BEW - Ongoing Dev Thread (DAZ Artists Still Wanted)

[Wolfschadowe]

Continued from the Adventure Creater thread at http://www.the-new-lagoon.com/viewtopic.php?f=11&t=2076&start=495 to avoid hijacking that thread

Would it be possible to limit them to the range 0-51? If you did that, it would be easy for us to encode one character per variable. That's still a 250 character string, but better than a 750 one. Bit packing the achievements is easy too.

Still, I think you've got too much state to effectively make a string for your users to cut and paste. Is there anything you can do to reduce it? From one game to the next do you really need every detail about the previous game? Realistically, couldn't you have a variable saying which NPC you're making progress with and how well the player did? Even Mass Effect didn't carry over all that much state from one game to the next.

And, if you're going to have 3 games, make 3 separate sets of achievements that don't carry over. It's really rare for you to need to have played game N-1 to get an achievement in N. I know you're thinking of this as one game told in 3 acts, but you really should think of them as standalone sequels.

Tlaero

I've been thinking about this a lot, trying to figure out what needs to carry over and what doesn't.

Regarding the achievement carryover, I'm undecided on whether to carry them over to act 2, or make each act independent. I'm not personally a huge fan of achievements myself, but am including them primarily for those who are. One of my friends, for example, is so motivated by achievements that his wife finally got him to help out around the house by giving him achievements that translate to gold stars on a board. Vacuum = achievment, do the dishes = achievement. After so many "achievements" he gets beer money and a night out. What's funny, is that it works on him. I think her idea is both brilliant, and maybe a little sad. heh.

I just returned from a business trip that took the second half of last week, so I've had time to think about different strategies for minimizing what needs to transfer between acts. Mass Effect is a good example since that is also a game where things that occur in the first game have significant effects in the second and third. My difficulty is coming from distilling what is really needed from Act 1, and what isn't. Ultimately the intent is for all three acts to meld together in a seamless experience that can be played from start to finish, but thinking about them as standalone sequels is helping. Keeley 1 and 2 melded together very well, and they originated as standalone games, after all.

Back on BEW, there are some achievements that are specific to the acts, and others that span all three acts. I could strip the overarching achievements out and just limit achievements to each act, but much of that undermines the overall achievement design I created. I definitely need more thought on this topic, although if it's easy to bitpack them then it may be moot. Still more thought is needed here on my part.

Regarding the actual game variables, it's probably not realistic to limit them to values less than 51, many of them, yes, but not the relationship values. On the other hand, the more I think about it, the fewer variables I can think of that will be needed to transfer between Acts. Each love interest really has a maximum of 3 variables that need to transfer, but really, only 4 of the interests are active at the end of Act 1. Basically those variables are the visible relationship, hidden relationship and status. Using Emily as an example, those variables would be a calculation of ehappy - eangry, earoused - ethreat, and brade. That could give a result like 62, 49, 14. Although, it may be possible to reset the relationship and just take the status to carry forward and start all the new calculations at 0 for each love interest when it comes to the relationship value. That means just the relationship status (brade in the example) needs to carry for each of the four active love-interests between Acts 1 and 2. That number could be less than 51, since there are less than 51 status possibilities at that point, or any point really.

There are some downsides to resetting the relationship values for the overall game though. One of those is that it lessens the impact of decisions made earlier in the overall story.

Act 2 concentrates primarily on two love interests, with one or more of the original three coming into the picture during the last third of Act 2. Essentially, Act 1 has 3 romanceable love interests, and the 4th is present but not romanceable. Act 2 has numbers 4 & 5 as romanceable as well as whoever the player succeeded with in the first act, if any. Finally, Act 3 reconciles the results of Acts 1 and 2, to conclude the story.

Wolfschadowe

[Wolfschadowe]

Status update from the first page:

January 20th, 2014 - Made good progress in rendering over the past couple of weeks. All three major branches are fully rendered, including the erotic scenes. There are two minor erotic branches to finish up on the render, and also the alternate path to render. It looks like the Day 2 office scene will have close to 600 pages when it's done, although only about 200-250 will be accessible for each playthrough, depending on the path. One of the reasons this is so big is that it's one of two Emily Relationship payoff scenes. For those who are on an Emily relationship path, this is effectively the end of that main branch of the relationship for Act 1. There's still some stuff that can be done in the evening, of course, including losing her as a possible love interest in Acts 2 & 3, or cementing the relationship more solidly. Most of the evening for Day 2 concentrates on Faith and Natalie though. Overall, I hope to finish up rendering for the scene within the next week or two.

