So I've engaged in mild speculation on new console specs in the past. We're getting to the point where there should be some new consoles coming. It lines up pretty well with (large chip) mass production 7nm in the coming year or two so everyone will be ready to offer powerful SoCs with high power efficiency for console enclosures. A 2020 release is 7 years on from the current high fidelity consoles - very normal for the console upgrade cadence. Even the half-generation updates would be getting on by the end of 2020: four years after the PS4 Pro was released.
Phil Spencer was at CES talking up their continuing partnership with AMD, which certainly sounds like an ongoing use of their IP (makes a lot of sense - AMD have all the IP under one roof and a track record; Jaguar cores are going to get a substantial update with a Zen derivative; and the MCP design AMD are working with could be a good match for a console that will live through at least a couple of die shrinks while uncertain about how much of a monolithic SoC would benefit from that). Vega (and even lower end Polaris) are not leading the power efficiency war but at this point everyone is experienced in making GCN work well and there are still a few bad memories around nVidia. Everyone hopes that the next Vega is going to bring that power efficiency to where nVidia are at on 12nm (if they can't when a generation ahead at 7nm and not wasting any transistors on speculative features, something will be going very wrong). There is also the deep in development Navi so even without turning it around next month, the future may be bright. The only wildcard would be Intel trying to sell their 2020 GPU design to Sony or MS (or even offering an AMD GPU: did not see that coming a year ago) - very long odds in my book but worth mentioning as Intel make a big move to compete on high end GPUs to avoid seeing server farms filled with expensive nVidia chips (getting a foot in the door with Tensor etc specialised silicon to expand their use cases). [Those with a long memory will remember that nVidia were paid by Intel to relinquish any claim to be able to provide x86 chips and so can't offer a complete package that would enable a rolling window of compatibility.]
I'm going to say that odds are good for something like a CCX cluster (4 Zen cores) or two with a pretty large Navi GPU block. What I'm curious about: what if the next Xbox is two new Xboxes? Rather than releasing one console then doing a major redesign when the power requirements are lower, what if there is a constant pair of consoles, like we have now. The next Xbox as a high end system with at least 8 Zen cores and a big Navi block. $500 at the end of 2020 with enough fast RAM to avoid the issues that plagued the original XB1 and hopefully at least a PCI-E SSD cache (64GB? 128GB? 256GB as a premium SKU?) to avoid the loading times getting even worse (even if most downloads and game installs would live on a large platter drive for cost reasons, even with NAND getting cheaper).
But the XB1X isn't that old and isn't a terrible target for an entry-level spec for games going forward from 2020. There would probably be a way to make a reasonably good approximation of that system using a smaller Zen cluster and much smaller Navi block (cheaper/less RAM and no SSD cache). Something that would run existing games just like an XB1X, with minimal shims to catch edge cases (in the low level API calls). Something that would effectively be sold as both the cheap entry to the next generation and the "Slim" edition of the XB1X. Something to ensure developers continue to make sure their new games work for existing XB1X customers (with whatever compromises were required for performance) while dropping support for new games on the anaemic XB1/XB1S system. The fact that the XB1S isn't identical to the original XB1 shows that the ecosystem is already built around supporting two low end specs that are only approximately the same.
When the next high end Xbox is ready for a slim-down, it might be time to look at another step forward (an "XB2X") and which devices get guaranteed access to 100% of new releases going forward (and which platforms are something a developer could optionally support in a "cross-buy" style system). Have two or three models always "current" that cycle out every three to five years until they can no longer continue to offer the right shims to ensure compatibility with their API choices. It would certainly be something not totally out of step with the current MS talk of slowly expanding their cloud gaming offerings (on TVs, tablets, and laptops) and making games work anywhere (Windows PCs and Xbox devices). It's the realisation of much of the "what if more like annual tablet/phone updates" talk from about five years ago.
Sunday, 13 January 2019
Saturday, 29 December 2018
Games of the Year 2018
So in recent years I've been doing this based on games that released (1.0 versions) in a given year. But as we talk of living games and Early Access, there's a lot more going on that maybe makes this harder to do right. I'm going to generally try to keep to that broad concept - a list of new, complete games of that year that stood out to me - while maybe considering a lot of games that are for sale but didn't necessarily announce a 1.0 version number. The big addition is games that feel close to release and could be seen as the equivalent of many 1.0 titles that go on to get bigger via incremental patches.
The previous requirement to see a 1.0 number has led to some slightly weirdness. When considering RimWorld's 1.0 release in the later part of this year, I'm considering a game I mainly played (and even wrote about) almost two years ago. Maybe that would have made more sense to be considered last year or even in 2016. I haven't talked about Factorio much here but I've been enjoying that once more and it's getting close to that 0.17 release which is basically feature complete and lays the way for a 0.18 update in early 2019 that will be the official 1.0. That seems like a game that can be considered for this year's list. Slay the Spire is missing their expected 1.0 release window but it's been close to done since I first talked about it.
On the other side of the fence, Mashinky continues to see new development along a roadmap but is very much not done in the way those earlier games are. I think it's not up for this year's list despite having really enjoyed the game as it exists so far. There is a clear point in each playthrough where you reach the end of the line and find the "coming soon" notice - it's not just a game that will get deeper with patches, it feels like the core progression is still very much being built (as a normal part of Early Access). The final balance of the progression may well change significantly during polishing because we don't even know how a full run plays out yet. Rounding out the examples, Life is Strange 2 is in no way finished yet so will be considered for next year. While Half-Life 2 had 'Episodes' as standalone expansions, LiS is this more modern understanding of seasons that work as one cohesive package (released at a reasonable pace and purchased as a season bundle).
Strategic 'Mech of the Year:
The individual skirmishes in BattleTech are just some of the most chewy tactical decisions around, with plenty of damage output and equally lots of armour that'll evaporate when you block that incoming fire. As I said when it came out, you can just play and play that layer but it's the arc of the strategic layer that really binds it all together and turns the grinder of each mission into a desperate search for the next contract between the story missions. Much of that has been further tweaked over time and several mods (and an official expansion) have really refined everything so you can get the challenge you demand to match the tone of the game. My view hasn't changed much since I wrote up that review (and the news the developer is now a Paradox subsidiary, so potentially a lot of piecemeal DLC is planned after this current season of expansions) and I expect to head back for even more in 2019 as they slowly expand the combat scenarios and develop the strategic layer with DLCs.
Platformer of the Year:
While still constrained by the limited GPU of the PS4 (you're not getting nearly the anti-aliasing you need for great VR visuals), Astro Bot is a really clean looking game that oozes light charm and playfulness. The sound helps set that stage but it's little things like bumping the playable character with your controller as you both float in space between levels that is not just playful but helps fix you in the world. Oh, a glowing orb during the short loading screens? I bet I can bounce that with my controller... *orb bounces with a satisfying chime sound*. The game itself is pretty traditional for a platformer, with enough variety to keep it engaging for the eight odd hours it takes to play through it. Hit checkpoints, collect coins, rescue the other robots (who fly up and into the controller you're holding - VR is still somewhat magical when the tracking works perfectly so you see the controller exactly as you're holding it but with the freedom to render it however the game wants), figure out the puzzles and secrets. That rarely used touchpad on the DS4 gets a good workout (will be a shame if the PS5 comes out and cost-saves that out of existence). You already probably know what this is, it's a very good one of those and one of the best games to showcase VR (sorry Lucky's Tale, the bar has risen significantly).
Tomb Raiding of the Year:
That's a lot of words about other games but it's important to understand this new God of War as being very much in a new lineage as well as being a sequel to the previous games in the series. The director having previously worked on the Tomb Raider reboot before taking on this project with a list of ideas that didn't make it there (like the camera that never cuts) is how to understand what is on offer here. The major genre-switch from those games being the primarily "stylish action" melee combat in place of ranged weapons or playing for stealth.
