IGB220 Blog

Nathan P Blog

[nathan3197]

Stuff from the textbook

Chapter 7 - Prototyping

A physical prototype is a great way to prototype as it is quick and easy to test out an idea or mechanic as it allows you to make quick changes to the original idea or mechanic. Prototyping is not limited to gameplay feature much time people and myself have prototype UI for the end-user to experience and 99% of the time my original UI get scraped after myself or others found it too difficult. One cool thing I found out from reading this chapter is that even first-person shooters go through physical prototyping. Great way the book reference is not just to make a physical prototype you can make a diagram explaining your mechanic. Prototyping levels have been created with diagrams or with a white board. when I was working on assignment 3 I was using a whiteboard to prototype different enemy types as I can quickly sketch up the player screen and easily add each enemy type to that screen sketch up. this gave me a scope of how to refine each enemy type for our game.

Chapter 8 - Tools to prototype and level editor

Unity is a really good way to prototyping mechanic. Unity Work on any platform. and you can easily install addons or plugins to help you achieve your prototype, iv even used unity to make a few prototypes during the semester breaks. But there also other engine that will help you make a prototype like gdevelop, scratch, game makers and unreal engine. if you are prototyping levels these engines allow you to create levels really quick without having to use external program to make the level. you can just make a quick change and click play and test your change. 

Exercise 3.3: Interaction Patterns

Exercise 4.2: Dramatic Games

Ch 10 Is Your Game Balanced?

Ch 6  Brainstorming

CH 11  Improving Player Choices

[nathan3197]

Assignment 3 Postmortem

Pre-production:

We didn’t get enough time to do a lot of planning besides taking apart my asteroids game and stripping it and only keeping the asset (code) that we need for this game.

Production:

For me, I wish the group would have spent more time working on the game itself beside me. but with the short amount of time, I was able to get the core gameplay of the game working and even added support for the switch(even though it is just an android version of the game running on a modded switch) this shows that our game can be run on weaker hardware and mobile hardware. But we couldn’t add everything that was found on the one page. most noticeable more enemy types, an item shop and customisation. These had to be cut. But the current version of the game is polished from a code point of view. 

Post-production:

I wish we had spent more time polishing the assets and replacing the placeholder assets. as a group, we only made menu icons and a boss sprite. All this can be contribute to the short amount of time we had to make this game.

What we can learn next time.

We really should have more time in all areas of production to allow us to create a game that we in vision.

[nathan3197]

Improved input detection

So for the quality of life improvement and the sake of having everything into one version instead of having two versions (1 for keyboard and mouse and 1 for the controller). So what do I have to do? make a variable that switches on or off if you are using a controller or Keyboard and mouse. So I made this variable iscontroller when 0 only keyboard controls are being used and if it a 1 using controller controls. Code   

Now for the sake of the teachers and anyone who is not going to use switch controllers (or play on a switch that runs android) to play our game so iv made it so when the game first detect controller input it will ask you what controller you are using. What you select will make the GUI show your controller button and the button layout will make a scene for your controller.

Now with this improved controller input, we can switch input at any time and the game will auto-detect and adapt the GUI to the current input you are using. IT IS ALSO FAST TO DETECT TOO.

[nathan3197]

Addressing playtesting feedback

Right time to address the feedback from playtesting.

The first thing to do is to give the player more visual feedback when they do an action or an event come up (i.e boss). The first thing I did was make the enemy's, boss and the player flash red when they take damage. Jodie made the damage sprite for the animation. So what iv done is when the player takes damage from anything I change the animation to the damage animation and once that damage animation has finished, switch's back to the normal animation.

This looks much better than having no visual feedback when the player gets hit and you shooting other cars.

Now to fix the health balancing issue.

what iv have done is change the fuel pick up from 2 to 5 and didn’t change the health decreasing over time. As that takes 0.5 per 2 seconds.

Jodie asks to add a distance meter that fills up over time. once full spawn the boss. As this was in the original one-page document. 

How does the distance meter work?

Fills up over time(every second add 2 to the distance variable (max 300) .

Once filled up spawn the boss.

Decreased when hitting objects.

This was easy as i can copy the health GUI code and use it for the distance meter. 

Next is to tell the player when the boss spawn. Iv just made a text element spawn on-screen when the boss spawn.  

