#semester 1

[cloverdaledevupdates]

Summary of Tasks for Semester 1

Created the game's dialogue system

Created the game's inventory system

Created the game's interaction system

Created the scripts responding to the player's interaction

Created a prototype for the present minigame

Created the game's flag system

Made adjustments to the scene transition script to match with persistent data

[cloverdaledevupdates]

Week 12 - (1/12/2025 - 5/12/2025)

Final Reflections for Semester 1

After last week, we all had a collective moment of disbelief - we'd accomplished a great berth of the mechanics that the game required, and we had two whole levels built out, as well as Reggie's house almost being completed! We were finished for the semester!

Now that I finally have time to truly think of what comes next, my mind is completely filled with everything I want to work on next. The list of things the game will need next semester is still gargantuan, but at least its many small things to implement, instead of designing core systems! One of the biggest priorities to me now is animating the UI and moving the camera in/out during conversations; I feel like this is a small aesthetic choice that will really make the game feel polished, as well as potentially adding customizable camera angles during key conversations to enhance the storytelling and effect desired from the writing.

While achieving this, I also want to edit the DialogueLines SO so that the characters appear in a typewriter fashion as well, and perhaps I can even use that to add audioclips to have characters 'speak' during their dialogue! I also obviously need to finish the present minigame's integration with the game. Lastly, I really want to make a node editor view to make creating dialogue trees easier for the non-coders so they can take part in the designing process as well! I always have the big list I made prior to check if I've missed anything.

I'm really excited to keep working on this next semester! Developing this game, with this team, has been a wonderful experience, and even though there have been lows to match the highs, I'm still ready to face the second semester's work with verve and passion!

Signing off for semester 1!

[cloverdaledevupdates]

Week 11 (24/11/2025 - 30/11/2025)

True to Week 10, the only things I really needed to do this week in terms of the project itself was bugfixing that the NewGame Flag wasn't refreshing when the player quit to menu and started a new game - this was easily solved by adding to the code connected to the new game button in the menu and making it set NewGame as true. The biggest point of contention was that at one point during the previous week, I had made a horrible merge to main that almost overrode my teammate's hard work; unpicking the changesets to save back to main and finding which empty shelves were preventing me from undoing those said changesets were a hassle to deal with after having just woken up that day. I think this tells me that exhaustion is not very good for making sound development decisions. Luckily, I was able to revert the error and merge properly, saving all of our progress.

Finally, I also reworked how the Rabbit and Frog moved when Reggie spoke to them, using flags to change their dialogue dependent on whether or not Reggie had spoken to the victim, acting on the feedback we had received at the confusion of why the NPCs reacted defensively if they were spoken to before even speaking to the victim. Now, this response and their teleportation is locked behind speaking to the Goat first! They now move individually, instead of together.

Small but meaningful bugfixes this week! It already feels so much more distinct and professional than it did only a few weeks ago!

[cloverdaledevupdates]

Week 10 (17/11/2025 - 23/11/2025)

Milestone : External Testing

This week was when everything I had worked on previously really came together like a game should. After asking for advice on how to manage it, I created a Flag Manager script that would instantiate a Flag Data Scriptable Object when the game was opened. This can be changed later, but right now this is the optimal setup so we can test the flags from scratch for debugging purposes.

By utilizing enums to account for each instance in the game that would require objects in the world to activate, deactivate or change, we have great control of the game's progression and can also replace a lot of the unsustainable methods to facilitate that progression quite easily. I made several scripts to affect change in the game world once flags were set, and scripts that would set flags after an action was taken! This took a while to account for all the different types of things that could happen to make a flag set (or unset), but eventually I assembled all of the previous scenes with these scripts in hand!

These are the current conditions in the game so far:

And here are how they can be affected:

The possibilities really are endless with the Flags in motion! It was very exciting to put things together and see them work so reliably and clearly, a massive difference from the chaos of debugging in the earlier days of the project.

Finally, as a last set up for next year's work, I created a prototype for the present minigame I had struggled with in the back of my mind after all of this time. Now that everything worked better and I understood how the systems would interact with one another, I was finally ready to take a crack at it.

It's basic, but it will be a massive timesave for next semester!

After that, that should be the bulk of the work I have to do this year! Next week will likely only consist of bug fixes instead of adding anything substantial.

[cloverdaledevupdates]

Week 9 (10/11/2025 - 16/11/2025)

After the internal testing, we recieved a lot of valuable feedback from the people who tested our game! More importantly for me, I got a bunch of bugs to review and fix. I was still happy to see how much people liked the art and the UI, since Charlie had been able to implement some of Lilli's UI elements into the game, making it look even more polished!