The story of family bonds, while leaning on a small mainly-male cast of often-stereotypes, is elevated by the performances and quality of rendering. Christopher Judge perfectly leads the VO cast in a game always ready to fill traversal with anecdotes (which have plenty of spaces to trail off and get continued later as you break for action) and then provide useful information during combat (where the close camera makes calling out attacks as vital as the danger indicator arrows). The core narrative feels like it's something much more reasonable (20 hours?) than what you're finding in open worlds this year (eg RDR2, AC:O) without there not being stuff worth doing if you've got 20-30 more hours you want to put into getting deep into the craftable equipment and ability upgrade trees or just hearing some more anecdotes and side stories while catching the amazing sights - Sony continuing to show how far devs can push their system. Huge areas of the game world are left for optional quest lines on top of a roguelike-like zone (reconfiguring dungeon with a poison mechanic that offers a poison resist gear progression to finally crack it with a run to the treasures at the core) and combat (challenge) arena zone.
I usually open my GotY post (or give the top picture slot) to the game that is my top game that year (everything else is not even trying to be ordered by preference) and this year I've flipped between this and BattleTech a few times. Ultimately, the performances here elevate a story that is too reliant on clichés (and women who are dead or wish to die - the small cast doesn't help allow a diversity of traits for any marginalised voice included) so despite being a gorgeous game I really enjoyed almost every moment with, it misses out on my top spot. But I really look forward to what comes from this team next and finding out if they course correct on some of the narrative stuff (or do a David Cage and double down on the issues with each iteration).
Lumines of the Year:
I had a GameBoy back around 1990 so I gave the original game some energy (but never experienced the Tetris effect with it, something I did experience with Arkanoid: Revenge of Doh and Columns around that time) and this takes that gameplay (with the few updates to the official formula now considered standard) to another level. Something about the expertly chosen music, sequenced to grow and adapt to the length of your session and button presses, and the visuals keep everything fresh. So does the decision to make the levels contain slower and faster drop speeds based on progression to a line-clear goal, giving the music slower and faster sections, rather than the constantly increasing speed of the classic game. Beyond the many Journey (campaign) levels that tie different visuals and music to the game, there is also plenty of challenges and leaderboards that ask you to focus on slightly different modifiers that can give you a more chilled experience or force you to deal with a single aspect of the game. Particularly, the Mystery challenge offers some extremely non-standard randomised elements that mix up the Tetris formula.
I'm sure it's great in 2D on a TV but to me the game is another that shows off what VR can do with low latency, making anything musical just perfectly timed and where a fully enclosed environment makes even a classic puzzle game swim around you and put you inside the virtual space.
Management Sim of the Year:
But Factorio has none of that pawn management; it's a builder with a single (controllable) protagonist who you inhabit while doing much of the construction stuff you'd normally do in a god game. You develop robot helpers in the later stages of each run to provide remote construction and management (usable via the map view that zooms into something that looks like the normal view in areas with radar coverage), at which point you can start playing it more like a builder. It's almost like the modded (huge) tech trees of Minecraft, except this retains the top-down view of classic management games. Ultimately, as I iteratively design these huge chains of machines that take in raw material and convert them to final products which can be used to fund more research or placed to expand the factory, it feels a lot like a more open ended version of the puzzles in many Zachtronics games like Infinifactory.
Always, I keep an eye on the power use of everything, watching the pollution which will waft out and enrage the warrens of hostile creatures that live on the planet I've crashed on. My playstyle leans on efficiency and relatively green operations (while still fundamentally mining out the local resources), unlike the really expansive players who ramp up and consume the world with their endless smog and high power factories. You can dump hour after hour into tweaking designs and working out exactly how best to feed various intermediate stages in the factory process here, which is where I feel it sits next to a Zachtronics game (almost offering a way to gamify circuit layout or similar tasks without overlapping with existing programming-style games around that theme - it's all conveyors, loading arms, basic detectors, circuit conditions, and very simple logic here). It's almost at version 1.0 with a load of great mods (and a simple interface to enable them) and a final rebalancing and polishing pass currently underway so now is a great time to jump into a game that's been in Early Access for years.
Mouse of the Year:
Runaway Card Combo of the Year:
I didn't spend the entire year going back to slowly refine my card preference and climb up the Ascension ladder (unlocked difficulty modifiers) but I've had a good time with this and been back to enjoy some daily challenges. It's exactly the sort of solid design that means you can dive in and spend a few dozen hours and feel satisfied, not the eternal treadmill of new seasons and boosters (and corresponding demand for more money). That said, now it is almost released, we are hopefully going to see some expansions that add a new character and cards at some point - a good excuse to dive back in once more.
Pawn (Sim) of the Year:
Crashing onto an inhabited planet, a bunch of pawns who all have their own wants and needs must be wrangled into actually doing the important task of living and building up a colony that can survive raider attacks, wild creatures, and natural disasters all while the seasons turn and friendly factions offer trade (and the opportunity to kidnap new members if you can woo them sufficiently). What if The Sims except occasionally a bunch of marauders turn up and you go into a direct-control battle mode? Oh, and less time spent dictating hobbies and more time making sure there are enough crops harvested and refrigerated before winter or actually constructing the items the commune needs from raw resources all based on skill levels of each pawn. Some of the rougher edges and design decisions around pawn psychology have been tweaked via the vibrant mod community and as the game has developed further towards 1.0 (hitting a few moths ago) then a lot of the older usability mods have become core features in the base game. Today, it's a huge story generator without the learning curve cliff that Dwarf Fortress is best known for.
Oscar for Just Falling Short Every Time:
In a year where Need for Speed is totally absent without it being a loss (after repeated failed attempts to release something meaty recently) and The Crew 2 arrived as a step back rather than a confident second title (diversifying vehicles while failing to capture the fun of the dynamic challenges that helped cruising round Weird USA in the first game or even retaining the online racing or bad F&F knock-off story elements). So it's time to evaluate the Horizon spin-off series within the crumbling ruins of open city/world driving games.
Horizon games are still good driving, and that means they're by far the best out there now. A new landscape every two years; enough stuff to dump hour after hour into, becoming familiar with every bend; and some nice tech updates each time (running ahead of the 60fps main series while coming to PC has - sometimes imperfectly - enabled 60 with those shiny new dynamics). This year the new tech involved repainting the landscape for each season, on top of the previous dynamic weather and time of day updates. Unfortunately the weight of all these iterations feels like it has almost crushed the core game, now lacking any sense of progression (outside the carefully paced tutorial hours) as everything rests on the roulette wheel of unlocks and endless treadmills of unlocking events that adapt to whatever vehicle you happen to have unlocked and bring along.
Everything adapts. So nothing ever gets harder or feels like a better vehicle would make easier. Whatever you drive up in will work, unless you pick a road car for an off-road challenge. The slight wrinkle here is winter and how that makes a few more cars unsuitable for the slippery conditions. It also means all of the small challenges dotted around the map (which all share a single leaderboard that doesn't account for the seasonal conditions) become pointless as you wait for summer to actually beat your friends. It feels like a live game (the season cycle weekly based on server time and come with "new" events that are actually just recycled existing events with a new marker over them) and the hourly challenges give a reason to meet up with other actual players and all contribute to some shared goals but it quickly runs out of new challenges to share (many of them are already just the existing small changes with a new communal total every individual attempt adds onto so it's repetition not high-score chasing as a group activity).
The lack of a feeling of real progression extends to every event, which are in the classic categories and almost too numerous to count. They all bring up a list of "balanced" opponents (the algorithm used feels like it needs some tweaking as I've noticed the same buggy selections that infested Forza Motorsport 7 last year) based on whatever you care to drive from your roulette wheel of winnings (gone is selling cars back to the garage so you can't even make some money on a car you don't want unless you're prepared to spend hours on the Auction House selling it to another real player and undercutting anyone else trying to offload yet another car they don't care about because it's all random). You don't even have to try and win an event or push down the assists as the small extra credit bonus for less assists doesn't really matter (when looking at $15m player housing with bonus effect) and not even the map tracks when you actually won anything. It's a mess of inconsistent iconography where only the small map challenges even denote stars for doing more than just finishing something in any time.
