Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - dtishin

Pages: 1 2 3
Fog: True that makes sense.
Icons: You know use color coding for the segments on th left (hints/status and so on). I thought diving them a bit more or adding an icon at start would make the screen easier to digest.


I'm actually using colour coding for different rumour classes (rumours about monsters = black; rumours about perils = red; rumours about artifacts = blue). I suppose I could additionally separate rumours about monster behaviour / stats from rumours about monster location - that totally makes sense from the mechanics viewpoint. Thanks again!

Very good job!
I like that you see the whole "world" infront of you and are motivated to explore.
Maybe some "fog of war" might be interesting to add.

One wish: A bit less text. maybe you could add icons in the left part under the portraits to make it easier to read/digest.


- I tried the fog of war in an earlier build and it definitely adds to the exploration part, but it also contradicts the rumour mechanic - the player doesn't know which clues to buy until the whole map is revealed.
- As for the amount of text and icons, can you clarify? Part of the rumour mechanic is interpreting the text into actual map phenomena, e.g. "evergreen plants" can be desert cacti or conifer trees. How would you make an icon for that?

Thanks Luyren!
Good points, added to my bucket list :)

Hi fellow Stencyllers!
What do you think about this game?
Fireside Hero is a compact turn-based strategy / RPG set in a fantasy world.
Very roughly the concept can be described as "Heroes of Might and Magic meets Minesweeper" :)

Key features:
- Cosy old-school look
- Original rumour mechanics
- Combat built on rock-paper-scissors principles
- Emergent story with a bit-size narrative
It's my first personal project, and I've been working on it for about 9 months, planning to release later this year.

You can get an idea of the gameplay in this video snippet:
Or access the latest alpha build on Google Drive:

Any feedback and comments highly welcome, especially if you love RPGs, strategies or puzzle games.
Thanks a million!

Ask a Question / Re: Cppia build not compiling / crashing
« on: November 12, 2019, 11:38:29 am »
Got debugging to work.
The original instructions (as far as folder paths are concerned) seem a bit outdated.
After adding the debugging parameter (Settings -> Advanced), I simply chose "Publish -> Desktop -> Windows", and then opened the .exe file from inside Visual Studio. The location of the problematic behaviour is quite clear from the Visual Studio log.
No folder with the name "cpp" or similar was created in the Worskpace folder though.

Ask a Question / [SOLVED] Cppia build not compiling / crashing
« on: November 12, 2019, 10:39:15 am »
Hi all,
I need to debug my project in Visual Studio using Justin's method (,49211.0.html). I added the debugging parameter to my game settings, please see attached screenshot.
As I understand, for the method to work, I need to launch the \Cppia\executable in my workspace folder.
Unfortunately, I couldn't find the cppia folder in my workspace (please see screenshot), and when I try Run -> Cppia, the game does not launch and creates an error log (attached).
Any advice how I can export the cppia project?
I'm running Stencyl build 10472 on Windows 10 Pro (64-bit).

First guess: lists and maps that are hidden must be initialized (set to "create new list" and "create new map" respectvely) before you can access any data on them.
Thanks Luyren, the error disappeared as soon as I initialised all lists and maps when created.
It seems that it doesn't matter if they are hidden or not!

First guess: lists and maps that are hidden must be initialized (set to "create new list" and "create new map" respectvely) before you can access any data on them.

I know about hidden lists, so I don't normally hide any maps or lists, please see attached. I had also added a default pair of (key + value) just in case, so that it doesn't consider it empty when compiling (though I have no clue what happens under the hood).

Hi all, I'll appreciate some info on this :)

I access a key ("name") of a map-type actor attribute ("monster selected") in a "creating" event.
When I test the game I get the "Null object reference" error.
Still, if I dismiss the error pop-ups, the code works as intended.
What is causing the error message, and how do I make sure it doesn't pop up?

Please see a screenshot of the game test (with monster names actually drawing in the scene) and the code attached.


Thanks Luyren, that was great advice!
Instead of checking the LOCAL list for all UNSUITABLE tiles I checked the GLOBAL list for all SUITABLE tiles, then copied them to a local list. Like clockwork :)

Depends on where the level is located in the game. For a starting level it's insanely hard.
It might help to allow players to change direction of elevators on the move, will save them a lot of time and frustration.

Hi, I encountered a weird issue.

TL,DR: The “remove item” function does not remove some items from the list even though the items meet the condition; at the same time (seemingly) identical items that meet the condition are removed. Does anybody have an idea what’s happening here?

