Graphics Tasks/Work-Flow Discussion

By no means am I the person best qualified to discuss this topic, but I got in a brief conversation with @Kertit on Discord and I am pretty energized, so I wanted to put something up to maybe generate some discussion.


I have a long-running inside joke (with myself) where I sarcastically yell at graphics artists to revamp the microbe backgrounds as it’s something I always asked the dev team to focus on when I was a new face on the community forums. It’s something the fan-side of my brain would really like to see happen, but my dev-side of my brain started remembering that we oftentimes mention that the current way in which we organize graphics-related tasks isn’t very consistent (the common saying is that the art team lacks a way to get pointed to tasks, and a direction). I realized a few things…

  • It isn’t really clear how we are trying to revamp the backgrounds. We have a few concepts drawn out, but those aren’t really listed anywhere prominently, so no-one really knows what to do and what can be done.
  • We don’t have a list of tasks which can be done to help revamp the backgrounds, which might confuse volunteers and make them unsure of exactly what should be done. One task is to remove the parallax layer in some of the backgrounds which contains static images of things that should be moving, like groups of cells and un-moving bubbles which looks uncanny. This is mentioned like once or twice on Discord, but isn’t documented anywhere.

This then lead to me wondering if we might benefit from changing up the way we organize tasks related to graphics (and perhaps sound as well). We have a task board which is utilized occasionally, but it almost never leads anywhere:

After skimming through the listed tasks, I realized that a lot of the tasks can be rather broad and ambiguous. Here are two examples:

I am not saying that the people who opened these issues did anything wrong, but I can see a new member to the graphics team not being sure what these really mean. How do we need to improve backgrounds, and which backgrounds need to be improved? What is our goal for backgrounds? Where are we currently at? Why do we need to improve the organism statistics’ panel readability, and how? What is our plan related to that? Again, this isn’t the fault of the people who created these issues; @hhyyrylainen opens some graphics-related tasks, but he deals with so much work that he can’t really afford to stop and create a detailed list outlining what is needed for a graphics-related task. It isn’t necessarily entirely on the person requesting an art-asset to detail every single thing they need since that should fall on artists, but our artists do come and go and might take time off, so it can be really hard to get back into the flow of things.

Some long-standing graphics-related tasks have this issue. As mentioned above, it isn’t mentioned anywhere what exactly needs to be changed for backgrounds and what we are working towards. @Nunz mentions that creating graphics for currents is pretty easy, but the underlying current generation system needs to be reworked first. This isn’t entirely why graphics-related tasks take so long - we have other priorities currently than revamped currents and backgrounds, and some graphics tasks just require a lot of effort - but it is something we can improve on. As a whole, it might be a bit confusing understanding exactly what needs to be done, and how.


So that leads to the question: what are some ways we can better document art-related tasks? Is there even a better way of doing things? Am I unfamiliar with the process of developing graphics?

I think there are two things we can do that won’t take too much effort…

  1. Review Some Graphics-Related Tasks on GitHub and See If We Can Better Document Them - There are some broad graphics tasks that can probably be defined better; for example, the microbe backgrounds. It can help to break these tasks down a bit in the description so we understand what we are working towards and what needs to be done. I can volunteer effort towards this, but I would like the opinion of a graphics wiz in this effort.
  2. Start Utilizing These Forums More For Graphics-Related Tasks - 90% of graphics discussions happens on Discord. I do think that it is normal for graphics-related tasks to operate like this, since a lot of the process involves sharing progress and transferring assets over. But utilizing forum posts to draw attention to what is currently needed in a certain graphics-related department might be a good way of addressing things.

I can provide some sort of a template for a post that might reflect the above principles below.


Microbe Backgrounds (I’d post this on the forums to start discussions, then once accepted, it can become a description to an issue)

Issue - The current backgrounds look nice, but are flawed in the fact that various elements in the background layers which should be motile are actually static.

For example, in this screenshot above which I grabbed from the community Discord, the specks and bubbles all indicate elements of movement and activity, but once the player stays completely still, so do those elements. The result is an uncanny awareness of the parallax layers, which draws attention to the fact that this layer actually isn’t dynamic or moving.

Goal: The image below, a concept from Uniow, reflects the “ideal” set-up for our backgrounds. In it, the background is a generally constant color (for patches such as the seafloor, texture is also needed). However, above the player, there are sources of light which alter the background slightly, indicating movement, activity, and life. Combine that with elements in the game scene, such as currents, items, other cells, and compound clouds, you get a lively and polished scene.

Steps to Take/Task Breakdown - The above concept is obviously a pretty big task to take head-on, all at once. But we can breakdown the task into smaller steps that can be added iteratively, bit-by-bit improving the microbe stage backgrounds.

  • Remove/Alter Static Parallax Layers With Motile Elements - For example, removing backgrounds with static bubbles in them to remove the uncanny feeling of false movement occurring. This would essentially just mean removing a layer from the backgrounds.
  • Adjust Backgrounds After Parallax Layer Removal As Needed - If removing the middle parallax layer with all the motile elements reveals that some backgrounds need to be adjusted slightly.
  • Add Light Sources - Implement light sources into the game scene, so that the background changes in hue/tint.
  • Vary Light Sources - Have light sources vary in intensity, duration, and movement speed to create a unique and dynamic scene.

These would be checked-off and such as completed, letting people know progress. Notice that things are organized chronologically, so that someone with less-experience/time/energy might be able to spend time on a small but first step.