And this is the best we have right now, by quite some margin. Hopefully the console generation transition (and being a now wholly owned studio inside the Microsoft beast) will give Playground the impetus and security to try and do more than iterate because if Forza Horizon 5 doesn't radically change the formula, it might be time to put a fork in the open world driving genre. We can all go back to playing remasters of the ten year old Burnout Paradise.
2017 overflowed with so many games that I missed a lot of 2018, these are all pending:
Exapunks, Mutant Year Zero: Road to Eden, Marvel's Spider-Man, Hitman 2 (not Silent Assassin), Assassin's Creed: Odyssey (Xena Simulator), Vampyr, Paratopic, The Missing: J.J. Macfield and the Island of Memories, Into the Breach, Red Dead Redemption 2, Pillars of Eternity II: Deadfire, Artifact, Valkyria Chronicles 4, Yakuza 6: The Song of Life, Phantom Doctrine, Frozen Synapse 2, Ashen, & Ni no Kuni II.
The previous requirement to see a 1.0 number has led to some slightly weirdness. When considering RimWorld's 1.0 release in the later part of this year, I'm considering a game I mainly played (and even wrote about) almost two years ago. Maybe that would have made more sense to be considered last year or even in 2016. I haven't talked about Factorio much here but I've been enjoying that once more and it's getting close to that 0.17 release which is basically feature complete and lays the way for a 0.18 update in early 2019 that will be the official 1.0. That seems like a game that can be considered for this year's list. Slay the Spire is missing their expected 1.0 release window but it's been close to done since I first talked about it.
On the other side of the fence, Mashinky continues to see new development along a roadmap but is very much not done in the way those earlier games are. I think it's not up for this year's list despite having really enjoyed the game as it exists so far. There is a clear point in each playthrough where you reach the end of the line and find the "coming soon" notice - it's not just a game that will get deeper with patches, it feels like the core progression is still very much being built (as a normal part of Early Access). The final balance of the progression may well change significantly during polishing because we don't even know how a full run plays out yet. Rounding out the examples, Life is Strange 2 is in no way finished yet so will be considered for next year. While Half-Life 2 had 'Episodes' as standalone expansions, LiS is this more modern understanding of seasons that work as one cohesive package (released at a reasonable pace and purchased as a season bundle).
BattleTech
My first taste of BattleMechs made digital was the oft-forgotten Westwood RPG (back before they were known only as an RTS studio) BattleTech: The Crescent Hawk's Inception from 1988. Two years later the turn-based combat gave way to real-time tactics battles in the sequel, The Crescent Hawks' Revenge, and paved the way for Dune II and ultimately Command & Conquer. That universe became lore I absorbed over decades of various games as the original wargame designers built their own FASA Interactive team and then Microsoft acquired them (leading to a slow decline). But Jordan Weisman is back crowdfunding all his old IPs and showing there is value left in them (previously: the recent Shadowrun RPGs). So BattleMechs are back on the menu and it's time to go back to basics with a turn-based tactics RPG.The individual skirmishes in BattleTech are just some of the most chewy tactical decisions around, with plenty of damage output and equally lots of armour that'll evaporate when you block that incoming fire. As I said when it came out, you can just play and play that layer but it's the arc of the strategic layer that really binds it all together and turns the grinder of each mission into a desperate search for the next contract between the story missions. Much of that has been further tweaked over time and several mods (and an official expansion) have really refined everything so you can get the challenge you demand to match the tone of the game. My view hasn't changed much since I wrote up that review (and the news the developer is now a Paradox subsidiary, so potentially a lot of piecemeal DLC is planned after this current season of expansions) and I expect to head back for even more in 2019 as they slowly expand the combat scenarios and develop the strategic layer with DLCs.
Astro Bot: Rescue Mission
I'm not a huge platformer fan. They're fine but I started out on hardware that was never great at them - lack of hardware support for scrolling meant each screen was much more about what we now call (action-)adventure rather than the purity of smooth platforming. I've enjoyed myself some Sonic recently but since the industry move to 3D, I've not found much that really pushes me towards the purity of jumping (when I can get more from games that incorporate that into a larger genre as good traversal). But there's something about VR that really helps ground you and physically peering round for secrets and hidden locations is a delight that I'll probably not even start to tire of for a few more years. Collecting all the things while bashing the baddies in a kid-friendly environment is just more fun as we wait for VR to become capable of supporting bigger games (Homeworld: VR when?) where that's just one element (Assassin's Creed isn't scheduled to become a VR title soon with VR's limited play times to avoid simulator sickness).While still constrained by the limited GPU of the PS4 (you're not getting nearly the anti-aliasing you need for great VR visuals), Astro Bot is a really clean looking game that oozes light charm and playfulness. The sound helps set that stage but it's little things like bumping the playable character with your controller as you both float in space between levels that is not just playful but helps fix you in the world. Oh, a glowing orb during the short loading screens? I bet I can bounce that with my controller... *orb bounces with a satisfying chime sound*. The game itself is pretty traditional for a platformer, with enough variety to keep it engaging for the eight odd hours it takes to play through it. Hit checkpoints, collect coins, rescue the other robots (who fly up and into the controller you're holding - VR is still somewhat magical when the tracking works perfectly so you see the controller exactly as you're holding it but with the freedom to render it however the game wants), figure out the puzzles and secrets. That rarely used touchpad on the DS4 gets a good workout (will be a shame if the PS5 comes out and cost-saves that out of existence). You already probably know what this is, it's a very good one of those and one of the best games to showcase VR (sorry Lucky's Tale, the bar has risen significantly).
God of War
I love me some Tomb Raider. But as we come to the conclusion of the most recent trilogy with Shadow of the Tomb Raider, I'm left feeling like the initial step forward in the reboot didn't find enough further progress and the narrative left me pining for 2013's ensemble cast (and what they might have done with developing those bonds). But this is a God of War sequel, why start talking Lara's recent escapades? Surprisingly enough, this is the sequel to that 2013 game I wanted: an action-adventure around a semi-open world; quite linear chapters with wide paths and a lot more open areas and reasons to return to collect items or complete side-quests in any order. The world feels more concrete and open (and is certainly less linear in how you approach the different quests) without sacrificing the crafted story and potential for an evolving environment (with some gating and areas that transform as the main plot changes, without making anything inaccessible). As we've seen Uncharted: The Lost Legacy start to poke at wider paths, Lara is starting to look outdated five years after popularising the transition (that avoids falling fully into the open worlds that feel like a different genre).That's a lot of words about other games but it's important to understand this new God of War as being very much in a new lineage as well as being a sequel to the previous games in the series. The director having previously worked on the Tomb Raider reboot before taking on this project with a list of ideas that didn't make it there (like the camera that never cuts) is how to understand what is on offer here. The major genre-switch from those games being the primarily "stylish action" melee combat in place of ranged weapons or playing for stealth.
The story of family bonds, while leaning on a small mainly-male cast of often-stereotypes, is elevated by the performances and quality of rendering. Christopher Judge perfectly leads the VO cast in a game always ready to fill traversal with anecdotes (which have plenty of spaces to trail off and get continued later as you break for action) and then provide useful information during combat (where the close camera makes calling out attacks as vital as the danger indicator arrows). The core narrative feels like it's something much more reasonable (20 hours?) than what you're finding in open worlds this year (eg RDR2, AC:O) without there not being stuff worth doing if you've got 20-30 more hours you want to put into getting deep into the craftable equipment and ability upgrade trees or just hearing some more anecdotes and side stories while catching the amazing sights - Sony continuing to show how far devs can push their system. Huge areas of the game world are left for optional quest lines on top of a roguelike-like zone (reconfiguring dungeon with a poison mechanic that offers a poison resist gear progression to finally crack it with a run to the treasures at the core) and combat (challenge) arena zone.