In detail:
-   I have a scene of 16 x 9 tiles (144 tiles in total). The tiles in the tileset have data attached to them that describes their terrain type or a combination (e.g. “grass”, “sea” or “grass;hills”). When I create the scene I programmably build a list (game variable) of all tiles in the scene. Each item in the list is a map. The map has keys “column” (number 0-15), “row” (number 0-8), “occupied” (number 0-1), and “data” (list built by parsing the tile’s “data” field).
-   Every turn I need “quest goal” actors to be randomly placed on the map. But I need them to be only placed on suitable tiles, i.e. 1) unoccupied by other objects, and 2) NOT containing data “sea” and 3) NOT containing data “town”.
-   For this purpose I set a local list variable to a copy of the global list of tiles, then cycle through it and remove all items that meet either of the criteria above (containing “sea”, “town” or “occupied=1”).
-   It generally works fine except some “sea” tiles are not removed from the list :(
I attached a picture of the scene (Map_unremoved_tiles.jpg) showing tiles for which the corresponding map items are not removed from the local list. E.g. file “log2” shows that in row 8 (bottom row of tiles in the scene) only three items are removed; the other two are not.
-   The log also shows that the local list does contain the items that need to be removed. E.g. file “log1” shows that both tiles with coordinates [0;8] and [1;8] are in the list and they have identical value of the data field.

The code goes attached for your convenience. It looks big because of all the “print” debugging functions :(

Thank you!

Thanks for your advice! Pathfinding is now fixed and expanded. I've implemented dynamic pathfinding, plus NPCs now run away from bombs and from the player when badly hurt (actual threshold HP depends on the current 'behaviour' profile of the NPC).

The main thing was splitting pathfinding into two separate events - 1) an Updating event that watches for changes of the player or the NPC's tile, and 2) a custom event launched from the Main AI Switch (essentially a behaviour tree). I also dropped the "direct motion to the player" behaviour since dynamic pathfinding is much handier anyway; I didn't need the check for collisions this way either. You can see code screenshots attached and a video of updated game mechanics under the URL :)

Thanks again!

Ask a Question / Re: Stencyl UI is teeny tiny - can't read it
« on: January 12, 2019, 04:55:03 pm »
I had a laptop with high resolution display and to fix it I had to:

right-click stencyl.exe > properties > compatibility > change high dpi settings.
Then checkmarked everything and set the dropdowns to: "I Open this program" and "System".

Wow, that was a life saver! It doesn't fit as much information on screen now, but it's definitely more friendly to the eyes.
Thank you very much! This piece of advice should be pinned somewhere by the Stencyl creators...

Hi folks,

I'm working on a prototype of a top-down RPG set in medieval North Asia, with focus on tactical problem-solving (varied combat options vs AI manipulation - avatar will be able to play a tune that changes the NPC behaviour patterns - cautious vs reckless, peaceful vs aggressive). Currently finishing the core of combat AI. The artwork is currently placeholder opensource tileset, and the level just a testing ground for the mechanics :)

While putting together Main AI Switch (the "updating" event, which manages NPC behavior), I ran into a problem.

Expected behavior: When the player attacks the NPC outside the NPC’s detection range, the NPC should stop its current “Wander” behavior, switch to “PullToAvatar” (move in a straight line towards the avatar), then if the NPC collides head-on with a Tile, switch to “PathfindingToAvatar” (simple movement on a tile-based path created with the AI Tools extension). Once “PathfindingToAvatar” is initiated, it should switch the “path_initiated” boolean attribute ON, and there is a special block in the “Main AI Switch” event (see the first “Otherwise if” block that initiates the “PathfindingToAvatar” event) that ensures that while “path_initiated” = TRUE or “path_to_avatar” list  contains any items, the NPC executes the “PathfindingToAvatar” event until reaching the end of the path or until getting within the weapon range of the player.

Current state: either the NPC doesn't move on the path at all OR the NPC executes pathfinding once but for new instances of pathfinding it just stands facing a tile. Please also note that in the “Pull path” video the number of remaining "steps" to travel is changing as the avatar moves on the scene, which is incorrect behaviour. It means that the pathfinding movement cycle is not properly initiated. Please also note that the pathfinding algorithm itself used to work perfectly in my previous (less complex) versions of NPC AI. I suspect there is a fault in my Boolean switches and their resets (or in how the engine handles the triggered events, maybe it doesn’t get enough time to turn the “path_initiated” switch ON and executes the “Main AI Switch” event again?

Please see an export of the game, a few screenshots and two videos here:

If you import the game, please note that controls are currently keyboard only: arrows or WASD for movement, space bar for attacking (single press for normal attack, very long press for aimed attack), tab and arrows for changing weapon (press tab again to close the weapon selection menu).

I’ll appreciate any leads, this bug has been really frustrating over the last 2-3 days...
Thank you!

Pages: 1 2 3