The last thing I work on was menus (with controller support two) How would this work with Gdevelop? With variables! my original design was the button in a grid so when you move your stick in any direction you would highlight the button you want to select. I only needed 3 buttons. Start, how to play and controls. A line them up were pushing up on the left thumbstick would select start button, pushing left on the left thumbstick would select a controller and push right on the left thumbstick would select how to play. This is great as you can get to any button easy with one push of the stick or Dpad.

Code side of things work. basically made it when the left stick or Dpad button been pressed to change a variable and what number that variable highlights the button and if the A button is selected. change the scene.

Here is working in-game.

Code

One thing I have noticed is when pressing the A button there is no delay when you can press another input on the controller so if your next screen has a button and the user is still holding A when the next scene just finishes loading the first button is selected will get press. in my case, I would just go back to the start screen. easy way to combat this is to place a delay timer once the timer is up allow input from the controller. At the start of the scene I set the elect variable to -0, 0 is the first button in the list.

What about the other 2 menus? The controller menu just changes the screen to show the controls for our game. currently, just show the switch controls.

The how to play button show you everything you will encounter in our game and explain to the player what they do. Using the bumpers will change the page. I quickly made these pages up in the photo shop as these would be placeholders until we make the final version of these pages.

So with that our game has menus. Besides a pause menu in-game, I didn’t have time to make it and it would take a lot of reworking of existing code to pause everything in place. maybe in the future, I will look into making a pause feature.

Next to make the final boss ending into a cutscene instead of instantly going to the win screen.

How I wanted this little cutscene to work is when the boss dies these events will happen in order.

lots of explosions will spawn where the boss die.

Player input will be lock and hud will be disabled (to stop the player from breaking the cutscene and to tell the player a cutscene is playing)

The player car will dive to the top of the screen.

Once the car hit the top of the screen the screen will fade to black.

Show the win screen.

How would the logic work? basically scene variables and timers.How it look in game. Look very nice to end the level.

[nathan3197]

Play Testing

Our first playtesting session went alright. The first thing we noticed was that everyone wanted to play the switch version because they found it cool. during the first playtester, iv noticed that there is no visual feedback when the player shoots the enemy's cars. same with the player getting hit. those are red flags in my book as communicating with the player is important to tell them if what they are doing is the right way to play. Without this important visual feedback, the player would just avoid the enemy's cars not being able to pick up the gas can to give themselves more health.

Another thing that I did notice was that the player didn’t gain enough health from pick up and they would lose health faster than they would gain health. So this was a gameplay balancing issue that needed to get a fix.

The last thing iv noticed during my playtesting is that the tester wanted to have more enemy types other than a car going down the screen and the boss. In a future version of the game, there would be more enemy's types but due to the short time frame, we couldn’t add more enemy types in time for this playtesting session.

Overall from this playtesting session, I found out what the tester wants to see in the future version of the game.

Visible feedback when hitting enemy's and when the player gets hit

Different enemy types.

Visual prompt to tell when the secondary weapons can be used.

Story (explain in the game)

Balance health pickup.

more hazards. 

I will be addressing some of the issues in my next post.

[nathan3197]

Switch Port*

Yeah that right iv gotten the game to work on the switch. (*technically it a android version of the game running on a modded switch) First thing came to my mind is how can i add controller support to the game because if i can do that all iv have to do is replace the controls from keyboard to button press. good thing that G-develop has a controller extension that i can enable. thought the extensions tab

Once iv enable that now to add controller support but adding conditions to detect input from a controller.

Since my modded switch runs android and the joycons just emulate a xbox controller it was easy to set the game up for xbox controller inputs.

after converting all the keyboard controls to xbox controls i have a functional version of our game that can run with controller. with this short amout of time just right before play testing i wasn’t able to set up a system to auto detect if you are using a controller or a keyboard. so i had to have 2 version of the game one with keyboard input and one with controller input. which isn’t best thing to do as maintain two version will be hard. In the future i will need to make a input system to allow more then just 1 input to be used.

So i just export the game as an android APK and install it to my switch and it works.

[nathan3197]

Assignment 3 Development progress (before play testing)

This week iv spent the day stripping most of my asteroid code that wasn’t needed for this game and only keeping the weapons mechanics system that iv made.

