mojical

I thought so as well, but given the Deck's popularity I can see both anticheats switching to an opt-out model or asking game developers about it instead of said devs having to take the initiative. Hopefully LO transitioning to EAC won't break compatibility. I'm holding off on making a more detailed tutorial or video until 1.30 reaches Live due to this. Nobara is a great distro choice for gaming on AMD cards, even if the method I described is universal and can be achieved with Flatpaks.

Surprisingly, it took very little effort to make this new build run on Linux via Proton/Wine, unlike Live which crashes immediately at the district select screen at best. That alone is a huge step up in my opinion and a great incentive to start playing again, because I have pretty much ditched Windows completely other than a tiny partition to run a few well known benchmarks and overclocking tools. This also means the Steam Deck and devices with a similar software stack should be able to run the game now, though it is obvious this game is best enjoyed with a keyboard and mouse. The (simplified) steps I took for this were to install the game using Bottles and the default Soda runner, install the dotnet48 dependency in the same prefix, then have a separate entry for APB.exe and APBLauncher as the launcher can not start the game on Linux yet. Once 1.30 is out on Steam the process will be much more simple. Performance is comparable to Frosi's results on Windows and there are no issues with the anticheat, both BE and EAC support Linux by now. Here is a frametime plot complete with test specs and settings: https://flightlessmango.com/games/1338/logs/3621 I saw this being requested by someone else in the AMA so I hope others (especially with Nvidia hardware) give this a try and check if it works on their setup as well.

I've tested the game on a 12700K CPU and didn't run into any issues whatsoever, no need to disable E-Cores. In fact, these E-Cores help get Windows background services off the main cores which is always welcome when playing a CPU intensive game. I use a tool called Process Lasso to manually assign E or P-cores to processes but automatic scheduling works just as well with APB, it gets sent to the correct P-Cores by default on Windows 11. What has very noticeably improved are initial load times and alt-tab times. Ingame performance is also great with stable FPS but the engine's shortcomings are still present, such as garbage collection, stuttering with higher detail settings... I have a feeling there won't be a CPU that can mask these completely until the game itself is fixed.

It's not just about bad PCs vs. good PCs, for example RTX card users face constant "out of memory" crashes due to a long standing bug if their graphics settings are set too high. This is something that can be mitigated with configs by improving graphics as much as possible while keeping the game stable. Similarly, garbage collection is far from ideal on Live so the ability for players to tweak it is more than welcome. Also, most of the time configs do the opposite of lowering graphics - they improve the stock Minimal preset which is arguably the only one that runs properly without stutters and random slowdowns. The point of using a config is to avoid unnecessary and costly 2010 "eye candy" and focus on APB's great customization potential. No one should be judged for playing settings that are convenient and unobtrusive to them, regardless of how good their hardware is. I've made that mistake in the past and realized how stupid it was over time. It's a real shame that this thread is still rife with posts that equate configs with cheats, but in a way I also envy those who are easily satisfied by 30fps with 3 second stutters every couple of minutes.

Nice idea, I tried to comment on every weapon (except grenades, don't think anything needs to be done there): https://docs.google.com/spreadsheets/d/1vTz7LxjALaLC25i89gwO0zuXulqG-G_NYHvxuO0b-QY/edit?usp=sharing I feel like one of APB's issues in terms of weapon balance is how haphazardly overdamage (damage over the minimum necessary for a gun to achieve its TTK) is distributed accross weapons. For instance, the ISSR-A desperately needs some more damage while the .45 and Obeya would still be more than good enough doing a bit less damage per shot. It seems to me that adding overdamage to the Harbinger helped it more than any of the other changes due to CA3 being everywhere.

When tracer mechanics are reintroduced with the engine upgrade, silencers will gain a purpose (hiding tracers that can give away positions), albeit situational. I do agree that it is pointless to assign them a downside in the current game, them occupying a slot/color is already enough of a penalty. As for the changes, I like most of them, especially the Scoped N-Tec and Tommygun. I'm not a big fan of the Misery's changes since it feels like a nerf to me, the weapon has issues more severe than its recoil, such as its effective range and extremely slow marksmanship movement speed, and its jumpshooting gimmick has now been taken away. I'm not sure about the OCA changes since SMGs are difficult to justify over shotguns without a range advantage. I do agree with the general idea (of OCAs needing a nerf) but perhaps 17.5 is going too far, especially for the Whisper which will be heavily penalized by this. Its extra accuracy will be near useless with such a low range and the inability to equip Improved Rifling 3.