I usually open my GotY post (or give the top picture slot) to the game that is my top game that year (everything else is not even trying to be ordered by preference) and this year I've flipped between this and BattleTech a few times. Ultimately, the performances here elevate a story that is too reliant on clichés (and women who are dead or wish to die - the small cast doesn't help allow a diversity of traits for any marginalised voice included) so despite being a gorgeous game I really enjoyed almost every moment with, it misses out on my top spot. But I really look forward to what comes from this team next and finding out if they course correct on some of the narrative stuff (or do a David Cage and double down on the issues with each iteration).
Tetris Effect
Lumines was my most played PSP game. Designed to be something like Tetris (without access to that license) while mixing in the synaesthetic reactive music and visuals the team had worked on a few years earlier with Rez. Now that team is back with the actual license (after offering an update to Rez in VR two years ago) and it's Tetris like you've never seen before, as the surroundings explode around you in VR.I had a GameBoy back around 1990 so I gave the original game some energy (but never experienced the Tetris effect with it, something I did experience with Arkanoid: Revenge of Doh and Columns around that time) and this takes that gameplay (with the few updates to the official formula now considered standard) to another level. Something about the expertly chosen music, sequenced to grow and adapt to the length of your session and button presses, and the visuals keep everything fresh. So does the decision to make the levels contain slower and faster drop speeds based on progression to a line-clear goal, giving the music slower and faster sections, rather than the constantly increasing speed of the classic game. Beyond the many Journey (campaign) levels that tie different visuals and music to the game, there is also plenty of challenges and leaderboards that ask you to focus on slightly different modifiers that can give you a more chilled experience or force you to deal with a single aspect of the game. Particularly, the Mystery challenge offers some extremely non-standard randomised elements that mix up the Tetris formula.
I'm sure it's great in 2D on a TV but to me the game is another that shows off what VR can do with low latency, making anything musical just perfectly timed and where a fully enclosed environment makes even a classic puzzle game swim around you and put you inside the virtual space.
Factorio
How have I not written about this on here before? It's been a good year for builder games, often incorporating detailed pawn simulation in new ways. The bitter cold and series of bad or worse decisions in Frostpunk built an atmosphere and progression that mixed in a lot of how survival games can weigh you down, all with very impressive visuals (and patched in some additional scenarios beyond the initial campaign). Meanwhile, Surviving Mars randomised the tech tree and offered a range of event chains rather than a limited set of fixed scenarios to mix up each playthrough. And later in this list I'll get to RimWorld.But Factorio has none of that pawn management; it's a builder with a single (controllable) protagonist who you inhabit while doing much of the construction stuff you'd normally do in a god game. You develop robot helpers in the later stages of each run to provide remote construction and management (usable via the map view that zooms into something that looks like the normal view in areas with radar coverage), at which point you can start playing it more like a builder. It's almost like the modded (huge) tech trees of Minecraft, except this retains the top-down view of classic management games. Ultimately, as I iteratively design these huge chains of machines that take in raw material and convert them to final products which can be used to fund more research or placed to expand the factory, it feels a lot like a more open ended version of the puzzles in many Zachtronics games like Infinifactory.
Always, I keep an eye on the power use of everything, watching the pollution which will waft out and enrage the warrens of hostile creatures that live on the planet I've crashed on. My playstyle leans on efficiency and relatively green operations (while still fundamentally mining out the local resources), unlike the really expansive players who ramp up and consume the world with their endless smog and high power factories. You can dump hour after hour into tweaking designs and working out exactly how best to feed various intermediate stages in the factory process here, which is where I feel it sits next to a Zachtronics game (almost offering a way to gamify circuit layout or similar tasks without overlapping with existing programming-style games around that theme - it's all conveyors, loading arms, basic detectors, circuit conditions, and very simple logic here). It's almost at version 1.0 with a load of great mods (and a simple interface to enable them) and a final rebalancing and polishing pass currently underway so now is a great time to jump into a game that's been in Early Access for years.
Moss
Sometimes you just want to listen to a story of a mouse as you play though some action-adventure dioramas that VR allows to look perfectly scaled, allowing you to understand the tiny world you're looking into. Weirdly, this wasn't the only mouse game that arrived earlier in the year; Ghost of a Tale was far more of a full (if unexceptional) RPG but lacked the way VR allows you to really feel the scale of everything. This second VR game on the list (also somewhat of a platformer) is also exuding charm but in a completely different way to Astro Bot. The cuteness is strong here and the mix of direct controller controls and reaching into the scene continues a theme this year or games getting more confident with the additional inputs VR offers while staying with a sitting-down, controller-led experience.
Slay the Spire
This year Valve brought in MtG creator Richard Garfield to design their new digital card game (that is unfortunately pay to play on top of an initial price tag) and Magic: the Gathering itself got a new digital version with MtG: Arena that more closely mirrors the online play rewarding booster packs used to great effect in Hearthstone (moving away from the buy and trade model used by Artifact and MtG: Online). But the digital card game I really enjoyed this year isn't about buying boosters, it's the purity of roguelike-like deckbuilding. When Slay the Spire left its small beta and arrived on Steam Early Access, I mentioned how complete it already felt. Back then the plan was to call it 1.0 later in the Summer but they decided that 52 weekly patches was what it needed before calling it gold so that means the final release is now imminent.I didn't spend the entire year going back to slowly refine my card preference and climb up the Ascension ladder (unlocked difficulty modifiers) but I've had a good time with this and been back to enjoy some daily challenges. It's exactly the sort of solid design that means you can dive in and spend a few dozen hours and feel satisfied, not the eternal treadmill of new seasons and boosters (and corresponding demand for more money). That said, now it is almost released, we are hopefully going to see some expansions that add a new character and cards at some point - a good excuse to dive back in once more.
RimWorld
The current generation of Dwarf Fortress inspired games have managed to move away from being far shallower clones and into the territory of actually offering something different. RimWorld is probably the most popular and it finally fully released this year, quite a different beast from the very early version that just looked like a Prison Architect asset rip.Crashing onto an inhabited planet, a bunch of pawns who all have their own wants and needs must be wrangled into actually doing the important task of living and building up a colony that can survive raider attacks, wild creatures, and natural disasters all while the seasons turn and friendly factions offer trade (and the opportunity to kidnap new members if you can woo them sufficiently). What if The Sims except occasionally a bunch of marauders turn up and you go into a direct-control battle mode? Oh, and less time spent dictating hobbies and more time making sure there are enough crops harvested and refrigerated before winter or actually constructing the items the commune needs from raw resources all based on skill levels of each pawn. Some of the rougher edges and design decisions around pawn psychology have been tweaked via the vibrant mod community and as the game has developed further towards 1.0 (hitting a few moths ago) then a lot of the older usability mods have become core features in the base game. Today, it's a huge story generator without the learning curve cliff that Dwarf Fortress is best known for.
Forza Horizon 4
My view of the original Forza Horizon has slightly improved with time (especially as the expansions were fun and the micro-transactions that I cited against it have now infected all of AAA). But my fear of it becoming a semi-annual franchise that fails to evolve sufficiently has really cursed the entire Forza franchise (the main series still finds Forza Motorsport 4 as a pinnacle of design, features, and progression that later games have repeatedly missed). It's a series I keep on not quite giving a space on these lists.In a year where Need for Speed is totally absent without it being a loss (after repeated failed attempts to release something meaty recently) and The Crew 2 arrived as a step back rather than a confident second title (diversifying vehicles while failing to capture the fun of the dynamic challenges that helped cruising round Weird USA in the first game or even retaining the online racing or bad F&F knock-off story elements). So it's time to evaluate the Horizon spin-off series within the crumbling ruins of open city/world driving games.