Rendering is the primary bottleneck at this point, but on the plus side I've got some techniques down that have reduced the rendering time per image from 20 minutes to somewhere between 5-10, with the same level of quality. This is primarily due to lighting techniques. Unfortunately some of my playing with the lighting is evident in the Office scene, so it gets a little goofy here and there. Hopefully it won't detract from the game too much, but I'd rather get the game out then re-render scenes over and over again until the lighting is perfect. One of the main goals of making BEW was to learn these digital art techniques, and hopefully the game will evidence my growing skill as it progresses. The worst images will be the first images, i hope.

I've been travelling on business over the last week, so that slowed me down a bit, but I also have a head start on storyboarding the next few scenes, so some extra time now is saving me some time later. I have the home transition scene storyboarded, as well as a start on a small supporting scene on one of the corruption paths. Both of those scenes are only about 10-30 pages, so very small in BEW land. heh. Next on the storyboard schedule is to start filling out the Bar scene for day two with Faith...or Emily and Faith, as the case may be. I've broken the scene storyboards into multiple Visio files to make it easier to storyboard on one and render to a second without having to scroll back and forth a lot. That's it for this update! Thanks everyone for your patience!

[hrmf]

Thinking about the bit-packing thing a little more, outside of the framework of Tlaero's game creator your game state is essentially a flat JSON object, which is to say a key-value mapping. The state consists of two items: firstly the URL of the page the player is presently looking at, and secondly the set of variables representing the choices they've made in their playthrough up to now. It might look roughly like this (I'm making the variable names up out of thin air based on other html games I've seen, like Chaotic's or whatever):

Code: Select all

var state = {
  "page": "bar45.html",
  "variables": {
    "emily_like": 3,
    "emily_lust": 2,
    "emily_angry": 1,
    "bought_camera": 1,
    "bought_bikini": 0
  }
};


In between Act I and Act II or III you know what page the user will start on, so you don't necessarily have to deal with the "page" part and you can just focus on the set of values in the "variables" object.

So the question is basically how you can serialize a pretty simple key-value map from strings to keys and persist it from one browser session to another. The idea of boiling it down to a simple string that the user can type into a form somewhere is somewhat appealing, Shark has done that for a bunch of his games and it works ok, and that approach is also used with some of Goblinboy's AIF games if I recall correctly. You could probably research how URL-shortening sites like bit.ly work and come up with a reasonable algorithm.

Really, though, there's no reason to make a user write down a seven-letter (or however many) string of characters just to store and retrieve state. The end user already has a computer and it's pretty good at storing state. You could use something like HTML5 LocalStorage to persist it in their browser - like, you can just take the whole "state" object from above and store it, and then load it back up when the user clicks a button.

Also, if you wanted to, it wouldn't be too hard to generalize the persistence mechanism out to a save-game system with multiple save states, so that you might have something a little more like this:

Code: Select all

var all_saved_games = {
  "My saved game": {
    "page": "bar45.html",
    "variables": {
      "emily_like": 3,
      "emily_lust": 2,
      "emily_angry": 1,
      "bought_camera": 1,
      "bought_bikini": 0
    }
  },
  "Another saved game": {
    "page": "pool3.html",
    "variables": {
      "emily_like": 0,
      "emily_lust": 2,
      "emily_angry": 0,
      "bought_camera": 0,
      "bought_bikini": 0
    }
  }
};


Again, I don't see any particular need to encode this stuff into a string. Just have a button somewhere, and when the user clicks it, take the current state of the game, construct an object like the above, and persist it to storage.

You do lose something this way in that the user can't move from one browser to another and get back to where they were by just typing in a string. There might also be complications with incognito mode, I'm not sure. And you'd probably have to get Tlaero to write some stuff into the Adventure Creator toolkit to support this, but I don't think it would be technically very difficult to do. (Tlaero, I know it's a lot easier to say that than to actually implement it, sorry if this seems like I'm signing you up for a bunch of work!)

Beyond that there's also a game design element, to wit that you don't necessarily want the user to be able to save and reload at any arbitrary page in your game. I see that as a non-technical question and don't have a particular opinion on it other than that I trust your aesthetic judgement a great deal from the stuff you've released to date. But I also don't see any particular technical need to limit the amount of variables you use in your game - personally I'd go completely nuts trying to keep track of them all, but from all your posts on BEW so far it's pretty clear that you are well on the way to becoming completely nuts anyways, to all of our benefits.

[Wolfschadowe]

Hi hrmf!

Thanks for all the great information and examples.

Beyond that there's also a game design element, to wit that you don't necessarily want the user to be able to save and reload at any arbitrary page in your game.