At some point due to the character controller being changed from using a Rigidbody component to another physics-based method, the interaction stopped working. This is because in Unity, a Rigidboy component is needed for trigger colliders to function. This was a quick fix; I attached a Rigidbody component on the Sensebox and disabled everthing that makes a Rigidbody actually effectuate an object.

Secondly, there was a bug where after the player completed the game once and tried it again, the dialogue with the victim at the end automatically played as if the player had collected both of the suspects. This was due to the boolean of that particular Dialogue branch not resetting after the end of the game, so I added a quick addition to the end of playtest script to reset it once the player had completed their task.

Lastly, I reworked how the DummyNPCs worked. Since their implementation was clunky, I decided to replace their function with the Cutscene Manager, which I could use to display dialogue without needing the player to physically interact with anything. I added a 'can Interact' boolean value to the interaction script, changed the interaction condition to whether that boolean was true at the same time that the use was pressing 'F', and then I could disable it when the Cutscene Manager was playing dialogue separately from the Dialogue Manager to avoid having to deal with a double input. Having this work out felt like a total victory! Since the Cutscene Manager's setup was so similar to the Dialogue Manager's setup, it also meant that it was flexible for any future alterations. Things are almost ready for the external playtest!

[cloverdaledevupdates]

Week 8 (3/11/2025 - 9/11/2025)

Milestone: Internal Testing

Finishing prep for internal testing

This week was spent testing the progression of the newest version of the game to ensure that everything was smooth sailing for the internal testing milestone. The first thing was, in order to account for the new addition of the city scene, to ensure that the bike scene transition would update itself based on whether or not the player had been to the park or not.

To do this, I used static boolean values to monitor whether or not the player had used the teleport once before, which would automatically switch the string to the city scene's name after the boolean value flipped. This was a very bad idea, and there are better ways to go about this so that you can actually play the game multiple times without needing to close the executable out everytime you want a new working play session, but it would function fine enough for the playtest. It can be made better later, it just needs to work now. Which it did! The transitions were stable.

Next, I set up my DummyNPCs around in reggie's house, in the park and in the city to trigger dialogue as needed.

Then, after I ensured that the logic for the conversation with the victim in the city scene worked well, I added a simple script specifically for the testing that would switch to a new scene thanking the player for playtesting the game, making it trigger after the player stepped on a platform in front of which suspect they believed to be the culprit, these platforms having appeared after the player finished their last conversation with the victim after collecting the suspects for questioning.

I also ensured that the scripts involved in triggering the DummyNPCs locked the cursor as well so that the player couldn't rotate the camera during a cutscene.

After painstakingly sweeping through the game a few times to make sure nothing else was broken, I realized that the way that the DummyNPCs were implemented were a big problem if the player came into the invisible collider at an angle. The core of progressing the dialogue (more importantly, freeing the player from the cutscene after its done) wouldn't be possible if the player's sensebox wasn't overlapping with the fake NPC, which wouldn't happen if you got caught in certain angles. At this point the only times this would happen poorly is in the city scene and the park scene, where we get extraneous internal dialogue from Reggie, instead of the more important context-laden dialogue that we could directly place the player into in the beginning of scenes or once they enter a particular doorway.

Since it was so close to internal testing, I decided to just axe these extra DummyNPCs and keep the more important ones that played more consistently. Again, they could always be reworked later, but if they impacted the playthrough negatively, there was no reason to keep it in.

All in all, I was happy to get over this leg of the journey. Now I needed to think a little more persistently.

[cloverdaledevupdates]

Week 7 (27/10/2025 - 2/11/2025)

This week wasn't too productive in comparison with the prior week. Just some more tinkering on the Dialogue Manager, some research and insight into what we could potentially look at implementing next semester, and adding a way to have dialogue trigger automatically in some situations without the need for interacting with an NPC.

I created a large list of all of the tasks that the finished version of the game would likely need, just to keep track of things in the future.

While tweaking the Dialogue Manager, I also went back to the Park scene and updated the Dialogue Blocks with the more recent dialogue that Lilli had shared.

After being more satisfied with the flow of the Dialogue Manager, I began creating a Prefab that I entitled 'DummyNPC', as it contained an invisible trigger box attached to the front of an invisible cube that had the NPCTriggerBehaviour script on it. I added a script to the trigger box that would activate the dialogue canvas and disable the trigger box once it detected that Reggie, tagged with 'Player' entered it. Afterwards, the player need only press F to interact and the dialogue would continue! I used this in tandem with Lilli's concept to contextualize how Reggie would know that a theft had occurred at the park in the beginning of the game, via a TV report. This could also be placed in the beginning of scenes to hear Reggie's internal dialogue.