Horizon games are still good driving, and that means they're by far the best out there now. A new landscape every two years; enough stuff to dump hour after hour into, becoming familiar with every bend; and some nice tech updates each time (running ahead of the 60fps main series while coming to PC has - sometimes imperfectly - enabled 60 with those shiny new dynamics). This year the new tech involved repainting the landscape for each season, on top of the previous dynamic weather and time of day updates. Unfortunately the weight of all these iterations feels like it has almost crushed the core game, now lacking any sense of progression (outside the carefully paced tutorial hours) as everything rests on the roulette wheel of unlocks and endless treadmills of unlocking events that adapt to whatever vehicle you happen to have unlocked and bring along.
Everything adapts. So nothing ever gets harder or feels like a better vehicle would make easier. Whatever you drive up in will work, unless you pick a road car for an off-road challenge. The slight wrinkle here is winter and how that makes a few more cars unsuitable for the slippery conditions. It also means all of the small challenges dotted around the map (which all share a single leaderboard that doesn't account for the seasonal conditions) become pointless as you wait for summer to actually beat your friends. It feels like a live game (the season cycle weekly based on server time and come with "new" events that are actually just recycled existing events with a new marker over them) and the hourly challenges give a reason to meet up with other actual players and all contribute to some shared goals but it quickly runs out of new challenges to share (many of them are already just the existing small changes with a new communal total every individual attempt adds onto so it's repetition not high-score chasing as a group activity).
The lack of a feeling of real progression extends to every event, which are in the classic categories and almost too numerous to count. They all bring up a list of "balanced" opponents (the algorithm used feels like it needs some tweaking as I've noticed the same buggy selections that infested Forza Motorsport 7 last year) based on whatever you care to drive from your roulette wheel of winnings (gone is selling cars back to the garage so you can't even make some money on a car you don't want unless you're prepared to spend hours on the Auction House selling it to another real player and undercutting anyone else trying to offload yet another car they don't care about because it's all random). You don't even have to try and win an event or push down the assists as the small extra credit bonus for less assists doesn't really matter (when looking at $15m player housing with bonus effect) and not even the map tracks when you actually won anything. It's a mess of inconsistent iconography where only the small map challenges even denote stars for doing more than just finishing something in any time.
And this is the best we have right now, by quite some margin. Hopefully the console generation transition (and being a now wholly owned studio inside the Microsoft beast) will give Playground the impetus and security to try and do more than iterate because if Forza Horizon 5 doesn't radically change the formula, it might be time to put a fork in the open world driving genre. We can all go back to playing remasters of the ten year old Burnout Paradise.
2017 overflowed with so many games that I missed a lot of 2018, these are all pending:
Exapunks, Mutant Year Zero: Road to Eden, Marvel's Spider-Man, Hitman 2 (not Silent Assassin), Assassin's Creed: Odyssey (Xena Simulator), Vampyr, Paratopic, The Missing: J.J. Macfield and the Island of Memories, Into the Breach, Red Dead Redemption 2, Pillars of Eternity II: Deadfire, Artifact, Valkyria Chronicles 4, Yakuza 6: The Song of Life, Phantom Doctrine, Frozen Synapse 2, Ashen, & Ni no Kuni II.
Saturday, 24 November 2018
µNote: Respecting the Player's Time
MicroNotes are something new I'm going to try (possibly through 2019). Rather than aiming for a long post once a month, which I am prone to kill (if it doesn't come together or looks slight after poking at a draft), I'm going to post some shorter writing here. I've been writing some thoughts elsewhere for a while and so might go back and put some of those posts here (and even go through my old drafts that I never fully shelved). This should also give room for smaller coding & rendering thoughts (that are bigger than fit on Mastodon or Twitter) to actually get posted here.
I have always found "respecting the player's time" to be a useful lens through which to consider game design. What needs to be here; what do I think makes a positive contribution to the core arc; just how much of a structure can be made optional and how do we signpost it so players can best dive into only the content they want? Extracting the most from the interactive nature of games means building structures that react to each player and customises how they experience a game, attempting to give the best progression to the most people we can reach (and even respecting that some players will not enjoy what we are making and should not be strung along). The thing is, this phrasing has become extremely common in games criticism, to the point of dilution.
“I don’t like this game as I feel like it is not respecting my time (and here are 6 paragraphs on exactly how)” is commentary on where I feel a game does not align with how I value the different activities it offers and what it considers core vs optional - I can’t fast travel in a game where I don’t agree with a design decision to make it more immersive and exploratory by not having those systems; I can’t sample just the narrative content I find engaging and think that the game should flag more content (as optional, as less important) that I don’t find core to the experience; I see mountains of “content” without enough signposts to let me understand it and a progression through it and am simply overwhelmed in a way I do not think benefits the game or possibly even was the design intent.
“This is one of those doesn’t-respect-your-time games” generalises specific criticisms about various systems vs a personal view of what it could be into almost a genre - the too-long game. It flattens a meaningful discussion from which the speaker can make clear what their values are and so why a game doesn’t align with them (valuable information a reader can compare to their own values) into basically nothingness. It’s a topic where you need to be detailed because otherwise you’re basically just repeating the ancient “game long so good value” vs “game long so necessarily boring” war.
To some Assassin's Creed Odyssey provides the latitude to inhabit Kassandra’s life and soak in the world, to others the content is a slog and they’d much rather a far shorter core experience that the game does not appear to offer. Respect-your-time is extremely respect-your-time and the forming of critical consensus around a game will always flatten that even if the critics who voice their views are individually detailed in their analysis.
Every player is different and ideally games would all have a certain level of give to help accommodate as many people as possible but we’re not yet at that point. I’m also not sure the tools for signposting & flagging are sufficiently developed to be universally readable by the audience even when we add them. As we pick through the big lessons of the last decade or so of game design advances, the big open worlds and structures for repeatable content that span genres and platforms, this is hopefully something that will develop.
I have always found "respecting the player's time" to be a useful lens through which to consider game design. What needs to be here; what do I think makes a positive contribution to the core arc; just how much of a structure can be made optional and how do we signpost it so players can best dive into only the content they want? Extracting the most from the interactive nature of games means building structures that react to each player and customises how they experience a game, attempting to give the best progression to the most people we can reach (and even respecting that some players will not enjoy what we are making and should not be strung along). The thing is, this phrasing has become extremely common in games criticism, to the point of dilution.
“I don’t like this game as I feel like it is not respecting my time (and here are 6 paragraphs on exactly how)” is commentary on where I feel a game does not align with how I value the different activities it offers and what it considers core vs optional - I can’t fast travel in a game where I don’t agree with a design decision to make it more immersive and exploratory by not having those systems; I can’t sample just the narrative content I find engaging and think that the game should flag more content (as optional, as less important) that I don’t find core to the experience; I see mountains of “content” without enough signposts to let me understand it and a progression through it and am simply overwhelmed in a way I do not think benefits the game or possibly even was the design intent.
“This is one of those doesn’t-respect-your-time games” generalises specific criticisms about various systems vs a personal view of what it could be into almost a genre - the too-long game. It flattens a meaningful discussion from which the speaker can make clear what their values are and so why a game doesn’t align with them (valuable information a reader can compare to their own values) into basically nothingness. It’s a topic where you need to be detailed because otherwise you’re basically just repeating the ancient “game long so good value” vs “game long so necessarily boring” war.
To some Assassin's Creed Odyssey provides the latitude to inhabit Kassandra’s life and soak in the world, to others the content is a slog and they’d much rather a far shorter core experience that the game does not appear to offer. Respect-your-time is extremely respect-your-time and the forming of critical consensus around a game will always flatten that even if the critics who voice their views are individually detailed in their analysis.
