Quality of Life"QoL" would have to be the largest single category of unplanned features, getting significant new additions with every Alpha release. While Cogmind has always been relatively accessible among traditional roguelikes, there are so many different ways players might want to approach the game's information and UI, and not all of these will be apparent until lots of different players are actually playing!
Honestly there was almost no playtesting during pre-Alpha (the public 7DRL before had already proven the core mechanic was fun, and that was enough for me at the time!), but once everyone (including myself
) started tackling full runs there were very obvious areas with room to streamline the experience. The first release was fun, but it would be annoying at best for the UI of today to go back to that state after getting so many extra features. Some of the bigger ones, for example...
Smart inventory management (Alpha 5c). Cogmind is full of inventory management decisions, but a number of them are no-brainers so why not have the game just take care of those automatically!
Automated part sorting (Alpha 9). Long parts lists are harder to parse when they're not organized, and repeatedly doing it manually is probably a waste of time when the game can just group like parts for the player. (Note: The visual design process for this particular feature I've written about before.)
Evasion breakdown (Alpha 10). There was originally no way to easily check just how effective your evasive capabilities were at any given point, since it's the sum of many factors, but now there's a single number and summary available right on the HUD.
Manual hacking code recall menu (Alpha 10). Before this feature it was generally necessary to write down the 4-character codes you discovered and might want to use later, but they're only valid during that run, and you may not even get a chance to use them anyway, so it was a pretty wasteful requirement.
Slot-based part swapping (Alpha 11). For a while those players using a large inventory as part of their strategy might have to spend too much time searching for the right item. Inventory sorting has always been a thing, which helped somewhat, but why not make it possible to quickly pull up a list of valid items to choose from? 
Manual hack autocompletion (Alpha 12). While a command buffer was added early on to simplify the process of repeating manual hacks, there are also somewhat lesser used hacks that would end up far back in the buffer, not to mention the continually growing number of hacking possibilities (91 as of today...).]
Additional data visualization modes--only four were planned, but we now have eight, as well as accompanying number values. This feature was initially designed for bar graphs only, since the point was to show relative values, but it turns out that having specific numbers is useful as well.At first I was reluctant to add separate difficulty modes (discussed in detail
here), but they do make a lot of sense for accessibility. So far they seem to be serving their purpose, but the true results won't be in until available on Steam, where Cogmind will be exposed to a greater ratio of players not already familiar with the genre.
Setting Cogmind’s difficulty in the options menu.I did later adjust them such that some areas are inaccessible to easier modes, partially because those areas are quite difficult by their nature (regardless of mode), and also to give players who can consistently reach them in easier modes some impetus to up the challenge. (It's still possible to win in any mode--these are just outlying special/lore areas.)
More?With any sufficiently expansive game there is nearly endless potential for QoL improvements, much less a roguelike that continues to get new content updates over a long period. Most of the existing features are taken care of pretty well, but I'm sure this category is incomplete even now--it will grow in response to playtesting and feedback on new features (and to a lesser extent as a result of new types of players joining).
MetaThere are many ways to add value to a game beyond the core experience, though naturally these kinds of things should wait until later in development so there's
actually something to build off (also so it's more apparent where the greatest value for effort can be found). While I've always had some "meta" ideas floating around, none of those considered early on have been implemented. Instead, as alpha development neared completion I worked in newer ideas that seemed best at the time.
Lore CollectionThe world of Cogmind contains quite a bit of useful (or just interesting) information among its lore. The way that lore is acquired, however, across many runs and from numerous different sources, doesn't really lend itself to forming a coherent picture of what's going on. This would be less of an issue if there were a single plot line, but the story is more a web of factions and NPCs with their own agendas and take on things. That and the lore includes quite a few tips regarding how the world works. So it eventually became obvious that we needed a central repository for all this lore, some way to conveniently reference it all.
Testing the lore collection UI (Alpha 10).Of course meta features are often also suitable for application outside a game (e.g. integration with external programs). For example players with a growing lore collection would like to be able to read it when not in game, or even in interact with it in some other format...
Sample HTML lore export, including links (Beta 2). (Also supports exporting to TXT and CSV.)As the world approached completion in late Alpha and a handful of players had seen much of what it had to offer at the time, special modes that add new rules or enforce certain restrictions on play seemed like a fun and easy way to increase replayability. There are currently eight such modes (ideas are sourced from the
relevant forum thread), with many more to come. They'll also need their own options menu UI, as opposed to the current method of control purely via the config file.
As with difficulty settings, early in development I'd never really thought about a need for extra challenges, aiming to create a world in which certain optional areas and approaches would naturally be more challenging. The ease of tacking on new challenges makes it really tempting at this point, though, considering there's such a high reward-to-effort ratio!
Gallery Item StatsPlayers had been asking for a while if they could have access to item information via their art gallery, essentially turning it into a makeshift stat compendium. During the Alpha design process I realized this is
not the ideal way to present such information, and came up with an alternative, but that alternative would also require a rather long time to implement... In any case, enough players had been asking long enough that it was time to do
something 
Item art gallery with stat access (Beta 2.1).Sure enough, after adding this feature there were a few comments about how they'd now like it to be able to do XYZ. At least I can say that if it ever comes to be, the alternative could do that and more. That's the thing about adding features as per player request--fulfilled requests often become the new launching ground for more requests xD
More?Other meta features are generally of secondary concern, so it's hard to say much about the future of this category other than that it's far from exhausted. I've always wanted to redo the score sheet, though, a project waiting on the sidelines since it's best done as late in development as possible where all the different needs it has to accommodate can be taken into account. (The only thing worse than redoing a feature is redoing it multiple times! Really though a first "redo" of the score sheet is warranted since it hasn't been redesigned at all since the 2012 7DRL, just expanded again and again
)
MiscellaneousThe broadest category of unplanned features is composed of numerous smaller bits used to fill cracks in the design, either for balance or sometimes just for fun.