By this time, the City scene had at least one NPC and all of the placeholder objects, so I assigned the Dialogue Blocks to them and set up the branching conditions, as well as a script that would trigger the Spoken to Target boolean value in the City's Dialogue Branch once two specific NPCs were spoken to.

The script for how the Banana Case would go was also helpful in generating a flowchart of progression for how the player should complete the levels.

I then debugged the scene a little to make sure it worked, even using paper to help the process to ensure I understood exactly what was going on. It seemed that there was another double input bug in the city, but it didn't seem to stem from the Manager's script for once. I decided to learn from the mistake of the past week and call it a day, leaving it to be reexamined with a set of fresh eyes later.

[cloverdaledevupdates]

Week 6 (20/10/2025 - 26/10/2025)

Milestone: Vertical Slice

The efficacy and productivity that I expected this week to bring was very suddenly obliterated over the weekend by a phone call from back home informing me of a bereavement of which I had no means to console my family except over a 3 hour time difference on the phone.

So I think it is a little permissible that, for a while, the rate at which I was able to complete my duties became considerably inconsistent.

One of the first additions to the project I made this week was to make a script that had the Interactable interface applied to it, so that when the player interacted with it the scene would be changed to a scene in the project that held the name shared with the string value inputted into the Inspector. Ben and Jess had worked really hard to create a wonderful scene of Reggie's home, a warm and cozy environment that the player would begin the game in, acting as a hub that they can return to in order to consolidate the clues they collect for their big case. The player would then use the bike as a diegetic means of transportation to reach the scenes of each investigation. For the vertical slice, this would be the park.

I then spent the rest of the week up to a few hours to the Vertical Slice showcase setting up the park scene's dialogue and pickups to align with the progression laid out by Lilli. In retrospect, I don't think I went about my process of filling out the scene in a very helpful way; I got frustrated at bugs frequently and ended up pulling an all-nighter trying to address two critical errors: one was that when the player interacted with an NPC, it counted as two interactions instead of one and as a result it was skipping dialogue. The second one was that the branching logic wasn't updating sufficiently when the player returned after picking up the decoy clover and speaking with the second Bear officer

This ended up taking far too long to fix than it should have, and I think my affected mood and disposition were the greatest factors in this, since the time it took to solve issues later on were highly decreased after I started to feel a bit better. On the day of the showcase, I felt deeply ashamed and unhappy with the quantity of work I had managed to display for the week; it could be charitably said that I became quite visibly displeased during the lecture. Luckily, my teammates were completely and entirely supportive of my situation and struggle to accept that the work was not what I had wanted it to be for that time frame. They gave me great comfort and sanity-saving understanding and compassion. It was thanks to their support that I was able to move forward with a less burdened heart and a clearer mind.

Being upset at the problems that arose wasn't going to fix them; systemically testing and debugging to find the source of the problem was. So instead of trying to intuitively find out the single line of code that was keeping the logic of the Dialogue Manager from working the way I wanted it to, I drew generalized lines around which methods could likely be involved in the bugs, and slowly debugged to see what exactly was happening behind the scenes. It was then that I could find out that the Condition checking for Reggie's inventory wasn't happening fast enough (again) for the conversation with the officer.

I tried a few different call orders to circumvent this, but brute-forcing the check in Update if the branch had been tested for once turned out to be the most consistent and effective way of handling it. It was a small but warming joy to see the bug that had just a few hours ago been causing me so much strife disappear from slow and steady work.

This incident made me rethink how I went about debugging this project mentally. I would have preferred it not to have been triggered by events of loss, but I can not effect every aspect of life. I can only control what I do. And for this week I chose to speak openly when I struggled, and to be comfortable, in the case that when steam, power and energy fail, to move steadily, knowing that progress will be reached eventually.

Now that the milestone had ended, it was time to work on the second half of our intended scope for the semester: the first real level of the game. While Charlie was working on reworking the controller and cameras to move more fluidly, the 3D artists were working on crafting the new scene for the case to take place in: a sprawling city street reminiscent of New York. This means a new script, and a new flow of progression to generate and set up. I created Dialogue Blocks and Speaker SOs based off of the script Lilli wrote for the 'Banana Case' and then spent a little more time tweaking the size and range of the Sensebox to better accommodate for interaction, since it was a bit tricky to find the right angle to speak to the officers in the scene prior.

That roughly concluded the work for this week. A new level to work in, a new mindset. Time to continue!