What got remove asteroid: spawning and logic Zones Player vertical movement What i had to add in Car spawning hazard spawning boss Let start with car spawning in we wanted to have cars spawn in a random lane to make the players move around the screen in order to shoot the car. so iv come up with a spawning system that pick a random number (1-4) and base of that number spawn a car in the lane. (the highlighted red X is the spawn points of the car)

What happen when the car spawn? That a good question, The car when on spawn will take 4 hits from starting weapon that the player start of with and a one hit when the player unlock the second weapon. Also when the car spawn they will move down the screen if the player hit the car the player will take damage and loose some distance to start the boss flight.

hazard spawning This was going to be a bit difficult because we have many different type of hazards that we want to spawn from people walking across the screen Road works Pot holes Fuel can (pickup) Coins (pickup) How would all this work? with some good old fashion brain storming on a white board. Brainstorming Road blocks How i wanted road blocks to work is they will spawn just like how the cars do but in more places. From what i was drawing on the whiteboard is different roadblock design and how they will look like on the screen and they size compared to the player.

Next things to brainstorm is people walking across the map and different variation of people.

Now let add this to our game. I first started off was the logic. basically it a random number get pick (picking the location of what to spawn) then a another random number get pick to see if it a road work or power up. if it a road work or power pick and random number to decided what variation of it will be. Code of the hazard

He how it look in game

Now to fix the boss that mat half finish. What Mat has done is make the boss spawn when the player hit max distance and make him move left and right and shoot. All that i have to do is make the boss take damage, give make the boss bullets give damage to the player, stop spawning in the cars. First things was making the bullet deal damage to the player. that was easy as all i need to do was create another object call boss_bullet and make it so when it collide with the player the player take damage.

Now to make the boss take damage. same with the car make the player bullet give damage on collision of the boss.

once the boss hit 0 or less health it end the game. to stop cars from spawning iv made a scene variable called (Boss_alive) when this is set to 1 (when the boss spawn) all cars will stop spawning. This allow the player to focus on the boss and the hazards. Now i wanted to do more to polish off before our first play testing but this what i gotten done in the time frame before the play testing.

[nathan3197]

Assignment 3 Development Progress

This week we had to form groups to create a game in Gdevelop, Mat and i wanted to work together, the assessments require groups to be in 3 so we saw a post from another students and both mat and i like the concept but we saw that we can readjust my second game code to fit into this other students concept. So we form a group. Mat and i started talking about what needed to be removed from my second game project and the same with what need to be readjusted.

[nathan3197]

Post-mortem

Pre production

I knew from the get go what i wanted to create and i know that i could get it down within the time limited and in Gdevelop. i didn’t spent a lot of time in here as i know my concept of a multiplayer racer would be easy to create. 

Production

During this time i spent most of my time fine tuning the player control once i had that down i can just copy it to the other players, one of my major issue was making sure the player doesn’t cheat a lap by just going backwards around the track. The great thing was i went and have a look at existing racing game and i found they all use zones to track the player on the track, this was easy to add to Gdevelop. Though out this phase i didn’t have to cut any features to the game but i learn new ways to handle curve wall is by placing lot of squares collision boxes around the assets and made moving around the curve wall smooth.

Post Production

Not a lot of bug i needed to fix but got a lot of good feedback during playtesting. overall was happy how smooth this production cycle went. The one thing i would use in future Gdevelop project, start with something simple then once i have achieve making it then try to do more advance things.

[nathan3197]

Play-Testing: Car racing game

Play testing went alright the first major issue iv found is if 3 or 4 players are playing at the same time the keyboard would get very cramp and feel awkward. easy way to fix this issue is to by adding controller support. 

Play Tester 1

From what this data show was the game felt well and the flow of gameplay, game play balance was great, the tension of racing other player was satisfying and the game was fun (what people found my game fun, im shock) the one thing the would make this game more replayable is playing with more people or with different people, (online / lan mode would make this game more replayable).

From the questions that the tester has answer he found the game was competitive which is one of my goals for this game, one the issue that this playtester noted was if he touch the wall there car would get stuck on the wall and made this playtester to be frustrating. also allow the player to chose there car color before the match would be cool. The only idea that this play tester want to see is power up (this would make the game more fun depending what type of power ups are included) 

Play Tester 2

This play tester Felt the game play well and had no issue really,

from the question that this play tester answer, What work was the game was fun, fast paced and had completive elements, this is great because that what my original design was meant to capture. What need to me improve is the wall collision, this was a common issue with all play testers. and again with the new ideas power ups. 

