Veles 最近的评测

Bottom line up front: Marred by technical flaws and questionable design decisions leading to tedious gameplay. There's a decent game underneath if you're patient - and I mean VERY patient - but better games exist to satisfy the same itch


I won't deny that Satisfactory's premise is an enticing one - that being a laid-back factory builder with creativity-focused building systems a la Minecraft - but after the opening hours and as you reach the midgame, tedium begins to set in at a rapid pace.

The game runs surprisingly well for what it is, with some hitches and stutters here and there depending on how built-up an area you're in. Thankfully, global illumination is an optional feature (as it should always be until even the most entry-level gaming rigs can maintain a stable 60 FPS with it), but depending on your rig you're bound to feel stutter in more built-up areas.

This is where the game begins to falter. In the beginning, the scale of production is so small that it feels overwhelming at first, takes a bit more time than you'd like, but it's fine because you're learning the game and you're surely going to get access to more automation tools as you progress, so you move on from your first objective of 50 smart plates, a relatively complex recipe for this stage of the game requiring a decently large logistical footprint, which are produced at 2 per minute, requiring 25 minutes for your first objective - arguably a modest pace. Then you're given your next objective, and now part of your next goal is a whopping 1,000 smart plates plus a few hundred miscellaneous items. That single factory that completed your previous task after 25 minutes will now take 8 hours and 20 minutes to complete a portion of your next task. If you built another smart plate factory, you would cut the wait down to 4 hours and 10 minutes, but what if you could just... double your output magically?

Somersloops are effectively output multipliers at no additional cost aside from a negligible increase to power draw which is usually less than building another factory if you use them on the end product. Most buildings accept them, and they essentially consolidate an entire extra production line into your factory for a cost we've already established as free or even net-positive. There are also power shards and the overclocking system, which increase production and consumption together in a linear correlation which lets you tweak a building's consumption for getting the right ratios, and finally there are dimensional depots which are a better version of Factorio's logi-bot network in that it's basically cloud storage you can push to and pull from while anywhere. But now you might be wondering how do you acquire these items...

Expanding the factory's industrial capacity with these items requires venturing outside the safety of your factory and fending off the local wildlife. Their AI is simple, bunnyhopping with the first projectile weapon you get (the rebar gun) is enough to cause even the highest tier of enemies to falter, the world is a chore to navigate with the random bits of terrain poking and prodding through the dirt at random and the obscene quantity of sheer cliffs. You get a scanner to locate your bounty, but it doesn't account for height differences, so whatever you're looking for may very well be in a hole somewhere whose entry point is hidden behind a swathe of overgrown moss and jagged rocks.

Remember those rocks and sheer cliffs I mentioned? These form natural chokepoints to and from different areas. If you were making an action/adventure game with RPG mechanics, this would be a great idea to funnel players through certain places and get them on a dedicated progression path, but in an open world factory builder, it falls apart given that you can build anywhere, so your solution to navigating this jagged landscape is building a bunch of ugly towers and ramps to the heavens, and building factories to meet your production quota in a reasonable timeframe requires ample space, which only a few locations in the game have, and then those locations are sometimes lacking resources, which gets us to the matter of long-range logistics...

Trains are serviceable. The rail building system is janky and requires hacking together what feels to be an unintended solution with the building system before, but trains feel serviceable. They're like a more primitive version of OpenTTD or Factorio trains albeit with one very fatal flaw in that their paths are pre-baked and they will not take detours or stack up in waiting areas, which hurts their versatility.

Trucks and tractors are... abysmal, so bad that they inspired me to write this whole review, in fact. You place a vehicle, you drive the route (and hope you don't make a mistake), and pray the vehicle doesn't reverse into any barriers, get itself stuck, drive into water, wander off its designated path, or get stuck on wandering fauna. The best solution to this issue is building a road, but in that case you could've used that time to build a train - at least trains don't need to be fuelled directly.

Which leaves, aside from rather simplistic drones, belts. They can clip through rocks and ceilings, be raised to infinitely tall heights, provide constant throughput, clip through one another with no repercussions, and they're cheap and easy to build at scale. There is no reason not to cover the entire world with belts to counteract the tedium of getting things from A to B - they can even be used to circumvent those aforementioned natural chokepoints by abusing their clipping.