Every player is different and ideally games would all have a certain level of give to help accommodate as many people as possible but we’re not yet at that point. I’m also not sure the tools for signposting & flagging are sufficiently developed to be universally readable by the audience even when we add them. As we pick through the big lessons of the last decade or so of game design advances, the big open worlds and structures for repeatable content that span genres and platforms, this is hopefully something that will develop.
Friday, 21 September 2018
The Free DLC: Driving Customer Delight
Quick aside: the recent posts about Rust have done very well. Enough that some of you are probably following the RSS feed just for that. I'm going to try and link some interesting Rust posts if I'm writing about a different topic that month. Today, Bryan Cantrill Falling in love with Rust. Both good introductory links and extensive thoughts on why Rust is worth caring about.
On to the main topic: Games are increasingly being updated after release as part of a "living game" strategy to continue to sell the game and any additional content. This is nothing new except in how common it has become. Back at the tail of the 1990s, if you wanted players to keep coming back and knew that the internet was now good enough to distribute patches then free content was how you drove sales and interest between boxed expansions. Total Annihilation was offering optional additional units back in 1997 on top of the balance patches and we'd come to each LAN and make sure everyone was up to date before jumping into the game. TA wasn't the first game to do it but additional maps, scenarios, and units was still notable in 1997 and became more common as games embraced online multiplayer as a primary focus (eg Quake 3 and Unreal Tournament). Even things like mods are part of this, being community-developed free DLC. Counter-Strike drove Half-Life sales at the development cost of building and maintaining those mod tools (which had originally been used to develop the game).
A decade later and consoles embraced online balance and even feature patches to expand games and eventually started pushing paid piecemeal DLC that wasn't just a different way to buy those traditional boxed expansion releases. Fancy buying a single new item for your RPG? What if you paid for it? What if you paid enough that, at those prices, a full game would cost at least several thousand dollars to buy? Hello the pricing model of many current F2P titles which offer to sell content (beyond the model that just sell temporary cheat codes to skip the most grinding activities or boost your stats).
Today, paid DLC of all types (from cheat codes to full expansions that massively extend the volume of content in the game) is everywhere. In a change from those original online games, multiplayer modes often sell additional maps which require you to keep paying to continue to play the current rotation of popular maps, somewhat like how MMOs require continued payments. But things are starting to swing back the other way with cosmetics being sold while the maps are given away to everyone who paid for the base game so they can continue to play. As someone with an interest in the longevity of the industry, I really hope this swing back continues. We need to delight customers, to make them feel like they got more than they expected. If we are making living games, we should avoid making things that feel like you've purchased a $60 storefront that's constantly asking for money rather than offering enjoyment. Over the last year, my feelings on two console exclusive (formerly tent-pole?) driving games have really driven this home.
I may have played more hours of Forza Motorsport 7 in the last year than GT Sport but by the end of Forza, I felt it was a slog; I saw the treadmill and only the proximity of the finish line kept me even considering continuing the career mode. I had access to over 500 of the 700 cars the game launched with; I'd become familiar with all 32 locations; but I had zero interest in paying for more cars (their plan for DLC). Regular challenges and (eventually) some more substantial patches to add back a working drag mode and fix the track limits were unable to really make me feel like I was engaged with the series continuing. The feast with a list of paid extras just made me feel bloated.
Forza was the series that got me to love cockpit mode, assists off, actually feeling like it was driving. But if it wasn't for the rental option then I'd probably not even look at the next release (I'll definitely be playing Forza Horizon 4 as a rental next month - once and done for pennies rather than the increasingly expensive bundled launch day editions that don't even guarantee access to all DLC content over the next year+ of updates; the last Horizon game being unable to run offline on Windows 10 and so effectively being a rental anyway). The monthly FM7 DLCs, heavily advertised in the game and covered in branding for TVs and snack food, offered extra cars on top of already so many, but with an order of magnitude higher price per vehicle and no extra locations. A handful of freebie cars (which appeared later on) are invisible when surrounded by the number the game launched with. But the lock icons stand out, as does the way the Forza series now includes day-one paid DLC in the form of a launch car pack (James Bond cars for Horizon 4) that only the most expensive edition ($100 up front) gets access to.
Meanwhile, Gran Turismo launched with under 200 cars (but most of them feeling distinct - no 10 cars just with different advertisers and identical handling/options) and with only 19 locations from which the various tracks (plus reverses) are assembled. A full career mode did not exist at all in the launch game but all of this was clearly messaged in the advertising (and a temporary sale price really helped push it from something ignored to a worthwhile gamble). As I talked about in my GotY discussion in December, GT Sport really appealed to me. It also found a way to keep me engaged with monthly free DLC that regularly added a new location and track options plus around a dozen cars each pack. The career mode arrived and every single area of the game received additional attention and content. All for free.
While Sony may be holding the next GT game for their next console, I feel extremely positive about the series (and buying that next game) and where it might be going. For a game that launched with less to do but then constantly gave me reasons to come back and try the new additions. The experience was of deciding that the game was probably worth taking a risk on, finding what was purchased was genuinely interesting to me, and then having my interest rewarded in the long term with new content all for the price I'd paid up front. There was enough content at launch, there wasn't a massive treadmill that ever felt full of filler just there to extend the total play time (without the variety to make it interesting), and then when I was ready to go back for more there had been patches that added the more I was looking for and didn't demand additional payments (with the associated question of if that value proposition was better or worse than the one offered by the base game or by choosing a different game).
There are many models for how to do this sort of continuous expansion based on player feedback and a long term plan. I'd call almost all games in Early Access an example of releasing a core experience and then iterating and expanding that with an engaged community who expect those additions to be included in the price they paid up front. We have two decades of using the internet to distribute feature patches and new content and how audiences react to that. The Paradox model ties a lot of paid DLC to base game feature patches that expand the core systems so they can milk a release while never letting someone feel like they paid up front for a game and now are left to just watch it progress without them.
My experiences certainly make me lean towards saying that a launch game shouldn't try to contain absolutely everything possible with a plan to aggressively monetise post-launch content. The audience will feel fatigued at too much content or start to divide it up into what feels fresh and what feels like filler (and when overwhelmed with content but lacking the tools to understand where it all can fit in as unique, it may increasingly look to be filled with filler). You can't under-deliver but over-delivery at the cost of having content to give away for free during the lifespan of the game feels like something to be considered carefully. The option is always there to push back more for an additional polish pass; rather than monetising it, use it to drive sales of the base game and delight your existing customers.
On to the main topic: Games are increasingly being updated after release as part of a "living game" strategy to continue to sell the game and any additional content. This is nothing new except in how common it has become. Back at the tail of the 1990s, if you wanted players to keep coming back and knew that the internet was now good enough to distribute patches then free content was how you drove sales and interest between boxed expansions. Total Annihilation was offering optional additional units back in 1997 on top of the balance patches and we'd come to each LAN and make sure everyone was up to date before jumping into the game. TA wasn't the first game to do it but additional maps, scenarios, and units was still notable in 1997 and became more common as games embraced online multiplayer as a primary focus (eg Quake 3 and Unreal Tournament). Even things like mods are part of this, being community-developed free DLC. Counter-Strike drove Half-Life sales at the development cost of building and maintaining those mod tools (which had originally been used to develop the game).
A decade later and consoles embraced online balance and even feature patches to expand games and eventually started pushing paid piecemeal DLC that wasn't just a different way to buy those traditional boxed expansion releases. Fancy buying a single new item for your RPG? What if you paid for it? What if you paid enough that, at those prices, a full game would cost at least several thousand dollars to buy? Hello the pricing model of many current F2P titles which offer to sell content (beyond the model that just sell temporary cheat codes to skip the most grinding activities or boost your stats).