[nathan3197]

GDevelop - car racing game

Car movement: So far with the template the car instantly movement speed is instantly set to 400 when the forward key is press. that not how a real car drive. a real car build up speed until it hit is max speed. For my game i want the players car to behave the same as a real car.

Making it: My though process of how im am going to do this is to have value that is store on the car which will be call movement speed. this value will be use to add force to the player car. and when the forward key is been press it will add 5 to the player movement value then that movement value will add force to the car.

Code:

Gif of it working:

Bugs with it:

so far with this current version of implementation when the players want to move backward while moving forward they will stop dead still. i could say this is a feature call “hand-break” and move on with this. so far for this prototype this is fine.

Track:

need to create a track for player to drive around. lucky itch.io has some good assets that i was able to build a track. One problem i have to face is the player can drive around the track and though any walls. So i need to set up collisions

Gif of bug:

Custom collision has been set up for the track pieces. wasn’t to hard just a bunch of squares

Time to give player collision:

The easiest way i can make the player collide with the wall is to give the player a 2D platformer behaviour and disable everything beside collision. By doing that the player car can now not leave the track. and the other thing i did was every time the player hit the wall there current speed get divided by 2.

GIF:

Laps:

How im going to do this by creating zones when the player drive though them their zone counter will go up. So when they come back to the finish line and if the player has hit all the zones in order a lap will be completed. This is very slimmer to how modern racing game handle laps and tracking who in 1st place.

Now for Player 2 or many a player 3 or 4? This was quite easy as i have already setup the player controls and gates. all i have to do was change the key bind for each player and assign them a variable telling the code what key bind are for what player and the same with the gate. i added the ability for players to hit each other so they can nock each other to the wall or let them team up to make a road block.

gif:

[nathan3197]

racing game - Pitch

[nathan3197]

Asteroids Post-Morten

Pre-production: I made a concept that i thought it could be possible to make in Gdevelop as i didn’t want to recreate the mistake i made in the first game of making an concept that Gdevelop couldn’t handle. The first thing i wanted to do first was the level design, i did a few drawing of how i wanted the level to be lay out, good thing was i had to worry about and if the player in a zone i can tell Gdevelop what to do. i didn’t spent much time on this as i wanted to get into making the game straight away. But my concept was solid enough to have a start and expand if things go well. 

Production: This stage went quite well more then i expected. i had a few issue but that was cause due to me miss spelling some names, was able to add most of what i wanted to the game beside the dynamic difficulty as that was taking more then 1hr to get working and balancing did add some cool feature to the game by out of curiosity, i wanted to see if destroying a large asteroid would be able to spawn smaller one and first try it work and felt fun destroying multiply of them at the same time.  by the end of this stage i was able to have a game playable from start to finish beside having all the pretty visual. which i will add during post-production  

Post-production: This stage i had some issue getting assets to work, as i had a lot of play space to fill with assets, my main issue was getting all the background object in the right spot and dealing with no hierarchy to manually select object behind another object. This resulted in frustrations but was able to get everything in place with of a method of working backwards. by placing objects that are back of the background like sky, distance tress, grass and etc. and then placing object at the front like bushes, god rays, clouds and etc, so i can move them easily around. What i also did here was adding visual and audio feedback to the player as they need to know when they shot has hit an asteroid or the player has taken damage. I did a lot of self play testing to balance out the weapons, player health, asteroid spawn and etc. Which help out for setting up play testing.   

What went bad:  Nothing really. beside that one issue where i change the player animation to another animation set and cause Gdevelop to crash. was an easy fix of removing the animations.

What went good: Having a simple and not overambition concept to start with help so much when it came to creating the game during production. as i consider the limitation of Gdevelop when designing the game.

What could be change for future project: Don’t spend 15hr working on a prototype. beside that nothing really as i had address the issue i face during making the platformer game.

[nathan3197]

Playtesting

Time for playtesting, I had 2 playtesters and the information is helpful to help improve the game even further. Playtester 1

What the data suggest is that this player the game wasn’t understood very well. i can come down to 2 things lack of a start screen with instruction of the goal of the game and the mechanics or when the game starts the player hasn’t had enough time to get use to the mechanics without being in danger.

