Tobye's Forum Posts

  • Hey jayderyu - how exactly do you get your games on there? Which export method etc.?

  • Link to .capx file (required!):

    dropbox.com/s/j9gu4ununyws6nq/ready%20to%20submit.capx

    Steps to reproduce:

    1. Play layout 'Town'

    Observed result:

    Of the 4 characters on the screen, 3 of them will be black boxes. If you delete more objects from the project, this will gradually stop happening (try deleting say 3 objects and one image will go back to normal. Try deleting 20 and all will be normal).

    Another method that worked for a bit was to relocate the affected objects to another object type folder. However, after doing this many times, the problem starts to occur again.

    *EDIT: an additional note is that when this issue occurs, it also prevented the Clay.io plugin connecting (the Clay.io plugin is not included in the posted .capx). Not sure if that helps but thought I would mention it.

    Expected result:

    The characters should appear normally.

    Browsers affected:

    Chrome: yes

    Firefox: no

    Internet Explorer: no

    Operating system & service pack:

    Windows 8 (unknown service pack)

    Construct 2 version:

    116

  • jayderyu - ah alrighty. Let's hope Cocoon get onto it soon ^^

  • Yann - Yeah I just searched it and boy do I feel dumb :( That took me a long time to sort out! xD   Well, it's already done now. Thanks for the guidance :)

  • Yann - Yes, I think 15 minutes at a time is better, it forces me to stop and think about what you've just shown ;)

    And about the dictionary, that is what I started with, but I had no idea you could save dictionaries...? The reason I did them as arrays was so I could save the array .asJSON.

  • Okay I know this is a long shot, but I want an artist for a small RPG. I already have stock art, but of course I would prefer some original stuff. The deal is a 40% cut plus $100 AUD per week. Sorry, but any more than that and oDesk becomes more attractive. The work would start in about 10 days.

    GAME OUTLINE

    -Very grim RPG, not a kidsy romp

    -No turn based combat. Actually combat is more of a real-time 'automated via commands' thing, and there isn't a lot of it in the game

    -Emphasis on story, puzzles and navigating very cunning conversations

    FEATURES

    -Full non-grid based pathfinding

    -Line of sight detection

    -Inventory system with point-and-click style usage options and item combining

    -1-6 hours of gameplay depending on paths chosen with huge emphasis on replay value

    -Over a dozen endings

    -Multiple main characters

    -Full custom soundtrack

    -Full text system with choices

    -Autosave feature that records every aspect of the game

    All of this without the use of any plugins what-so-ever to ensure compatibility with as many platforms as possible and also so I can make my own modifications.

    So far latest tests have it running at 30fps on Android when maximum number of units are on screen, which is pretty nice.

    PUBLISHING

    The game runs on pc, web, android and I am going to assume mac and iPhone (as the new .exe exporter works for mac and Cocoon's APK works fine. I just haven't got around to testing on them). I am also hoping to get it running on Ouya, but as everyone is aware there are issues with the controller.

    While I will be approaching some publishers, it's not a good idea to rely on that. And if I do, I will be approaching some of the local successes (in Australia) like Halfbrick and Voxel Agents as they have been much more receptive in the past than big companies, as you could imagine.

    The final monetisation structure is not solidified, but leaning strongly towards 'pay after a certain time'.

    This game is not aimed at becoming a block-buster or the next Angry Birds, so don't expect that, it is simply trying to be a great experience. An actually good game, not a 2-minutes-of-joy-for-a-dollar affair.

    CURRENT PROGRESS

    The game is still heavily in development, but the entire engine itself is almost complete. The only mechanic I still have to work on is combat. I expect the entire engine to be complete within a week, then it's about adding content. I have a very clear view of what is going into the game as I have been working on it as a side project for about 6 months now and have also built a substantial game doc. So I expect to have the game completed by late March.

    If you are interested, you would need to work very busily to try to keep to that timeline.

    WORKLOAD

    It will be about 10 days until I can give you the exact number, but to outline:

    The primary things that need to be done are backgrounds and characters. As I have stock art for everything, I will simply send you a level and you recreate it to fit, in the exact same dimensions. So once style has been confirmed, there really should not be any misunderstanding as to what is wanted. I will handle UI and effects myself.

    The game world is very small for an RPG, but small for an RPG is enormous compared to an endless runner. There will be about 20 different characters too that need basic animation.

    So that's about it, if you're interested or want more info, feel free to send me a message!

  • rfisher jayderyu - so you guys have controllers working on Ouya? Care to share? :) I would love to see my work actually moving about on it!

  • Yann - Haha, yeah I am halfway through and gotta say thanks! Covers stuff I had no idea about (like printing out values of arrays - so helpful!).

    How I have it set up is a different array for each level that records every enemy position, their direction, chasing player or not, health. That same array also records the self-switch for each item (so, if self is 0 the item has not been collected, if 1 it has basically).