Today, paid DLC of all types (from cheat codes to full expansions that massively extend the volume of content in the game) is everywhere. In a change from those original online games, multiplayer modes often sell additional maps which require you to keep paying to continue to play the current rotation of popular maps, somewhat like how MMOs require continued payments. But things are starting to swing back the other way with cosmetics being sold while the maps are given away to everyone who paid for the base game so they can continue to play. As someone with an interest in the longevity of the industry, I really hope this swing back continues. We need to delight customers, to make them feel like they got more than they expected. If we are making living games, we should avoid making things that feel like you've purchased a $60 storefront that's constantly asking for money rather than offering enjoyment. Over the last year, my feelings on two console exclusive (formerly tent-pole?) driving games have really driven this home.
I may have played more hours of Forza Motorsport 7 in the last year than GT Sport but by the end of Forza, I felt it was a slog; I saw the treadmill and only the proximity of the finish line kept me even considering continuing the career mode. I had access to over 500 of the 700 cars the game launched with; I'd become familiar with all 32 locations; but I had zero interest in paying for more cars (their plan for DLC). Regular challenges and (eventually) some more substantial patches to add back a working drag mode and fix the track limits were unable to really make me feel like I was engaged with the series continuing. The feast with a list of paid extras just made me feel bloated.
Forza was the series that got me to love cockpit mode, assists off, actually feeling like it was driving. But if it wasn't for the rental option then I'd probably not even look at the next release (I'll definitely be playing Forza Horizon 4 as a rental next month - once and done for pennies rather than the increasingly expensive bundled launch day editions that don't even guarantee access to all DLC content over the next year+ of updates; the last Horizon game being unable to run offline on Windows 10 and so effectively being a rental anyway). The monthly FM7 DLCs, heavily advertised in the game and covered in branding for TVs and snack food, offered extra cars on top of already so many, but with an order of magnitude higher price per vehicle and no extra locations. A handful of freebie cars (which appeared later on) are invisible when surrounded by the number the game launched with. But the lock icons stand out, as does the way the Forza series now includes day-one paid DLC in the form of a launch car pack (James Bond cars for Horizon 4) that only the most expensive edition ($100 up front) gets access to.
Meanwhile, Gran Turismo launched with under 200 cars (but most of them feeling distinct - no 10 cars just with different advertisers and identical handling/options) and with only 19 locations from which the various tracks (plus reverses) are assembled. A full career mode did not exist at all in the launch game but all of this was clearly messaged in the advertising (and a temporary sale price really helped push it from something ignored to a worthwhile gamble). As I talked about in my GotY discussion in December, GT Sport really appealed to me. It also found a way to keep me engaged with monthly free DLC that regularly added a new location and track options plus around a dozen cars each pack. The career mode arrived and every single area of the game received additional attention and content. All for free.
While Sony may be holding the next GT game for their next console, I feel extremely positive about the series (and buying that next game) and where it might be going. For a game that launched with less to do but then constantly gave me reasons to come back and try the new additions. The experience was of deciding that the game was probably worth taking a risk on, finding what was purchased was genuinely interesting to me, and then having my interest rewarded in the long term with new content all for the price I'd paid up front. There was enough content at launch, there wasn't a massive treadmill that ever felt full of filler just there to extend the total play time (without the variety to make it interesting), and then when I was ready to go back for more there had been patches that added the more I was looking for and didn't demand additional payments (with the associated question of if that value proposition was better or worse than the one offered by the base game or by choosing a different game).
There are many models for how to do this sort of continuous expansion based on player feedback and a long term plan. I'd call almost all games in Early Access an example of releasing a core experience and then iterating and expanding that with an engaged community who expect those additions to be included in the price they paid up front. We have two decades of using the internet to distribute feature patches and new content and how audiences react to that. The Paradox model ties a lot of paid DLC to base game feature patches that expand the core systems so they can milk a release while never letting someone feel like they paid up front for a game and now are left to just watch it progress without them.
My experiences certainly make me lean towards saying that a launch game shouldn't try to contain absolutely everything possible with a plan to aggressively monetise post-launch content. The audience will feel fatigued at too much content or start to divide it up into what feels fresh and what feels like filler (and when overwhelmed with content but lacking the tools to understand where it all can fit in as unique, it may increasingly look to be filled with filler). You can't under-deliver but over-delivery at the cost of having content to give away for free during the lifespan of the game feels like something to be considered carefully. The option is always there to push back more for an additional polish pass; rather than monetising it, use it to drive sales of the base game and delight your existing customers.
Friday, 31 August 2018
Rust: Fail Fast and Loudly
So recently I was chatting to some Rustaceans about library code and their dislike of a library that can panic (the Rust macro to unwind the stack or abort depending on your build options). The basic argument put forth was that a library should always pass a Result up to the calling code because it cannot know if the error is recoverable. The chapter of the Rust Programming Language book even lays out this binary: Unrecoverable Errors panic! while Recoverable Errors return a Result. As a researcher in debugging, I reached the point where I basically banned these terms from my lectures because they can potentially lead to this thinking that libraries cannot know they're in an unrecoverable state and so can only defer to what calls them.
An error (sometimes also called a fault or infection) in program execution or modelled / simulated execution is the consequence of executing a defective block of code and the resulting creation of an erroneous state for the program. This effect may not be visible and not all errors will be exposed by surfacing as a failure. An error is the result of a defect being exercised by an execution which is susceptible to that defect.
A failure in program execution or modelled / simulated execution is the surfacing of an error state by the observation of behaviour in violation of the program specification. It is therefore correct to say that a failure was experienced for a given test case due to a chain of erroneous states that originated with the execution of a defect that caused the error.
It's good writing but it can also potentially be read by those who really wish panics didn't exist as saying you just need to make sure every possible input into your library is valid. My stance is that an occasional small slice of invalid input being possible is actually important for code quality when writing in a language that can fail (it can also make some things a lot easier to write in practice). However, it must be clearly labelled as such, with no question about when you might panic. This is the contract you're writing and every good library should fully document the interface so there is no possibility of an unexpected panic.
To give an example from the Rust standard library (which is totally just a library and we should expect other libraries to conform to the same standards it uses - this is even more true of Rust than in other languages as Rust splits out the really core library code into the Rust core library). When you've got a vector and you need to divide it in two, split_off(at) is what you need.
Why does this panic rather than returning a Result and letting us decide if the error is recoverable or not? I can imagine many places where trying to split an array may not be the only thing a program can do to continue, a backup path could be constructed to continue operating under some circumstances if that failed but this library decision means the calling code cannot decide that. If you ask for a split at an invalid point then you get a panic.
It is because the library set a trap. It asks the calling code to know something about the object it wants to be manipulated. Because there is no reasonable way of asking for the array to be split in two beyond the end of the array, the only conclusion that the library can make about such a request is that it is unreasonable. We are past the point of executing a defect, we are swimming through an erroneous state, and it is time to fail so this can be caught and fixed. That also means no room to let the erroneous state accidentally ask to zero the entire storage medium it has access to and trash a last known-good state that might be used to recover later (or debug the defect). Any calling code that wishes to avoid this should catch the erroneous state and recover (if possible) before calling over the library boundary. The potential to get unreasonable requests allows our blocks of code to keep each other more honest by surfacing errors as failures.
It is a restriction that provides a higher chance of fixing a defect before we ship a product. We must strive to fail fast and sometimes that means using some small gaps between what is possible and what is permitted as traps to catch when errors have occurred. A library can be poorly constructed to panic when not expected (and declared) but the existence of panics should not itself be used as a sign that a library is of poor quality or to be avoided.