What works:

Weapon fire rat:e balances for the different weapons types. you don’t want a laser having a fire rate of a shotgun. it good to know that my weapons balance. Things to improve:

The player doesn’t know the core goal (getting enough energy to leave the planet) i have try to tell the player through the text on the top of the screen and it looks like the player hasn’t even noticed it, so my only suggestion is to have a visual prompt with audio cue when a player has unlocked a new area of the map and weapon and tell them though visual prompt with audio when they can’t access it. A start screen explain the object of the game would have been great for the user to see as this would explain how to play the game.

New Ideas: Boost mechanic would work as this can help out in certain scenarios where the players are getting trap by asteroids and the only way out is boosting through them.  maybe add damage to the astroids when boosting. all this would cost energy to help balance out the boost.  What do you think was the most exciting or fun moment for you as you played?: destroying multiple asteroids at the same time, the only way i think this was fun is because every time an asteroid get destroyed the explosion effect plays. collecting orbs is one of the key parts to gain access to new weapons.

Playtester 2

From what this playtester data shows that he enjoy the game but felt the game wasn’t too balanced. my original idea was to have a dynamic difficulty when the player energy bar gets more fuller.

[nathan3197]

Gdevelop - 30 seconds of fun and getting things ready for playtesting

Now i have a working game, now to make it pretty, fixing tiny bugs, tweaking mechanics and getting it ready for play testing.

Visual:

trying to create the visual for my game would be hard as i can’t draw so i turn to my good friend itch.io to find some good pixel art assets that will fit into a space them game. sadly i couldn’t get assets for a frozen planet. placing the assets was easy as i can replace the placeholder assets that i have made when creating the game.

Before 

After

Commutating to the player:

This is important to tell the player what they are doing and what going on. as an example, if a player is shooting an enemy that doesn’t die in 1 hit and the object they are shooting doesn’t indicate that the enemy has taken damage then the player will think that there shot didn’t do any damage to the enemy. this is one many problems that im going to run into when i do play testing.

First thing i did was when the player destroy the asteroid an explosion effect plays on the location of the asteroid die at. the same apply when the player get hit but i have change the colour of the explosion to red. from play testing myself and destroying multiply asteroid as the same time felt good seeing the explosion effect as it look cools seeing the destructions that i created.

The next thing i have to indicated to the player is the out of bounds.

how i approach this was to place a bunch of clouds in the out of bounds zone. as the playable area has no clouds. This will hide unfinish areas of the map (mostly just leftover assets for creating the level) 

Making the game fun: 

In the video documenting halo 2 development (VIDEO) a designer Jaime Griesemer said this “ you can have all the great graphics, and all the different characters, and lots of different weapons with amazing effects, but if you don't nail that 30 seconds, you're not gonna have a great game.“

When i came to playtesting my game i would stop every time i felt like I'm not having fun and ask my self these questions

1. What making feel like im not having fun

2. What would i change to make it bit more fun

These two question will help me find out the cause of me not finding this game fun. First time i did this was when i was testing the game from start to finish. the first 2min of game play i felt like i wasn’t having fun as there was nothing to do. turns out i had the spanner for the asteroid way to low so as the player can destroy them faster then they can spawn. What i did was to make them spawn faster to make the player feel bit overwhelm but not to overwhelm. After a testing changes to the spawn timer i found the perfect balance as when you start off you have to plan which asteroid you want to attack once you unlock more weapons you can start making havoc while avoiding asteroids.  

The second time i uses these question was tweaking the energy at first i found my self dying so quickly because i gave the player so little energy to start and the energy they get from destroying asteroids. tweaking it fix this issue after constantly stopping when i felt the game wasn’t fun and asking myself those questions.

After 15hr working on this game Here what the game look like https://games.gdevelop-app.com/game-3be16aaa-9568-4e40-9fbe-237ed0c41a52/index.html

[nathan3197]

Gdevelop asteroid prototyping

So this post is going to break down how i implement features into my prototype The world and zones So i want the player to explore more of the world the more time they survive and when they gain more energy. One of the major feature of the game is the player managing their energy so they can explore more areas and use weapons. So the way im going to limit the player play space is by zones. The player will gain access to more zones when they hit a certain amount of energy as each zone has different type of enemy's. If the player goes below a certain amount of energy the zone will be lock and telling the player they need to get out of their due to not having enough energy to keep the engine warm. once the player has filled their energy they can enter the win zone finishing their gam.e Here a mock up of the zoning system and level layout

Implementation

First Test gotten the start area and out of bounds working by having the zone collision collided with the ship to change the state of the ship. so if i want to do something to the player ship i just need to make it check the ship state before anything can be done.