• Blueprint scale is limited, and rather than starting you with one universal blueprint format, the game gives you 3 different blueprint formats (none of which are directly usable in blueprinters of the next tier). You cannot load a blueprint from a mk1 blueprinter into a mk2 or mk3 blueprinter, but you can place the blueprint's contents within the blueprinter, save it, and then load the new blueprint's contents.
  
• Movement around the factory comes far too late. The hoverpack (which I'd argue is essential for building vertically) is unlocked in the lategame despite having the limitations of requiring a power network, effectively making it a less versatile jetpack
  
• Selected paint swatches are server-side and not client-side. If your friend in MP selects a paint swatch, then that's everyone's paint swatch.
  
• Some dialogues (primarily for picking up alien artifacts) are strangely client-side, but others such as milestone progressions and research unlocks are server side and play for everyone
  
• The (honestly quite impressively-simulated) fluid dynamics system could use more explanation in-game

There's a good game somewhere underneath, but the game feels unfinished despite having long been divorced from its Early Access phase. Better games in this genre exist, and I'd recommend Shapez 2 instead
https://psteamproxy.yuanyoumao.com/app/2162800/shapez_2/

It's been a while since I got started with the factory automation genre. First it was Factorio (which I'm not a fan of), and then it was Minecraft tech mods - and I preferred the latter.

Shapez 2 feels more like a Minecraft tech mod pack.

Whereas in Factorio you start with a bunch of raw iron, copper, stone, etc, and you progress up to mildly complex recipe chains that require more number balancing than procedure, Shapez 2 feels like taking Factorio's scale and combining it with a well-made Minecraft tech mod pack in the sense that procedures that get expanded upon with the same kind of open-ended problem solving you'd find in those mod packs

As a hobbyist programmer, it reminds me of developing APIs- let's say, for instance, you want to make a platform that cuts shapes. Well, you figure out your target, design fit it within your space constraints in mind, and after having developed it, you have what's essentially a giant black box which you can call upon whenever you need this thing done, no need to stress about the low-level "how" of it anymore, it'll just do what you want. Even though other games in the genre have this same blueprinting feature, here it feels less tedious in a way - and in others, more meaningful and less same-y. I don't have to wait for slow construction bots to navigate over to the build site and place things down, in short...

In Factorio, everything's essentially what the Minecraft tech mod community would refer to as a "magic block" - only much simpler, as assemblers will cover just about all of your manufacturing needs, and loading machines is as simple as throwing inserters around the building and letting you control what direction things go out. Contrast this with Shapez 2. The logistical footprint of a cutter requires different considerations when designing for tileability as opposed to designing for tileability in painters, and your input and output directions are static.

While I'd argue that the math required for this game is much simpler as well (the game will even spell the ratios out for you), I'd also argue the math required for Factorio wasn't anything brain-bending or pertinent to the experience. This is, of course, not counting if you decided to go for the wholly optional endeavour of learning the circuit mechanics, where an adequate grasp of boolean algebra and computer science helped (which Shapez 2 also has, and I'm currently in the process of experimenting with).

The music is quite good. I even bought the Supporter Edition for the soundtrack and listen to it outside of the game. It's good study music in a similar vein to the Satisfactory or Factorio OSTs, but a bit higher energy while still firmly remaining solid background noise. Sound effects are oddly satisfying (even when it's just copy pasting a platform for the nth time), Blueprint menu has great sorting capabilities with its folder system, the UI does its job like any good UI should, and performance is great under Linux - no sore spots here.

I'd wholeheartedly recommend Shapez 2 if you like the procedural nature of Minecraft tech mods, enjoyed the factory design aspects of Factorio, or are just looking to get a feel for the genre.

Unfortunately, as is the case with most games in the modern day (Especially those running UE5), this one seems to have foregone any and all optimization. On my setup (Ryzen 7600, Radeon 7600, 32GB DDR5, Loaded on an nvme drive), I could easily measure my framerate in the slums and forest maps in seconds per frame, and that was on the lowest settings at 1080p with DLSS on. Stay away until the game's performance gets fixed.