I really can't believe the response to what was said in my thread about the Scoped N-Tec was to nerf it further. At least the Harbinger changes make quite a bit of sense. It will be interesting to see how it compares to RSA/ACT now.

This is exactly the opposite of what I meant to say in my post. To clarify, Battleye does NOT currently work with Proton/Wine. For DEP no special patches are required anymore AFAIK.

DEP can be fixed, Battleye can not. When EAC was a thing it was possible to run the game but there were some post-login functions that caused a crash, and the launcher is broken unless original .NET Framework DLLs are loaded manually. If you absolutely must single boot Linux and run APB, plugging your GPU into a VFIO virtual machine is your only option and it's incredibly complicated to set up properly. I've had the exact same results with Proton, both the official and GloriousEggroll variants. What I haven't tested yet is the new engine, although the BE issue can only be solved if Little Orbit officially supports Proton.

This is a very good idea in my opinion. Both Kevlar and Flak arguably suffer more from not being stackable with CA3 than from their own downsides, even though consumables partly alleviate this. However, I wouldn't reduce CA3 to a single level, but turn CA1 into "CA2.33", CA2 into "2.66" and CA3 being the exact same as it is now, to allow for more flexibility in between current CA2 and CA3. If D is the base (unmodded) delay before regen and T is the time it takes to regenerate, we would have according to the mods' stats: Current CA2: 0.5D, 1.5T (-50% delay, +50% time) Current CA3: 0.2D, 2T (-80% delay, +100% time) Here's what new stats for CA mods could look like: Proposed base: 0.5D, 1.5T (like CA2) Proposed CA1: 0.4D, 1.65T (-20% delay, +10% time) Proposed CA2: 0.3D, 1.80T (-40% delay, + 20% time) Proposed CA3: 0.2D, 2T (-60% delay, + 33.3% time) (just like current CA3)

I feel like this is a special case because unlike all other variants where changes described as a buff were still a buff, in this particular case it has the exact opposite effect. This and OCA/RSA fire rate changes aside, I think the latest balance patch was spot on. However, it is unfortunate that I don't get to read how the design team justifies to downgrade the Scoped N-Tec, because it all doesn't make any sense to me as an actual user of said gun. It wasn't overused/overpowered at all with a 1.6 vs 2.4 advantage so why would it be with a 1.6 vs 2.0 one? Reducing weapon variety is a sensible idea, but if you are to normalize away a variant's advantages then you might as well do the same with its disadvantages. Else the game will be littered with dysfunctional variants of free weapons which are harder to obtain or cost real money. Besides, if the weapon was overpowered enough to warrant this change, this would have been noticed at most days after the patch, not weeks. ARMAS states that there is less recoil and from my experience using the weapon this makes a lot of sense. The heightened zoom increases perceived recoil which is likely why it was designed this way to begin with. Flavor wise this can be attributed to an improved stock as seen on the pre-skins model, much like its maximum bloom advantage over the regular variant that it has now lost. I will try to test it more in-depth when I have time to make sure that I'm not talking nonsense, but in any case if I am, then ARMAS is wrong as well... I completely agree with this. In fact, the only way to obtain a 3 slotted version of the Scoped N-Tec is to buy the VAS Sceptre reskin, which in my opinion looks even worse than the remade skinnable model. At the very least, ARMAS versions could have 3 point sling added, because arguably HS3/IR3/3PS3 is the only viable mod setup on this gun and there is no downside to that particular mod.

As much as I agree that the Scoped N-Tec is underpowered overall compared to the standard version, it does have upsides, they just do not show on APB DB. Upward recoil is slightly reduced as is damage dropoff due to using a different curve. This however only translates to 2-3m extra 6 shot range in practice and doesn't compensate for worse FOV and the practical loss of Hunting Sight benefits when strafing. I feel like the Scoped was at its very best during the Improved Rifling (fire rate affecting) changes, so somewhat increased range in return for slightly decreased fire rate would be a very welcome change in my opinion.

