Call of Duty: DevOps (Patreon)

'look only at the game', and on sunday i had a fantastic day of doing just that and flying around with a ton of little pieces that would have taken forever to try and organize/delegate, just vibin', and isn't that how it should be?
absolutely wonderful.

one of the big reasons i was able to fly forward is because i solved one of the biggest issues i had with making scenes before. we've got a lot of new message art coming up, and exporting it used to be kind of a pain. in the demo, our exported character art folder looked like this:

so, what happens here is, once a message art is done:

• i had to spend time displaying every face, naming it, saving it, and then going next for as many combinations as i think i might want to use in some eventual cutscene.
• and if i make a small change to the art, i'm gonna have to do that all again. 
• then when it's time to make the scene, i need to use this folder view, which is the built in Microsoft Windows Preview Pane to quickly view each face and choose the one that i think best suits the dialogue and scenario
• and if i have the right eyes but want a slightly different mouth, i better hope i remember exactly how i named all of them so i can jump right to it
• it is a huge pain 
• it took a lot of time
• it was not very fun to work with
• help

but now, our export folder looks like this:

there's a lot fewer files, and you can see the base art is only exported once. the things that change, like the eyes and mouth, are all in separate layers. and the base art is a faceless demon like in your nightmares:

there are a lot of advantages to doing it this way:

• fewer files
• smaller file sizes (~12kb vs ~500 kb)
• i can automate the export to spit out files per layer instead of combining and naming manually, which is way faster
• i can use any combination of expressions instead of just the ones i pre-exported
• while we're at it, let's separate the eyebrows from the eyes to give us even more flexibility in expressions
• we can get extra bits like sweat drops, and blush in easier than before

disadvantages:

• how do i possibly see what face i need to use in a scene?

that disadvantage is the main reason we didn't do this before, though believe me i considered it so hard when we were first working on the demo. it turns out maybe i should've done it all along, because writing a scene to help compile these images for me was way easier than i expected!

in specific part, it was only as easy as it was thanks to the work nicky had already did on blitting messages art together. blitting means we take one image and stamp it onto another. we do this in the demo for things like sweat drops, but also sephany's wings so that we didn't have to export her expressions both with wings and without.  but, because we were stamping directly into preloaded memory, that also meant that if you didn't close the game and reload everything, there was actually a cat where for one or two specific messages (depending on her facial expression) sephany would actually regrow her wings for a moment despite having had them cut off.
oops.

because that was a larger structural issue to do with loading, we couldn't fix that until taking a hammer to the preloading system, and then a second one to our message blitting. in fixing that, nicky set it up so that we could actually blit a chain of images together all at once freely-- for instance, the body and wings and eyes and mouth-- which is what opened the door to this now!

so i took a couple days and hijacked our old test scene for animations and made something a tool for myself. actually, i was having trouble for a bit because this relies on the mouse support that perry has slowly been propogating throughout the script, and for some reason i wasn't able to click on random areas of the screen, shown in rectangle-testing here:

i realized with a lot of pain that the top left was unclickable because of the at-the-time-invisible FPS counter:

all of the words in that image really sum up how i felt about that.

apparently the big rectangle in the middle was-- something about an error message? anyway, hewaje swooped in to save me from it. so really this was a huge team effort. anyway, the tool looks something like this:

so i load up the game in test play mode, jump to this scene, and now this will let me compile message art for any character on the fly, showing me the files that exist for them and letting me click between them like-- like a picrew, basically? wait, shit did i just make a picrew? i,...........

um, i'm going to choose not to think about that. anyway, once i'm happy with the expression, i can click the button in the bottom right to copy it to clipboard and easily insert it into the scene.

there's still some improvements i can think of there-- maybe we can make some more over time to speed this up even more! because there is going to be a lot of dialogue in this game (i don't learn my lesson), and that means a lot of message arts... all the incremental advantages i can get, especially early, will probably matter a lot in the long haul. 

i've also been able to backport some of these mouse and folder-scanning improvements into our animation test scene to help out there too:

before, this had to be loaded into an existing battle, the unit had to be in that battle, and we had to control it through console script commands. now we can just see every armature and animation with a couple clicks. 

so this is great, and i'll be thinking about what else we can do to speed up eventing and scene creation in the future too.

that said.
part of thinking about the feasibility and speed of development also means acknowledging when i need a little bit more help getting all of the assets together-- because isn't that always most of it anyway? 

i talked previously about how i am looking for someone who can help out with tile art and that is still true. i am also now in the market for someone to help out with sound design. i'll probably be querying the greater community as well, but if any among ye are skilled in and interested in helping out with that, or if you have the time and energy to commit to learning, shoot me a message. 

sound design has joined the list though because, while it isn't the momentum-killer that tiles are, it has historically been a weak point for me. i tend to develop mute or to music and forget about sound until going through and polishing, and although i'm capable of doing sound work myself, i would like it to be done more thoroughly, and it sometimes does become a task that i have to stop and attend to. so, i could use a more dedicated hand. 

it's just bottlenecks to solve... i hope to get this altogether well enough to the point that we can just blaze through content for you all like we were eventually able to do with postgame.

as always, thank you for being here <3