I'm already using a save state function to effectively save and reload a game. That used to be done using a cookie, but Tlaero's tool recently changed that functionality to localstorage. I think that savegame functionality in a game like BEW is going to be vital, primarily because of it's size. For example, the full game of Chaotics Rachel 1.2, is about 700 pages, which is about the same size as one of the larger scenes in BEW. Right now, according to my outline, BEW will have 3312 pages in Act I, although that number will change up or down. In this case, being able to save, run down a path a dozen pages, and restore back to try something else is beneficial, in my opinion anyway. It will save time on replay number 12 when someone is looking for a new ending they haven't seen yet without having to restart from scratch.

Here's my thought process about generating a string:

Where localstorage fails my requirements is where it comes down to persistence. First, per my understanding, the size and storage of localstorage is limited by domain (localhost in this case). Let's assume there is six months between Act I and Act II....Ok, more realistically a year or so, based on Act I. During that time, I can assume more games will be played from the localhost domain that use localstorage which is FIFO meaning BEW save states may be overwritten over time, or the computer will fail, or localstorage will be cleared along with the browsing history of playboy & brazzers, or...whatever.

I don't know how many hours of playtime there will be in Act I, but there will be around 15 endings that will carry over to Act II. I haven't actually played BEW yet, heh. Maybe one of the playtesters can chip in with an estimate of playtime available. For completionists that want to find every ending and achievement, that could be upwards of 20 hours if they take the time to enjoy each unique story. When Act II comes out, it will actually come as a contiguous seamless game with Act I. New players can start with Act I and play straight through to the end of Act II, and the same when Act III comes out.

I am looking for a failsafe method for those who played Act I to death to use so that they don't have to replay it in order to continue their games into Act II, assuming that they have lost their save-state information in localstorage. I don't want to do a checkbox style system like SITA does in the Arieaneb series either, that is just outside of my cheaty threshold. I can already envision a forum thread listing ending codes, if the game is popular enough, which is fine because there's a little effort involved in that for those who will want to go that route.

So essentially, once Act II comes out, there will be three options to play it.

1. Play Act I and continue seamlessly to Act II content
2. Load an Act I savegame, and continue seamlessly to Act II, the cliffhanger ending page for Act I will have a reminder to save the game in one of the save slots.
3. Enter your ending code from Act I and continue seamlessly from the Act I ending page to Act II. Again, I'd like to have an option of the ending page of Act I to generate a failsafe code.

But I also don't see any particular technical need to limit the amount of variables you use in your game - personally I'd go completely nuts trying to keep track of them all, but from all your posts on BEW so far it's pretty clear that you are well on the way to becoming completely nuts anyways, to all of our benefits.

Yeah, I've definitely gone nuts over this: The full spectrum, walnuts, peanuts, cashews, tennis balls, pecans, and hazelnuts. But...I'm still having fun with it.

Wolfschadowe

[karrek]

I have no experiennce in web development so forgive me if this wouldn't work, but would it be possible to package up the state into a file that the user could save on their computer somewhere, and then load up that file when they want to start the next act? I have no idea how hard file input/output is in html or javascript, but if I were developing the game in Java that's how I'd do it. That saves the user having to type in a huge string, but it also prevents automatic overwriting of localstorage.

[tlaero]

For security reasons, web browsers don't have access to the filesystem. IE used to allow it with ActiveX, but I don't know if that still works and it certainly doesn't work for other browsers.

Tlaero

[karrek]

Again, please forgive my ignorance of web development, but would it not be possible to generate a file that the user could "download", the same as if they were downloading something from any other site? I mean I know a website can't write directly to the hard drive, but obviously there's some mechanism for saving files locally through a web browser.

[Wolfschadowe]

Again, please forgive my ignorance of web development, but would it not be possible to generate a file that the user could "download", the same as if they were downloading something from any other site? I mean I know a website can't write directly to the hard drive, but obviously there's some mechanism for saving files locally through a web browser.

I like the way your thinking, and on the surface it makes sense. In practice though, web browsers can only display files, or download them. They can't actually create them.

Creating files is generally handled by the web server for the website, not the browser. A web site may have a link to generate a file (Like downloading a pdf version of your credit card bill). The server doesn't actually have a pdf version of your bill on hand, but clicking that link runs a separate application on the server that generates the PDF file on the fly from a database, and drops that generated file to a temporary folder. The web server will then tell the browser on your computer where to find that file and your browser will download it or open it.

The key is that creating the file is handled by a web server and various applications running on that server. BEW and other AC games are serverless, meaning they can be downloaded and run locally and don't use a web server.

Internet explorer can use an ActiveX control to pretend to be a web server to generate a file, but not by default. You would have to change your browsers security settings to such a low level that your browser would effectively be teabagging a grizzly bear with a honey coated scrotum.

Wolfschadowe

[karrek]

Well, learn something new every day. Too bad it wasn't any help to you guys.