Little Orbit explicitly stated that they did not change the scoped N-Tec. It's in the January 29th patch notes, which is when the N-Tec was nerfed to 2.4: https://www.gamersfirst.com/apb/patch-notes#collapse37

The Scoped N-Tec has had its shot modifier cap changed from 1.6 to 2.0 in the latest patch. For the N-Tec this change was a buff (it started at 2.4) but for the Scoped variant, which was at 1.6 previously, it is a nerf. I am not sure if this is an oversight (since nothing about this weapon in particular is said in the blog post) or if this is intentional. Given the already very low usage of this weapon I think the former is more likely.

When comparing the ACT and RSA, has the lack of HS3 availability on the RSA been taken into account? I think this is one of the reasons why RSA is seldom used currently. I don't think High Magnification Scope can be considered an equivalent as it is way too clunky, especially for a secondary. In my opinion a new JT/Armas RSA variant with HS3 would be welcome, because I believe the ACT GM will still be dominating in the long range secondary role. Changing the Hunter could also be an option but messing with weapons players have paid money for tends to be a slippery slope.

This is exactly the way I see it, there is just too little competition for the PMG now that the OCA has been left without any (significant) advantages over it. Some prefer the OCA's firerate since it helps mitigate the effects of RNG and prevent "bad streaks", but looking at the cold hard numbers the PMG will simply hit more on average due to its better base accuracy. Also, for some reason the PMG can have a crouch bonus and equip Cooling Jacket at the same time, which the OCA cannot as crouching only improves accuracy on the Whisper. If the OCA was actually changed according to these values instead of receiving an unexplained (and aside from APB DB, undocumented) last minute nerf, it would have helped both F2P options stay competitive. This would not benefit the likes of the Norseman and Tommygun but it is certainly better than to have the category dominated by PMGs and Manics. Surely it's a tiny difference, but I do remember the OCA feeling best balance wise after RP's initial 0.68s buff. The 0.64s buff afterwards, on the other hand, was complete overkill and is what caused LO to nerf it again bringing it back to square one. All things considered though, the OCA is still above average as far as SMGs go, it's just stuck in a strange no man's land compared to the PMG/Manic, and the likes of the Norseman/Curse/Tommygun.