Bottom line up front: Too expensive, lacking in certain areas, modules are so granular that their price is hard to justify. Better alternatives exist

Train sims are an interesting niche - one that I happen to take an interest in, and one thing that I believe inhibits its potential growth is its most mainstream representative being overpriced. Let's look at the price for a standard TSW module for instance. I'll use MBTA Boston Commuter in this example, as it's the one I have the most experience driving. You get access to two locomotives with their respective cab control cars, giving us 4 drivable vehicles in total. On top of this, you get a route that's advertised as having 45 miles of track which takes roughly an hour and a half to drive down. Now, looking at this, how much do you think the DLC would be worth? $15 USD? Maybe $25 USD if we're generous? Well, you're mistaken. The price sits at an outrageous $40 for a route wherein hardly anything happens, you rarely if ever have to react to signals, and you almost never stop anywhere but your designated stations. The only saving grace here is that the route happens to go on sale for a much more generous $15 - a price that I would recommend it at.

Well, let's ring off a few more

• LGV Méditerranée - $30 for a single locomotive and a route with practically zero variation (at least MBTA commuter has a branch line)
  
• NEC: Boston Providence - $30 for 3 locomotives max (1 new locomotive if you already own MBTA Commuter like me), 46 miles of track, 1 hour of back and forth driving. Potential for more train variety on the route if you spend $23 entire USD on the Acela. Yes, that's right, $23 for a single trainset that doesn't even have a fully functional cabin.

Unfortunately, no. On my current setup (Ryzen 7600, Radeon 7600, 32GB DDR5, Linux Mint, latest Proton version), frame drops seem inevitable on even medium settings, and I've heard reports of this game bringing much more robust rigs to their knees. On top of this, the game seems to throw in advertising at every nook and cranny it can worm its way into, as you will be greeted with ads for more overpriced low-value DLC until you find the setting in the options menu to turn them off.

• MaSzyna is a great FOSS railsim based around Polish rail operations - not only is it free, but it's also addon-friendly (helps to have a translator ready if you don't speak Polish, though)
  https://psteamproxy.yuanyoumao.com/app/1033030/MaSzyna/
  
• SimRail is yet another railsim based around Polish rail operations. The DLC pricing at the moment seems a lot more generous than Dovetail's offerings. It also has multiplayer. Haven't played this one yet as of writing this review, but I've heard positive things.
  https://psteamproxy.yuanyoumao.com/app/1422130/SimRail__The_Railway_Simulator/
  
• Open Rails is a bit of a deeper cut. For one, you can't find it on Steam (and I haven't played it myself), but the modules seem a lot more fairly priced and sometimes are even free.

While I don't have as much to say about this game as I do about most others I've played, I will say this - it's a fun little sandbox to kill time in with plenty of depth to really sink your teeth into should you choose to delve deeper (it's more than possible to build a working computer in this game, for instance). Give it a try if you have the time, as that's the only cost.

That's what Steam asks me when I post a review, and for Factorio, as much as it goes against the grain, I'm going to have to say no, unfortunately.

That is definitely true. Some standouts:
- The amount of QoL features present is something other games in the genre should strive for.
- The optimization and developer transparency is second to none in the industry and should be displayed as an example of how you develop software and keep your community informed
- It runs flawlessly under Linux without the need for Proton compatibility layers thanks to having a native Linux build

Unfortunately, the game is marred by some... questionable pricing and design choices. The elephant in the room first - the pricing. I bought this game back when it was much cheaper (sometime back in 2018), and I'd argue that's a great price for this game given the content on display (about 30-40 hours for a playthrough if you take your time), but looking at it now, with its current price tag, I cannot recommend it. While yes, I may seem a tad hypocritical with my 130 hours in this game, aside from a bit of mucking around in multiplayer, most of those hours consisted of a much more clueless version of me bouncing off this game repeatedly over the years before finally sitting down and the game seriously, only to be beat the base game underwhelmed and go, "wait, that's it?"