End result: had no issue beside the out of bounds timer not working as intended all i need to do was reset it to fix the issue. now the player can’t leave the play space and leave a zone without having enough energy

  Moving the player I wanted the player to be move with my movement control (WASD) i had to change how the ship is move with the mouse to WASD. this was quite easy as i need to find a point to add force for moving forward and backward and change it angle with left and right.

Energy: Player will have to manager energy. if there energy hit 0 they will not be able to shoot and will start losing health.

making this was quite simple as it just value just change base of if energy = this Shooting weapons and going into a zone you can’t enter will make you lose energy. Play testing will help out if value for energy will need to be tweak. every time a player destroy asteroid they will drop energy balls that will give the player more energy. 

as the player energy get high and higher different asteroid will start to spawn base on what zone are unlock For player feedback there is a energy bar and status of the engine to tell the player that they can shoot

This what the level look like now after everything been added

Weapons There are 3 different weapons that the player can use Turret (require 1 energy to use) Shotgun(require 25 energy to shoot, will remove 2 energy) BFL(Big Freaking Laser)(Require 75 Engery to shoot, will remove 7 energy) This was easy as i just needed to add more points to the ship and change the bullet behaviour to allow to shoot more then 1 bullet type.

For player feedback there icons on the energy bar to show what weapon can be use at when the bar it the marker

Weapons working in game

[nathan3197]

Exercise 1.1: Become a Tester

Game: Cod:Moblie(zombie mode)

A brief overview of the mode: players have a small amount of time to gather resources to help them defend a base from the zombies during nighttime. Once night time comes players must kill zombies and keep the base health above 0 to complete the mission. Players can craft ammo and equipment with the cash from killing zombies.

My playtesting

At the beginning of the game, I had to find weapons and ammo before the timer went to 0. Spent time running around not sure what I'm doing as iv only played the game once. Once I found a weapon I felt safe knowing I had a weapon to defend myself from the zombies. Once the timer went to 0 the game teleported me back to spawn and had to defend the base until another timer went to 0. During the defending, It felt like my gun wasn’t doing enough damage as some of the zombies would not die within 1 ammo clip of bullets. Lucky I found out how to place turrets and they felt weak against the zombies. Once the round was over I felt like I just barely made it as the base was half health before the timer went out. Once the game ended I saw the result screen and I was at the top of my team I felt kinda good knowing I was the best.

Friend playtest:

My friend has never played the mode before so he going in blind. At first, he just runs around for a while trying to find a weapon once he had found one he started to find zombies to kill but since the round was in the day time he needs to find supplies like ammo after he had killed all the zombies in the area he started exploring the town that the game has spawn everyone in at the beginning of the game. When it hit night time the whole team got teleport to the base, while everyone was setting up their turrets, my friend just started to fire at the zombies wondering why the shots are doing little damage. Once he found out there was a button on the screen he was able to place a turret down. After the first night, I notice he started to follow the objectives markers on the screen which lead him to look to upgrade his turret. Once night 2 came my friend didn’t upgrade his turret as he didn’t know that was a thing. Day 3 he went to explore more of the map picking up better guns and more ammo but when it came to night 3 he had a hard time killing anything as he didn’t upgrade his turret so he felt useless to the team and blaming the gun being weak. After the match, I ask him what he felt he said but was confused about how it works.

Compare the two sets and what iv learned from the process

Both experiences were from a new player point of view. Both players quite didn’t know what to do at the being. In my session, I noticed the turret earlier than the other. While I finish thinking I just made it while the other believes he did no damage to the enemy. From what I learned is that a tutorial should be included during normal gameplay to show how the mode work, but also feel like the weapon balance between day and night is horrible as at day time all guns are good at destroying enemy’s while at night they barely do any damage but the turrets do most of the damage. Visualising to a new player what to do for a new game is important as it will give them less confusion when playing the mode. Note: I just found out that Tumblr text post can contain image (i never used Tumblr before this semester