Most likely not as this tweak is meant to help APB put some assets in RAM that would otherwise go into a HDD/SSD. APB has no choice as to use the HDD/SSD because it is limited to <4GB RAM due to being a 32 bit application. Most other games are either 64 bits already or don't have issues with the 4GB limit to begin with, if they are older games. For those asking whether this will eliminate every stutter, garbage collection stutters are (as has been said already) not affected by this, but unfortunately GC is only one of the multiple causes of stutters. I have collected some GC times to prove this. On all 4 runs, I logged into the same character, went to social and idled a bit in front of the clothes shop to cause the maximum possible disk load: Spinner/ HDD without ramdisk Line 6645: 13:43:38 - DevGarbage: 71.101600 ms for realtime GC Line 6914: 13:43:59 - DevGarbage: 57.993300 ms for realtime GC Line 9494: 13:45:02 - DevGarbage: 71.593300 ms for realtime GC Line 11535: 13:45:07 - DevGarbage: 61.530200 ms for realtime GC Line 11970: 13:45:07 - DevGarbage: 60.692800 ms for realtime GC Line 19820: 13:47:07 - DevGarbage: 68.735400 ms for realtime GC HDD plus ramdisk Line 6645: 14:15:57 - DevGarbage: 70.825400 ms for realtime GC Line 6914: 14:16:14 - DevGarbage: 58.260500 ms for realtime GC Line 9429: 14:17:16 - DevGarbage: 71.052600 ms for realtime GC Line 11313: 14:17:20 - DevGarbage: 60.997500 ms for realtime GC Line 11748: 14:17:20 - DevGarbage: 59.838800 ms for realtime GC Line 17087: 14:19:20 - DevGarbage: 64.772400 ms for realtime GC NVMe drive without ramdisk Line 6645: 13:54:22 - DevGarbage: 72.887600 ms for realtime GC Line 6905: 13:54:30 - DevGarbage: 59.166000 ms for realtime GC Line 9476: 13:55:38 - DevGarbage: 72.855800 ms for realtime GC Line 11541: 13:55:43 - DevGarbage: 61.931400 ms for realtime GC Line 11988: 13:55:43 - DevGarbage: 60.429500 ms for realtime GC Line 19900: 13:57:43 - DevGarbage: 67.482200 ms for realtime GC NVMe plus ramdisk Line 6645: 14:02:13 - DevGarbage: 70.539800 ms for realtime GC Line 6904: 14:02:19 - DevGarbage: 58.487600 ms for realtime GC Line 9232: 14:03:09 - DevGarbage: 70.705000 ms for realtime GC Line 11278: 14:03:11 - DevGarbage: 60.021700 ms for realtime GC Line 11713: 14:03:11 - DevGarbage: 59.999700 ms for realtime GC Line 18967: 14:05:12 - DevGarbage: 67.676100 ms for realtime GC Relevant specs: i9 9900k @ 4.9Ghz, 2x8GB DDR4 3800-CL16, Samsung 970 EVO Plus 500GB NVMe drive, a "desperately slow" 5400rpm Seagate 2TB HDD and Windows LTSC 2019 (version 1809). These values reflect that GC is done in RAM, so all configurations perform mostly the same within a small margin of error. Notably, GC times are not worsened by using the RAM disk despite the higher amount of reads/writes to RAM. As for HDD/SSD access, the game just crams everything into the AssetCache folder ("everything" being mostly other players' models and cars) until either: - The game is closed and contents of AssetCache are cleared. - Disk space runs out, causing the game to crash. That means that there is no GC being done periodically, unlike with main RAM, as the game simply assumes that storage space won't run out. GC times aside, even though I did not measure framerates, the "HDD without ramdisk" setup created some very noticeable irregular stutters whenever characters loaded in, plus it also took longer for characters to load. Using a RAM disk greatly improved framerate stability even when the game's install remained on the spinning HDD. The performance difference between a high-end SSD and a RAM drive is minimal to nonexistent, but using a RAM disk will also reduce wear on SSDs and prolong their usable life. So therefore I strongly recommend this tweak to anyone using a spinning HDD or a pre-2014 SSD who has enough RAM at their disposal. The RAM drive can be set as dynamically allocated so memory is freed up again once the game closes. Of course, this will not be necessary when the new engine arrives as it will be able to use as much RAM as it needs, being a 64 bit executable.

If you can't counter the weapon by equipping it yourself (as a criminal) then it is fundamentally unfair, regardless of whether it's just "a little" overpowered or "extremely" overpowered. No other secondary comes close to having the same damage per shot so it's easy to see why this is an issue with the PIG in particular. Thus criminals should have access to a reskinned version, perhaps based on concealed illegal stun weapons, as it makes little sense to introduce a secondary with 90%+ health damage. In my opinion fact that it stuns rather than kills keeps it somewhat balanced aside from the fact that only one faction can equip it.

I agree and in fact I believe ramp/dropoff range is still too low on the CSG, as Rifling variants will outperform standard ones by quite a lot at 15m, and since all CSGs come pre-modded it will feel a bit like a cash grab for those who already have a non-Rifling CSG. Of course with the JG there are similar issues, but these can be solved as the JG is available with open slots for free, so just sticking IR3 in it will "fix" the issue. The numbers were probably designed with IR3 in mind. I would leave ramp start at 20m (at that point spread will be the limiting factor anyway) because the extra 5m won't really help Rifling variants due to spread, but it will help non-IR ones close the gap. After all IR3 has no downsides on the CSG, and most of the other shotguns. Other than that small detail the changes look great, looking forward to test the CSG in Asylum soon.

Does this mean that there will be no changes to live/prototype weapon stats or will they be outlined in a separate post?

I agree about this, the PIG is probably the most annoying gun to play against overall, and it really doesn't help that it is locked to one faction. Even if all other LTLs may need a small buff, the PIG really needs either a damage decrease or a wind up timer like the Ogre. At the very least the dreaded Perc combo should be prevented. As for the Scout I think many of its issues stem from the fact that a 3-slot version is unobtainable and it's always a choice between the 3 mods that would go well on it. If a 3-slot Scout or Pathfinder existed I have no doubt that it would be more than viable.