And what's in the base game? For a game that advertises itself as a logistics sim, there's little in the way of complex logistics required to actually beat the game. Look at a diagram of a printed circuit board and you've already solved the game's biggest logistical challenge - massing resources and splitting them off where they need to go. I had hoped that maybe the mathematical side of the game could change things, but no, it's all stuff you could solve in 5 minutes with algebra you'd learn in secondary school. There's an entire digital logic system you can play around with, but aside from one optional case (which I found quite fun, granted), it has next to no practical use when it comes to your main goal of launching a rocket. So how do we make an easy game harder?

Every hour - or every half hour if you're unlucky, the fine residents of the planet will decide to waltz up to your buildings with an assortment of instruments of deconstruction and subject you to forced rapid disassembly, of course!

Jests aside, the biters feel like an impedance and an afterthought to pad out a simple game with more uninteresting busywork: clear their nests with a shallow combat system whose automation comes too late, mindlessly build defences by clicking and dragging straight lines of belts full of items whose recipes are so simple you can just mindlessly plop down factories making them without second thought, or just turn them off and live a happier life without their constant tantrums...

The game weren't designed around them existing. There exists an entire research tree dedicated to killing them better, and there's an entire crafting tab dedicated to making the gear that kills them faster.

Admittedly, I didn't touch mods, and truthfully, they shouldn't be taken into consideration when reviewing a standalone game such as this. This game is not advertised as a modding platform, and thus it should be judged for what it provides. Mods may eclipse the quality of the base game, yes, but they don't render any verdict on the base game's quality moot.

It's a great piece of software, but as a game, it stands marred with game design choices that are either frustratingly simple or frustratingly inconvenient. Spend your money elsewhere.

It pains me deeply to recommend against this game.

But there's one thing I have no qualms with - the art direction is stunning, as is to be expected from a Team Cherry game. It really does feel like an interactive painting at times and you can really tell that's where most of their effort went.

Keyword: Most of their effort.

Act 1 is great.Aside from some teething issues (and one particularly savage beastfly whose difficulty feels thematically inconsistent), it feels good learning to use Hornet's kit to deal with what bosses throw at you in a 1v1. This even extends to some Act 2 bosses (Cogwork Dancers is one of my favourites, despite not being a 1v1).

But this is where my praise ends. See, instead of giving us a good variety of those aforementioned 1v1s (and we will be focusing on those, as the game's basic enemies are hardly enough to write about on their own), Team Cherry instead decided that the best way to make bosses difficult was by... prepending a gauntlet to the boss or allowing them to summon ads instead of fleshing out their moveset. Even on the main path, sometimes you'll find a room that looks like it would make for a great boss arena only for it to simply throw about 10 waves of enemies at you in lieu of a well-made encounter.

So, late into my playthrough, I realized something. It felt like this aspect of the game - the simple boss movesets - were an attempt to cut costs and pump the brakes on an already overambitious project, and that their solution for these simple boss movesets was to disguise it with a facade of difficulty by layering on enemy gauntlets and ad spam. The end result pales in comparison to most fights in the original Hollow Knight (yes, even the Watcher Knights).

Now, given that there are a handful of great boss fights in this game, would I recommend the game? Unfortunately, no. I don't believe it's worth diving face-first into mud hoping to find a nugget of gold.

Alternate media recommendations: The original Hollow Knight, Castlevania Symphony of the Night.

From what little gameplay I saw of it prior to purchase, I thought this game be something along the lines of a Diablo-like. After that, I got into the game and realized just how core base-building was to the experience. "Okay, so this is a Minecraft thing", I thought to myself - though I still had an open mind, and I was still thoroughly enjoying myself. This game's opening act is indeed very strong.

Progressing further, I found that the game seemed to want to do nothing more than waste your time: for some reason, you cannot fast travel with certain resources in your inventory. "Can I turn this off?" I asked myself - turns out you can, in fact, turn it off, but the reasoning behind why it remained a default option eluded me.