    Inventory array that holds all items. This one is simple for me as there is a low max number of items and no duplicates, so I just add to index of 0 if it exists, if not show an 'inventory full' message. However, I think most people doing RPGs will want massive arrays with many duplicates and want to know how to display that, so my method isn't helpful.

    Then I have the player's array, which holds health, location, direction, sanity (my main character does not have many stats, but again most other people will want all that luck, agility, etc.).

    Then there is also the transition array, which basically holds all the information about where you came from in the case that, say, you go to a cutscene on another layout then return and want everything to be the way it was. So, the previous layout name, the value of any timers that happen to be active, etc., etc.,.

    So yeah, although now I have it nicely streamlined and it all loads and saves automatically, it's pretty messy stuff.

    Here's an image of the sort of way I'm going about it: dropbox.com/s/k275o9sxm2aespl/Untitledv.png (sorry I don't know how to put a picture in my post).

  • Yann - Yep! It was one of the most confusing parts for me to learn. Webstorage itself is easy enough, but when it came to filling the arrays correctly, saving them and loading them, it gets pretty tricky when you're dealing with lots of information.

  • Now I haven't actually played that game, but imagine you click one spot and drag along and it creates track right?

    First off when you create a new object for your track pieces, I suggest adding 1 animation for each piece of track you will have. So animation 1 is straight track, animation 2 is right corner, and so on. Now, on any frame of any animation, select the image point button on the left of the image window (looks like a little target).

    When you click on that, a little window will pop up showing you the list of image points, the first one being called 'origin'. Now, add 4 more image points, and right click on them. Then choose 'quick assign' and make one left, one right, one top and one bottom, then name them appropriately. Once you have done them all, again right click each one and select 'apply to all animations'. And now you have image points on every side of every piece of track!

    Now you do: 'is mouse pressed' + is overlapping track (inverted): create track (there are many ways to make the new track snap to grid, but I don't have time to go into it. If unsure, just look around the tutorials - there's a lot of info for it!).

    Now create some small square images called 'detector left/right/etc.'. Give them the variable 'track UID'.

    Then have an event: 'On track created': track spawn object 'detector left on image point 'left', 'detector right' on image point 'right', and so on. Then on detector created: 'set 'track UID' to 'track(track.count-1).uid''. This will make sure that they only affect the latest piece of track.

    Now you can do something like:

    'For each track: if detector left is overlapping track + detector left track UID = track.uid, detector up is overlapping track + detector up track UID = track.uid: set track animation to 'turn up''

    This isn't 100%, you will need to decide many things like when to destroy the detectors, whether the track can split in 3 ways, etc., etc., but this is the general method I would start with.

    Sorry I don't have more time to help, it does sound like a fun project to work on :)

  • Try Construct 3

    Develop games in your browser. Powerful, performant & highly capable.

    Try Now Construct 3 users don't see these ads
  • Thanks Yan! It was really interesting just to see a different workflow. One suggestion I have for a future video is complex saving, as in how to manage saving the game in an RPG. I have my own autosave system, but it is very labour intensive to use and was a real pain to setup, so I would love to see how you tackle it.

  • Just off the top of my head, you could try this to automatically set the rail section to the right piece (corner, straight, etc.) while dragging:

    Whenever you create a new piece for the rail, spawn a detector object on all sides (set image points for left, right, top, bottom and spawn them on those). Then as you drag past and create a new piece, check which detector it's overlapping. With that information you can set the pieces of rail to the right animations (left corner, straight, etc.). Make sure to destroy all the detectors after a certain number of rails have been placed.

    I'm just heading out the door, so sorry it's not very precise. If you want I can clarify a bit when I return. Good luck!

  • I remember when I did playtesting for a company along with 2 other guys. The developers were amazed that none of us used multi-touch, as they thought it could only be played that way and it was a big design consideration.

    So yeah, get some people you don't/barely know to have a shot as often as possible. Half Brick (the Fruit Ninja/Jetpack Joyride guys) would go out each week or so and just hand the game to strangers on the street to get their impression of 2 minutes gameplay.

    And polishing can go on until the end of time. To use Fruit Ninja again, it took 3 days to make the game itself and 3 months of polish.

    I'd suggest going through and getting everything to the bare functional requirements (backgrounds scrolling cleanly, no clipping in movements, smooth layout transitions, etc.) and then when you have the whole picture in front of you, you can see what really needs the attention. Nothing worse than spending a week on something only to find it has made minimal impact and then discover one section of your menus looks like an abattoir floor.

  • I found the simplest solution is to just use instance variables on arrays/dictionaries as they don't seem to reset when you change layouts. Lets you name and sort your groups of variables too, a big plus!

    So if you want to access variables that are just for that level, you just compare to the corresponding array. Downside is you can't clone them for some reason, so if you have heaps of variables to add it becomes very tedious.

  • WaterlooErik - Phew, I can only imagine how busy you are...thanks for your help with this and I hope developing for BB is actually this easy!

    Good luck!