Not only that, but it reduced the number of situations where one team could find itself without any options whatsoever to come back and win the mission, especially on Item Hold. It is not good game design if the mission keeps going on with one team already guaranteed to win. Therefore overtime was one of the better changes APB received over its lifetime in my opinion. I also like the removal of P5/N5 since, besides unfair, it was way too situational for the "it helps new/struggling players" argument to make any sense. I have had missions which I have both won and lost due to the bounty system in the past (fortunately very few) and the negatives always outweighed the positives. Also, it made factions unbalanced as Criminals were able to prevent bounties by paying off, whereas Enforcers had to always be subjected to them. All in all, I consider this one of the best things LO has done for this game despite a rough start with weapon balancing, and it makes me look forward to more changes of this style in the future.

Test districts were populated after the patch so it was possible to get a decent picture of how the changes work in practice. The JG, Showstopper and Shredder seemed alright. Thumper looks like it is also in a good spot but I do not currently have one. However the NFAS, Strife and CSG are rather weak compared to the above. While NFAS is a monster on live, this is mostly due to per ray scaling, so the huge damage nerf compared to 2017 stats was not really necessary. I would give it the Thumper's 444 damage value, which would still leave it below the (certainly not meta) pre-LO NFAS. On current prototype stats what would have been easy kills with live stats sometimes require the whole magazine. However, 396 damage does seem appropriate for the Ogre provided nothing else changes on it. Strife is also very underwhelming, if a damage buff is not an option I could see its spread lowered to 85cm, as otherwise it is hard to justify picking it over other shotguns. And also, while not relevant to the balancing itself, I would fix the typo in the description ("piece" instead of "peice"). CSG is the most frustrating one, as I looked forward to it becoming less underpowered compared to other shotguns as it is now. The removal of center pellet inaccuracy allows for jump shooting, but this feels gimmicky on a weapon with good mobility to begin with, and the rest of its stats are a downgrade from 2017. I don't see a solution besides increasing its damage to a maximum of 630 or even 651 by slightly upping per pellet damage, to further compensate its firerate reduction compared to pre-LO.

I agree with most of the changes from the 2nd table, with two exceptions: - The DOW's damage should perhaps be lowered slightly as it already has lower spread than the NFAS, which is enough of a tradeoff for not being automatic in my opinion, it does not need to do substantially higher damage than its automatic counterpart. - The JG's firerate seems too good. Even with its spread the high amount of overdamage goes a long way to render the CSG's accuracy advantages moot, so there is no point in the JG having a substantially better TTK. Something like 0.72-0.73 would make more sense. Keep in mind that reducing its "range" as in dropoff start will accomplish next to nothing, thanks to IR3 (see below). As for IR3 it won't make a big difference on CSG if the dropoff starts at 20m as the table says, since practically all shotguns very quickly become ineffective at that distance due to pellet spread. In case of the JG it will be a factor, but since players are free to mod it as they wish unlike the CSG, I believe it will be much less of an issue. Finally, I expected prototype Asylum to replace standard Asylum rather than exist beside it, so sadly the districts have been empty most of the time. I realize taking away "standard" Asylum districts might upset a small number of players, but such drastic measures might be necessary to gather enough data.

I fully agree with the idea that carbine has RNG issues, and more recovery could certainly help since it often has to compete with weapons that do not suffer from bloom at all (or almost at all) such as PMG or OSCAR. I don't see the point of reducing its movement penalties but more accuracy in general is just what it needs right now. However, I disagree with the idea suggestion making the NFAS 4 shots to kill, before pellet spread was introduced it was much harder to use despite its higher maximum damage, and it was far from meta. I also can't really suggest anything since it's not clear yet how the NFAS will behave once current shotgun changes hit live. The biggest issue with its current iteration is the 0.84 per ray scaling (in combination with 381 max damage) which is far too low for a weapon which can shoot such a large amount of pellets in a short time. Perhaps other changes to the very least used weapons would also come in handy, such as the Harbinger getting bumped up to 300-320 damage due to Clotting Agent prevalence.