, you should read it!

Back on the technical side of things: I have a functional menu that lets players select the language in which they'll experience the game.



The Spanish text is mine, not Marc's, and I know it's not totally correct. This is just to show the functionality.

I want to get some solid programming work done today. My first goal is to get the game building on Windows, which could potentially (probably) be a nightmare. If I have time later tonight I might revisit the log to give a detailed technical explanation on how I'm handling game menus and language customization.

[Museraco]

Hola, hola. Congrats on the interview. We all hope the game will get greenlit.

PS: The correct translation for "Choose a language" would be "Elige un idioma"


Keep up the good work! ¡Ánimo!

[nathy after dark]

How The Whisperer in Darkness processes multilingual content

I promised a technical update! Whether anyone wanted this in the first place is another story, but have it ye shall.

I want to explain how the game loads and displays text in the language selected by the user at any given point. From the beginning I planned to keep game text in separate "language packs" so a new language could be added without any interference in files containing another language's text. Inside the game's content directory, text would be organized in JSON files like so:

Code:

content/
  text/
    english/
      menu.json
      chapter1.json
      ...
    spanish/
      menu.json
      chapter1.json
      ...
    other-language/
      menu.json
      chapter1.json
      ...

Thus, acquiring text in the current language is as simple as modifying an enclosing directory on a file path. To load text used in the menu, the file path would be either "content/english/menu.json" or "content/spanish/menu.json". In terms of pseudocode, "content/" + language_directory + filename.

What would one of these JSON files look like? Glad you asked.

Code: (English)

{
  "message-pt-1": "Hello",
  "message-pt-2": "world"
}

Code: (Spanish)

{
  "message-pt-1": "Hola",
  "message-pt-2": "mundo"
}

Every passage of text in the game is associated with a message key, which is the same from language to language. JSON files containing translated dialogue will have the exact same structure of their English counterparts, with identical metadata and the only difference being in actual content (Hello world -> Hola mundo). Writing code to say "Hello world" to a player of unspecified ethnicity, I would reference this dialogue by the two keys "message-pt-1" and "message-pt-2".

The process of loading text from a given file would have two steps, then: A) Load the text file from the selected language directory. B) Parse text inside the file into a dictionary for later use.

Easy enough to implement: I wrote a TextManager class with the power to load and parse JSON files, placing their contents in a standard-library map.

TextManager.h / TextManager.cpp

This was all well and good, until the requirements changed. My friend offered to voice act the game, and I decided descriptive audio would make an excellent accessibility feature. Suddenly, the game would be managing unicode data and audio in multiple languages. Eager to get voice-over implemented in the game engine, I started expanding TextManager to handle voice-over as well.

TextManager.h / TextManager.cpp

Ew, gross. A class called TextManager should never, ever #include "SoundManager.h". That's disgusting.

LanguageManager.h / LanguageManager.cpp

I created a class to simply manage the selected language. This would allow two separate classes, TextManager and VoiceManager to perform operations on filepaths dependent on the language setting, without duplicating language management code.

TextManager.h / TextManager.cpp

VoiceManager.h / VoiceManager.cpp

And that's about where we're at now. I haven't been working on the voice-over features since I switched to Final Cut Pro for cutting the trailer. There's still a weird bug with voice-over I've left unaddressed for quite a while. I think it was introduced by the refactor but I'm not quite sure?

But I think that sums up Whisperer's multilingual content processing. Only time will tell whether anyone was interested in reading this.

[nathy after dark]

I just updated the first post with a restructured to-do list that accurately reflects my current goals for the project. It makes a 40% completion estimate seem a tad optimistic, but this week I plan to start implementing chapters and since that's mostly a matter of porting old game scripts to the new scripting system, it should go very fast. I hope.

Looking back on the first post, there's another thing I want to address: fan translations in the Steam Workshop. My mind is split on that issue at this point. Part of me wants to use proceeds from the game to start hiring professionally/academically qualified translators and releasing more and more official translations. Another part of me really wants to play around with the Steam API (once I have access) and implement some kind of modding support. So, as of now, it's up in the air. (The two courses of action also aren't mutually exclusive.)

[nathy after dark]

New planned feature: previous chapter recaps

With the compo version of the game, problems would arise if I ever tried to show it to someone who had not enough time on their hands to complete the whole thing. The mood and pacing of the game are so fragile (as it's such a short experience) that taking a break at any point and attempting to come back, can be fatal to the player's enjoyment. (The plus side is that the game is short and this is usually not a problem.)

This is an idea that's been rolling around in my head for a while. Games like Alan Wake, The Wolf Among Us, etc. have recaps of previous content before the player starts a new chapter. This helps bring the mood back to where it was when the game left off, and also refreshes the player's memory on what the hell is going on. Just like "Previously..." with television.

I'm gonna add something like this to The Whisperer in Darkness. If a player starts the game at a chapter in the middle of the story, short scripted recaps will play for each of the preceding chapters (or just the immediately preceding one. Not sure.) I think this will go a long way in solving the problem, especially for a minimal amount of work on my part.

So I'm updating the to-do list with "recap" for all 8 chapters.

[nathy after dark]

Progress Update #11 (November 24 - December 7

I took a long break period over Thanksgiving to play more games, finish a book or two, watch movies with my family, and relax all around. I want to give shout-outs to Downwell, an amazing action roguelike platformer I played on my iPad, and The Mechanical, the newest book by my favorite novelist, Ian Tregillis. These were both inspirational experiences for me, although not in any ways directly pertaining to The Whisperer in Darkness that I can think of.

I got back to work last Friday night, and since then I've completely implemented the game's Chapter Select screen. It lets the player start the game from any chapter they've unlocked, and shows a visual thumbnail for each chapter. Along the way I made a first pass at prettying up the game menus in general. I think everything looks nicer now, although I'll continue iterating on the look until I'm totally satisfied and have sufficient positive feedback on it.

The fruit of my labor:

[foreignsubstances]

Looks great! I like the chapter select design.

[nathy after dark]

Thanks for the encouraging feedback! Helps keep up the morale. However, I'm still insecure about the look and feel of the UI, which means it's not ready yet! I'll definitely be hoping for more critique on menu layouts and appearance as I iterate on them. These things are not my strong suit which means I'm taking them extra seriously.

Progress Update #12 (December 8 )

I fixed the bug that's been holding fullscreen support back for months, and finished implementing the graphical settings menu. That's two more items off the to-do list. With the new dynamic menu system, I've set up a menu flow which lets the player make certain settings decisions the first time they open the game, and never have to worry about them again.

First the player chooses the language to play in. Their choice is saved in a JSON file, and every subsequent time they launch The Whisperer in Darkness, it will remember the choice and skip the language menu.

Then the player picks their graphical settings in the next menu. Their choice is saved, and once again the game will remember it next time. If they pick fullscreen, the game launches in fullscreen, etc.

[foreignsubstances]

I understand what you mean, when working in a text-heavy game it has been a challenge for me to figure out the best way to communicate what is interactive and what is just for reading.

However, I'd say what you have here looks clear to me, which is probably half the battle

Would I be able to compile your game on a linux machine if I downloaded the source now? Would love to try it out.

[nathy after dark]

Absolutely. There are build instructions on GitHub and I'll PM you my email address in case you need help with anything.