It obviously isn’t that simple, and this obviously wouldn’t result in an all-encompassing graphics revamp where artists blaze paths and get things done instantaneously. Again, I definitely am not the best qualified person to discuss this topic. But documenting tasks better, perhaps like above, might better help an artist understand what needs to happen. There could be a post on these forums outlining these issues, and then a GitHub task can be opened linking the post and briefly outlining the tasks which need to be completed. If we do this more consistently and redo some existing tasks to fit more alongside this work flow, perhaps we will see a better utilized task board.

What does everyone think? Any ideas? Am I neglecting to consider something? Is there a better way? Does this make sense at all? Again, if we want to re-organize the art task board, I’d be willing to help out.

1 Like

Discussing graphics in a dedicated forum post is definitely the way to go in my opinion, after which a final and concrete conclusions are recorded as GitHub issues and documented in the wiki (this one’s an important point). I think it has been that way for a long time but has since been a neglected practice. Microbe backgrounds for example has multiple related threads spanning years, which I’m also surprised to find out tbh (according to the search bar) with the oldest being 2016! Only problem being not much activity goes on, on those threads.

I do agree that GitHub issues need to be as detailed as possible, it would be ideal to have artists or proponents of such ingame changes to write them in some sort of a “proposal” format.

2 Likes

That template is great, and it might actually be worth making into a forum post outside of this one.

2 Likes

I also think that’s the optimum way of approaching things. Let’s see what a graphics artist has to say about this approach, as well as everyone else of course.

Glad that you think so! If people generally agree with the sentiments expressed in the original post, I’ll edit the header of the OP to include a general template guide (though now that you mentioned it, perhaps it can be a more generalized template for suggestions as a whole, such as for the sound team or even the suggestions website). Then, I’ll essentially copy-paste the microbial background proposal in an appropriate thread, then try to alter the existing pull request. That can be a soft start of this system, and I’ll try to encourage it to be used.


I also realize now that I keep a pretty active ear towards graphics-related tasks, so I think that I can ultimately serve as a “faux consultant” for artists in helping to identify tasks that can be done. I wouldn’t really have any input on the technical side of things of course, but I can try and pitch in.

1 Like

Hello there

As one of the graphic artist, I can tell I use Discord way more because it allows a waaaay faster sharing of progress and instant advice about current WIP.

BUT, I agree that it get lost pretty quick.

One other solution could be to use a brand new Discord Feature: Threads creation!
It allows some good and clear separation about one specific conversation without flooding the channel.

We’re talking about graphic art but I think it can be used for any dev branch.

I’d like to point out that @NickTheNick opened that issue purely because he was working on it, so it was to show other people what’s happening. I think now with Nick being inactive again, that issue should be split up / reworded a bit as reading through it is pretty clear that the graphics task is making like 10 icons.

I opened the improve backgrounds issue because people kept complaining that something should be done, but no one has any concrete ideas. That complaining part is literally why that issue exists, I feel like the issue could be closed if people promise to stop complaining about it and instead either not say anything or make concrete changes.

I always agree with this. Discord information basically disappears in 24 hours, so I’m always in favour of more stuff happening on the forums so that future developers can see what has already been discussed to not have the same discussions again and again.

I’d like to point out a huge problem with this: this is a massive change that requires programmer-artist pair to work on this pretty closely. The artist will need to first create the mobile assets while consulting with a programmer on how they can be done. Then the programmer needs to put those things in the game. And only after that can the background redrawing happen. If not all of these steps are completed then the game is worse off than if things weren’t started at all.

This is not a graphics task really, as the parallax layers are separate files and can be easily unset by a programmer.

But I’m not convinced that removing all parallax is actually even a good idea. It’ll make it super obvious that the background is just a static image and that will potentially look really cheap.

I think a big part is that artists don’t comment on Github what they are working on and move the cards to the in-progress column.

Also if people think the template is good, it should be put up on Github so that the template can be selected when opening a new issue.

I’d really hate to split things even more

Did you know that Discourse (this forum software) has drag and drop image posting, and realtime new posts appearing in threads? So in theory if people had a tab open to these forums / we had a general in-progress art feedback thread, it’d be almost the same as a chat, but with the added benefit of actually usable history.

I agree for basically everything else except programming. I think in the programming side of things, the workflow is working much better with active progress being much more visible and discussions centralized in Github issues and on the actual changes (PRs) themselves.


Edit: @Deus I split that organism statistics panel rework graphics task into a separate issue: Create icons for microbe organism statistics · Issue #4253 · Revolutionary-Games/Thrive · GitHub

1 Like

Also, the benefit of utilizing the forums is that discussions are transparent and visible to the public.

2 Likes

I admit I didn’t knew ! That’s pretty cool. So that fix the problem x)

1 Like

Anyone remember the talk a year ago when a new artist joined who wanted us to switch to using trello for art tasks? Well, that trello was actually created, but I doubted that it would very quickly be forgotten as a separate thing. And well I made a calendar reminder which triggered now to check the status on that.

Sufficient to say, that trello did not live long at all, though maybe it would have fared better with more initial support. Anyway, I wanted to bring this up to say that any new workflow needs to be supported by team members who will be present for years in order to be a success. As a concrete measure I think we should kind of disallow these major new workflow ideas from new members of the team (who have been on the team actively for less than a couple of months), as I think the risk of those efforts fizzling out is high and as a result just waste everyone’s effort and energy.

2 Likes

A post was split to a new topic: Reworking the microbe backgrounds