[rabbie]

I love your dedication, only reason i didn't jump forward as a beta tester was because i wanted to come into the full game fresh... selfish i know, but you seem to have had no shortage in willing volunteers.

[Wolfschadowe]

Well, learn something new every day. Too bad it wasn't any help to you guys.

Hey, it helped you learn something! Besides, even things that don't seem to help on the surface often make me think about things differently and end up helping in unexpected ways. Keep the ideas coming and don't worry if they help or not. I personally probably have 99 bad ideas for every good one, and I'd never have that good one without the thought process invoked by the other 99.

I love your dedication, only reason i didn't jump forward as a beta tester was because i wanted to come into the full game fresh... selfish i know, but you seem to have had no shortage in willing volunteers.

Nah, not selfish at all. I don't know how many times I've burned myself out on betas. It's a nasty habit of mine, and now the Steam Early Access is just making it worse. Curse you Kerbal Space Program!

Wolfschadowe

[Wolfschadowe]

Update to the Status post on page 1:

February 2nd, 2014 - So, it was a year ago that I rendered the first image for BEW. Ok, technically, it was a year ago, four days ago on January 30th, but close enough. It's amazing how much I've learned in the past year, and just as amazing how much I know that I still have to learn about digital art, storytelling, and AIF game creation. I naievly thought that I'd be done with the game by now, and I'm only about 3/4 of the way through the first act. While there are some things I'd do different if I started again, overall, I'm happy with the way the game is shaping up. I could have made a much smaller game and released it by now, but then it wouldn't be the game I wanted to make. Sure, I could have left one of the love interests off for this round, and added her back in as an expansion later. That's probably the one difference I would make if I was starting again, but I don't regret including that additional story line to the first version either. It's still the game I wanted to make when I started, and a year later, I'm still having fun making it. I'd like to thank everyone who is following this and hanging in there, not only for your patience, but also for your continued interest in the game.

As for status, I have completed rendering all the main paths and almost all of the branches. As I am going through my logic pass I keep running across little 4-10 page offshoots that aren't marked as rendered yet, but most of those get filled in with already rendered images. Occasionally one justifies a new render, and I still have a few images from the first half of the office scene to re-render, but I'm probably less than a week from being done with all the rendering for the scene. Once I finish the logic pass and pick up any stray renders, I'll build out the web pages using AC, run spell check and have it out to playtesters. the next few scenes that I'll be working on are small transition scenes. One is the main path of Brad at home before going on his date for the evening (if he has one), or deciding where to go if he doesn't have one yet. That one is only 20 or so pages, once all the branches are accounted for, but only about 10 are actually going to need a unique render. Another small scene is Brad picking Emily up at home before their date for one of the branches. That on is only 10 or so pages since it's a consolidated path. There is also a side branch at the local university to support one of the corruption paths which will be another 20-30 or so. Those should come fairly quickly. After those are completed, the Day 2 Bar scene for all the branches, then the Day two Strip Club. Both of those scenes will be fairly large. There is a third destination available on some paths as well that will come afterwards. Finally, to finish out day 2, there will be fairly short scenes at either Faith's house, Brad's, or Emily's that will lead into the Day 2 Recap. To finish out Act 1 we will have a 40-50 page scene for the morning that will set up Act 2.

That's essentially the road map for the rest of Act 1. Thank you all for your support and interest over the last year!

Wolfschadowe

[karrek]

Eagerly awaiting the next playtest.

[davori]

Hotfile seems to be down. Can u uploaded again somewhere else please?

[Wolfschadowe]

Eagerly awaiting the next playtest.

Getting close Karrek. If everything goes perfectly I could wrap up the next release by the end of the week...but everything won't go perfect, so probably a couple weeks. This should be the longest gap between test releases.

Hotfile seems to be down. Can u uploaded again somewhere else please?

Well, crap monkey. I lost the 0.0.8 version. I have posted the latest version containing only the same content as 0.0.8 in the first post. There is very little new content for those that played 0.0.8, just continuity cleanup and enhancement stuff here and there, but the Recap has been re-enabled and achievements are present. It was a fast hack job to pull this out of the current build so I may have introduced a bug or two. Let me know if there is something game breaking and I will post a patch.

Here is the link again for those following:
http://www.mediafire.com/download/outr26ybd5wn865/BEW_0.2.3_PoC_Beta.zip

Changes from 0.0.8 for the 0.2.3 POC Beta that is posted:
Changes from 0.0.8:
Re-enabled the end of day Recap
Minor story changes and enhancements in the Bar scene, plus general bug and continuity cleanup.
Abbreviated achievements are enabled for existing scenes.

Wolfschadowe