One of the more frequent goals here has been to make combat more viable and varied, since the environment and overall mechanics otherwise tend to give the strategic advantage to those who prefer (and are capable of) avoiding confrontation. While that tendency will never be fully reversed, combat became a lot more interesting, and reliably winnable, as Alpha progressed.
Some examples involving differentiation of combat styles:
"Gunslinging" was only added in Beta 1, allowing a single volley composed purely of guns to be redirected across multiple consecutive targets. Before this it was easy to waste shots on a target that was already destroyed. I originally considered that part of the tactical considerations necessary for turn-based volley combat--fire many weapons and risk overkill, or fire fewer and risk the target surviving to continue the fight? But the volley timing adjustments later in Alpha made that consideration less and less of a factor anyway because firing extra weapons required almost negligible additional time, just additional resources.Cannons also got their day in the sun, with an overflow mechanic allowing surplus damage after destroying a single part to transfer to another part. This had the greatest effect on powerful late-game cannons that might otherwise blast a weak part but then have no other effect, making them randomly ineffective and unreliable. (There is still some room here to further differentiate cannons and guns, for which I have some ideas.)
Melee combat wasn't part of the
7DRL version, and while I added it to the commercial version because it's cool (and players would expect it!), it was meant to be relatively niche, had to be balanced against ranged combat, and ended up being underutilized for a while. By late Alpha the core combat styles were pretty clearly defined, so it was time to revisit melee and do something to make it more desirable. Many somethings were done
Actuators and other utilities were added to help builds with a melee focus (
Alpha 11), sneak attacks provided another tactical edge (
Alpha 12), multiple targets could be hit with a single swing (
Alpha 14), and most importantly, melee-specific multiwielding became a thing.
Attaching multiple melee weapons at once (Alpha 14). Players designate a primary weapon, and other weapons might contribute to attacks.Additional damage types were also created (secret, so I won't mention them here), and some of the existing types were given new effects:
Thermal weapons were given a dedicated heat transfer stat, rather than being a simple side effect of damage (
Alpha 12). By decoupling the two, heat transfer could become a goal of its own, either to cause meltdowns or for its other negative modifiers. Before that the thermal weapon design space was quite limited because heat transfer always had to play a secondary role to damage, and a weapon with better damage would also be better at transferring heat, so it could only make good weapons even better rather than serving as a point of differentiation.
Overheating a robot with lasers.Electromagnetic weapons were given a "spectrum" (
Alpha 4). Their attacks sometimes causing chain reactions in robot power sources made the EM tactics game somewhat more interesting, especially earlier in a run where it happens more often (due to the types of weapons).
Triggering a chain reaction in Grunt's power source via EM Shotguns' spectrum.Impact weapons were always something to be avoided, in both senses
. Essentially they were quite dangerous in enemy hands since they could smash your processors and weaker parts, but useless to the player who would rather outright destroy targets than cripple them. Once given the ability to inflict significant system corruption by smashing parts (
Alpha 14) they became pretty decent.
Corrupting bots with a Shock Maul.Once the main game was completed earlier this year, it became easier to explore extra features that
might work out, rather than focusing mostly on what absolutely had to be done at each phase of progress.
One result of this experimentation was "visible sound effects." For a while I'd been thinking of closed-captioning sound effects for accessibility reasons, but simply showing color-coded sound effects would already be a big step in that direction, and far easier to implement. Plus it's a tactically interesting new way for
everyone to view the map, rather than just those who might benefit from captioning. It's worked out pretty well!
Visible sound effects for combat occurring nearby but out of view (Beta 1).I also came up with an optional way for players who really want every single combat detail to get that rundown on the map rather than squished into a small secondary log not really made for that much info.
On-map detailed combat log (shown here closer to the action than it really is) (Beta 1.2).But the miscellaneous features I'm most interested in exploring, beyond the design doc, beyond even strict balance considerations, are unique parts with either excellent stats or sometimes completely new mechanics. Content like this won't have to take into account balance in the same way as previous developments, because these parts are not at all guaranteed (unlike pretty much everything else in the world--an intentional design decision enabling players to somewhat plan their build/strategy ahead of time). I've relied on this principle in a few places in the world already, but prior to Beta it was more important to keep the core game balanced for proper testing. So there's still a lot more that can be done here, and it should make Cogmind even more fun--after all, there's nothing wrong with getting to feel OP for a while if it's not something you can do every time!
Imagine...Now, obviously a massive collection of features
were planned from the beginning, but not nearly everything we have now! Technically none of the features discussed above would've happened had I strictly gone by the docs and worked straight through Alpha to 1.0. Imagine how different Cogmind would be without garrisons, trojans, imprinting, automated UI features... and all that other stuff!
The point is, design docs are really important, especially in clarifying the vision for the experience and setting down basic principles to build a world based on that vision, but in the best of cases it can also be treated as a flexible framework on which the development process can expand. Pre-Alpha and Alpha finished the design doc over four years and even let a lot of other features leak in at the same time, while with Beta it's clear sailing with more ideas to explore on the horizon
Author
Topic: Cogmind (sci-fi robot-themed roguelike) - BETA RELEASED (Read 286823 times)