Working Speed, or “Why So Damn Long for Just a Text-based Game?” (Patreon)
Content
I've read such claims. To my game. To other text games. “It's just a text game”, “only an [pseudo] html game”, blah, blah, blah. When it came to me, I usually either reacted to such things with humor or ignored them. But the unpleasant residue still remains. I don't know how many people really think like that, or start to think like that, while months of work on the next update are going on, including here... That's why I decided to tell you how I work, in detail, in all stages, with numbers. To show a rough picture for understanding, so to speak.
Stage 1: Draft.
I get out to write about 7-8 to 12-13 pages a day. We'll round it up to 10 for simplicity. That's not a bad rate, really. However, this figure alone doesn't give us much insight unless you pull up the numbers of each update that has been released.
Well, luckily I saved all the drafts after Chapter 1 came out. Let's take a look:
Update 1.2 - 217 pages.
Update 1.3 - 149 pages.
Update 1.5 - 369 pages.
Update 1.6 - 298 pages.
Update 1.7 (1.69 + 1.7) - 631 pages.
Actually, 1.7 came out in two parts and the first one (1.69) ended, if not to say was cut off, somewhere around page 300.
That is, if you round up roughly, the average update takes on the order of ~ 300 pages. At the rate of 10 pages a day, that comes out to about a month of work on just one draft.
A full month of work. But it doesn't work like that. At least once a week, I take a day off. Sometimes something happens that prevents me from working properly. Like I just don't get enough sleep or I get sick or anything else like that. Shit happens.
Anyway, to just write a draft of the text for the update, these conventional 300 pages, in reality takes at least a month and a half under ideal conditions. And I have ideal conditions as the exception rather than the rule.
Well, the draft is written, okay. But you can't play in the draft. So here comes the next stage.
Stage 2: Editing.
I usually write those very “raw” 10 pages/day, and edit them the next day, with a fresh head, before I start writing another 10 pages. Editing is generally faster than writing... Unless you get stuck on some piece you don't like all day, which is rare but also happens. Still, this work takes its toll too, given the scale, we're talking probably a week or two in total.
Especially since sometimes I edit individual sections 3-4 times. After all, this is artistic text. It's not the same as posting tweets with a personal opinion. I suspect a lot of the complaints about "just” text games are because of this. Everyone can write. And they usually write something on the vastness of the internet. An opinion, an evaluation, that sort of thing. However, few of these “critics” realize what it means to work with text. And it's not like posting a couple of tweets a day.
So, the edited draft is ready. Alas, you can't play it either. Because...
Stage 3: Translation.
I'm writing a draft in my native language. And it's not English. This way it is more convenient for me to evaluate the whole text and edit it. Alas, my language is not the current “lingua franca” (language of international communication). Which means translation is necessary. Otherwise, few people will be able to play my game.
The translation stage is the only stage where I allow myself to use specialized AI translators, feeling perfectly fine. After all, languages don't belong to anyone. The one case where AI is really progress, not degradation with the substitution of human creativity.
I use two such translators, one for raw translation and the other for checking terms and individual words. But while much progress has been made, the universal translator from Star Trek is still insanely far away. While AI can quickly translate large amounts of text, those large amounts of text still need to be edited and edited, edited, edited... A lot of editing, yeah.
On average, I translate about 20-40 pages a day. And a full next day for editing and revisions. That is, based on a rounded average of ~ 30 pages in two days, it takes at least another 20 days to translate 300 pages.
But you also, yes-yes, still can't play in the translated draft.
Stage 4: Placement into the game (coding).
This is where the most uncertain stage comes in. I usually take a section, as I call them “scenes”, and transfer it into the game in its entirety, piece by piece. I've had the code templates for dialogs and ability checks ready for a long time, so for the most part, it's Ctrl+V and Ctrl+V with extreme concentration so as not to misclick somewhere wrong. But for the dialog variables (well, so that the game knows when and what a character says), you have to prescribe manually, otherwise it is impossible to do it. As well as a number of other “insertion” code sections with boolean algorithms. Since the scene takes, usually 10-20 pages, this is the speed of placement directly into the game.
But it is not always so fast. For example, remember the moment when Selene found herself on the island and cooked dinner? Selene could have just picked berries. Or she could have caught fish. Or she could have also collected bird eggs. Finally, she could have caught a crawler (by finishing it off in two different ways). So, the game displays all of these options in any combination of them. It took me all day to write this piece of code, and I remember it well. And a bunch of foul words were sent into the void in the process.
In the game, you see anywhere from one to several paragraphs of text there (depending on what food your Selene found). And it took me a full day of work just to get all those "if's" and "else's" to correctly display Selene's efforts as those few paragraphs of text. A full day's work. That happens, too.
Well, the raw version of the text is placed into the game and... it's already playable. But not desirable. Cause...
Stage 5: Testing.
Bugs. Every time I place a scene, there are always bugs in it. Usually a lot of them. If I'm lucky, not so much. But they're always there. Somewhere a sign is missing in the code and a line is broken, causing the game to swear a “red” warning. It's worse when some hidden variable works incorrectly. Then the game does not signal the error in any way. Nevertheless, there is an error and it is fraught with consequences for you, the players.
To exclude this, I test the game in two stages:
Quick testing. Immediately after placing a scene into the game. Check if everything works as it should. Double-check hidden variables by adding their display to the game itself and then removing them from there, because you, the players, don't need to see them.
Long Testing. Usually, as soon as an update is ready, I replay the whole thing several times, double-checking everything again and cleaning the game of bugs not found by quick testing.
How long does quick testing take? It varies. It depends on the amount of added scenes and the number of bugs found. But, as a rule, all day after the scene has been added to the game.
As for long testing, it usually takes 3-4 days. It is certainly faster to edit something that is already mostly ready.
Untitled stage: Arts.
There are arts in the game. Good enough for some, not so good for others, but the fact is that they are there and they also need to be done. I usually make them before I start working on an update. But not always. Sometimes it's more convenient for me to write the text first and then make the art for it.
The time of work on them also varies a lot. When it comes to the “forest” arts typical for Chapter 2, I do them pretty quickly (because I'm used to it). Several in a day. However, there are also some like the “Golden Fields” one. Remember all those tiny grainy yellow dots on it, representing endless fields of yellow flowers? Each such dot is placed by hand. Otherwise, the program in which I make art did not allow me to realize the idea. This one took me a couple of days to make if memory serves me correctly. That happens too.
Portraits are a separate story. I'll just say that sometimes you get the picture right away. And sometimes... Well, like the Gold Fields. Not right away at all.
“Just” a text game.
And then when it's all done, placed in the game, and tested, only then can you play it. And, I hope, immerse yourself in the story and enjoy it, rather than drowning in bugs and loose text.
And even despite all that, every now and then, I replay my game from the beginning. And there's never been a time yet that I haven't fixed something, somewhere in the course of such a playthrough.
P.S. I'll tell you more about the progress of the 1.8 update at the end of the month. However, I can already see from here that you should lower your expectations for the release date (it won't be soon), but boldly double your expectations for the content. Yes, I'm aware that the last update was in the summer... But the next update will be the biggest one yet. Unprecedentedly big. Of level 1.7, but in its entirety, not in two parts like it came out. I'll save that topic for a post at the end of the month, though.
P.P.S. It took a full day of work to write and prepare this post. Oopsie.