Past that, my friend and I reached the midgame, and it dawned on us both: we're sitting here twiddling our thumbs for an inordinate amount of time waiting for a meter to fill up enough times until our resources are refined into something that we can actually use, but then those refined resources take *time* to craft into something usable, further compounding the issue, so the thumb-twiddling continues. On top of that, we spent hours running in a loop farming for research materials which, due to the RNG nature of the research system, didn't give me the armour set for my particular build until it was the last unlockable option. Grinding can be fun, yes, but only when the game is built around mastery of content. Running through the same village mindlessly grinding the same trash mobs waiting for them to drop research mats isn't *good* grinding - it's padding. If you want a good example of how to do grinding right, go look at Monster Hunter (although I have my fair share of grievances with MH, it at least handles grinding in a more enjoyable way.)

But then it dawned on me.

This isn't Diablo, nor is this Minecraft - this is just vampire ARK, Rust, or Conan. The game is built around PvP and big servers while solo or 2-player gameplay feels dull and monotonous. Even after maxing out crafting modifiers (which only seemed to make the crafting timer tick down faster without changing the displayed crafting time for... some reason), the crafting still felt sluggish. The default fast travel limitations? Meant to make resource runs feel more dangerous. The default crafting rates? Meant to give an edge to bigger clans - it all made sense.

And, as an aside, for whatever reason, there is absolutely zero possibility of resizing the UI so, if you're hard of sight like me, you'll be hard-pressed actually glean any info from the user interface. It feels like a good game marred by a lack of focus full of questionable UX. It starts out fun, but eventually, as you progress, the cracks begin to show.

Prey is essentially what happens when you give a bunch of ImSim devs a Triple-A budget, and that budget shows from the world design all the way to the interactivity present in the game world. It feels like the devs thought of almost every playstyle you could imagine and then some. Playing with weapon durability and no repair skill for whatever reason? You're in luck, because now all those useless weapons lying around that were basically junk items before are now your only lifeline and the only thing between you your alien adversaries. Never in any other game have I been so happy to find two basic shotguns in a single room.

Maybe you want to play the game without upgrades? You can do that. Want to play without alien powers? You can do that. Want to play with only alien powers and roleplay a space wizard? WELL OF COURSE YOU CAN DO THAT!

Even the little environmental details have lots of love put into them. The whole space station has this Neo-Art-Deco look to it that gives me the vibes of a more future sci-fi Bioshock, and much of what would usually get relegated to its own separate UI (interacting with computer screens and ammo counters on most weapons come to mind immediately) is a diegetic part of the game world - it's really top-notch stuff, and zooming in on little charts and whiteboards put into the levels is without a doubt going to show you something neat eventually (your observations may also be rewarded)

I can't say much about the sound design other than it's also suitably great - weapons sound punchy as they should, it's mixed well, and things overall sound how they should (though, some sounds like the sneak attack crit SFX can get repetitive).

Make no mistake, though, this game can be BRUTAL for new players, but don't feel discouraged - it's all a part of the atmosphere. As a bonus, the game runs practically flawlessly on Linux.

Recommended similar media: Dead Space, Deus Ex (2000), John Carpenter's The Thing, 2001 A Space Odyssey

What if you made Deus Ex, but we took out the streamlined level design, replaced the smart writing with memes and off-colour humour, and made the setting "Polish" to the point of feeling like disingenuous satire written by a foreigner?

You'd get Peripeteia.

Somehow, despite looking about the same as Deus Ex graphics-wise (worse in places, actually), it runs worse than Cyberpunk 2077.

Weapons feel wonky (the muzzle climb on some of these weapons, ESPECIALLY the Glock, is exaggerated more than your average "realistic" stalker mod; aiming down sights with the Tokarev may as well involve the character putting their nose on the front sight, etc), weapon inertia is exaggerated whenever you turn your body, and the inventory system adds unnecessary bloat between combat sections, and speaking of bloat between combat sections, I sure do hope you enjoy spending 5 minutes climbing stairs to find an item that seems to exist in its current position with a basis in logical consistency rivaling the average Fallout New Vegas weapon mod.

That's what this game feels like - it feels like someone's mod: down to the amateurish overuse of techno music blaring into your eardrums whenever you're around enemies signalling that they exist - it's like if Deus Ex constantly played combat music when you were sneaking around.

For $25, it's not worth it. Go play Deus Ex; go play Prey; go play Dishonored - don't play this.