Terminology
Throughout CompSci literature, some terms relating to debugging are not used consistently. I'll start with the words I use (so I never have to write this in a blog post again). To illustrate the scale of the terminology issue, enjoy this quote from the 2009 revision to the IEEE Standard Classification for Software Anomalies:The 1993 version of IEEE 1044 characterized the term “anomaly” as a synonym for error, fault, failure, incident, flaw, problem, gripe, glitch, defect, or bug, essentially deemphasizing any distinction among those words.A defect (also called a fault, error, coding error, or bug) in source code is a minimal fragment of code whose execution can generate an incorrect behaviour. This is for some input (which includes the environment of execution), against whatever specification exists to declare what is and is not correct behaviour for the program. A defect is still a defect even if it is not exercised or does not cascade into error / failure during a test case. Defects can be repaired by substituting in a replacement block of code into the block reported as defective; this returns the program to executing in a way that does not violate the specifications.
An error (sometimes also called a fault or infection) in program execution or modelled / simulated execution is the consequence of executing a defective block of code and the resulting creation of an erroneous state for the program. This effect may not be visible and not all errors will be exposed by surfacing as a failure. An error is the result of a defect being exercised by an execution which is susceptible to that defect.
A failure in program execution or modelled / simulated execution is the surfacing of an error state by the observation of behaviour in violation of the program specification. It is therefore correct to say that a failure was experienced for a given test case due to a chain of erroneous states that originated with the execution of a defect that caused the error.
Setting a Trap
Having muttered about the language choices made in the Rust book at the top, I'm going to also praise how they actually resolve that chapter. The final section goes into detail about the pros and cons of calling panic from your code. It defines bad state in a way I might write myself in a practical programming guide. It even offers up the type system as a way to ensure your specifications for input aren't violated, with as much of the burden placed on compile time as possible.It's good writing but it can also potentially be read by those who really wish panics didn't exist as saying you just need to make sure every possible input into your library is valid. My stance is that an occasional small slice of invalid input being possible is actually important for code quality when writing in a language that can fail (it can also make some things a lot easier to write in practice). However, it must be clearly labelled as such, with no question about when you might panic. This is the contract you're writing and every good library should fully document the interface so there is no possibility of an unexpected panic.
To give an example from the Rust standard library (which is totally just a library and we should expect other libraries to conform to the same standards it uses - this is even more true of Rust than in other languages as Rust splits out the really core library code into the Rust core library). When you've got a vector and you need to divide it in two, split_off(at) is what you need.
Splits the collection into two at the given index.Here we have a clearly defined operation that does exactly what we want and comes with some important guarantees about how it operates. One of those details is that if we ask to split beyond the end of the array then it will panic.
Returns a newly allocated Self. self contains elements [0, at), and the returned Self contains elements [at, len).
Note that the capacity of self does not change.
Panics if at > len.
Why does this panic rather than returning a Result and letting us decide if the error is recoverable or not? I can imagine many places where trying to split an array may not be the only thing a program can do to continue, a backup path could be constructed to continue operating under some circumstances if that failed but this library decision means the calling code cannot decide that. If you ask for a split at an invalid point then you get a panic.
It is because the library set a trap. It asks the calling code to know something about the object it wants to be manipulated. Because there is no reasonable way of asking for the array to be split in two beyond the end of the array, the only conclusion that the library can make about such a request is that it is unreasonable. We are past the point of executing a defect, we are swimming through an erroneous state, and it is time to fail so this can be caught and fixed. That also means no room to let the erroneous state accidentally ask to zero the entire storage medium it has access to and trash a last known-good state that might be used to recover later (or debug the defect). Any calling code that wishes to avoid this should catch the erroneous state and recover (if possible) before calling over the library boundary. The potential to get unreasonable requests allows our blocks of code to keep each other more honest by surfacing errors as failures.
It is a restriction that provides a higher chance of fixing a defect before we ship a product. We must strive to fail fast and sometimes that means using some small gaps between what is possible and what is permitted as traps to catch when errors have occurred. A library can be poorly constructed to panic when not expected (and declared) but the existence of panics should not itself be used as a sign that a library is of poor quality or to be avoided.
Saturday, 28 July 2018
Empty Rust File to Game in Nine Days
I've been doing Rust coding for a bit now. Recently that's involved briefly poking at the Core Library (a platform-agnostic, dependency-free library that builds some of the foundations on which the Standard Library are constructed) to get a feel for the language under all the convenience of the library ecosystem (although an impressive number of crates offer a no_std version mainly for use on small embedded platforms or with OS development). I'm taking a break from that level of purity but it inspired me to try writing a game just calling to the basic C APIs exposed in Windows.
So I'm going to do something a bit different for this blog: this post is going to be an incremental post over the next nine days. I'm going to make a very small game for Windows (10 - but hopefully also seamlessly on previous versions as long as they have a working Vulkan driver), avoiding using crates (while noting which ones I'd normally call to when not restricted to doing it myself). I'm not going to rewrite the external signatures of the C APIs I have to call but I'll only be importing the crates that expose those bare APIs.
The Win32 calls go via Winapi, which is a clean, feature-gated listing of most of the Win32 APIs. The shellscalingapi is a bit lacking as it doesn't expose all the different generations of the Windows HiDPI API (and Rust programs don't include a manifest by default so you typically declare your program DPI-aware programatically) which means you have to declare a few bindings yourself to support previous editions of Windows. But generally it makes calling into Win32's C API as quick as if you were writing a native C program including the required headers. You could probably generate it yourself via Bindgen but the organisation here is good and it's already been tested for potential edge cases.
Vulkano exposes the underlying C binds via the Vk-sys crate. It has no dependencies (so it's what we want: just something to avoid having to write the signatures ourselves without obfuscating anything going on) and while it's not updated to the latest version of Vulkan (1.1), we're only doing a small project here (so it shouldn't matter at all). The function pointers are all grabbed via a macro, which is a bit cleaner than my previous C code that called vkGetInstanceProcAddr individually whenever a new address was required (to be cached). Of course, other areas are down to just the barest API which means looking up things like the version macro.
So at the end of day 1, we've got a basic triangle on the screen working (with a 40kB compressed executable, most of which is Rust runtime/stdlib as Rust libraries default to static linking).
So I'm going to do something a bit different for this blog: this post is going to be an incremental post over the next nine days. I'm going to make a very small game for Windows (10 - but hopefully also seamlessly on previous versions as long as they have a working Vulkan driver), avoiding using crates (while noting which ones I'd normally call to when not restricted to doing it myself). I'm not going to rewrite the external signatures of the C APIs I have to call but I'll only be importing the crates that expose those bare APIs.
Day 1
The day of building the basic framework for rendering. Opening with a Win32 surface (normally something you'd grab from eg Winit) and then implementing a basic Vulkan connection (which would normally be done via a host of different safe APIs from Vulkano to Ash, Gfx-rs to Dacite).The Win32 calls go via Winapi, which is a clean, feature-gated listing of most of the Win32 APIs. The shellscalingapi is a bit lacking as it doesn't expose all the different generations of the Windows HiDPI API (and Rust programs don't include a manifest by default so you typically declare your program DPI-aware programatically) which means you have to declare a few bindings yourself to support previous editions of Windows. But generally it makes calling into Win32's C API as quick as if you were writing a native C program including the required headers. You could probably generate it yourself via Bindgen but the organisation here is good and it's already been tested for potential edge cases.
Vulkano exposes the underlying C binds via the Vk-sys crate. It has no dependencies (so it's what we want: just something to avoid having to write the signatures ourselves without obfuscating anything going on) and while it's not updated to the latest version of Vulkan (1.1), we're only doing a small project here (so it shouldn't matter at all). The function pointers are all grabbed via a macro, which is a bit cleaner than my previous C code that called vkGetInstanceProcAddr individually whenever a new address was required (to be cached). Of course, other areas are down to just the barest API which means looking up things like the version macro.
So at the end of day 1, we've got a basic triangle on the screen working (with a 40kB compressed executable, most of which is Rust runtime/stdlib as Rust libraries default to static linking).
Subscribe to:
Posts (Atom)