I was thinking some new models (skins) for supporters / vip would be a good idea
As the ones we have currently are basic ones
1 Like
+1. We need moar!
Some are really nice, like the Killzone caracter: http://gamebanana.com/skins/141104
Only 4 Mo.
Also, whereās the new VIP godās skin you promise Morellā¦ 
I know itās not the same version, but as an idea we could have this skin as prize for the mako v6 winners when it comes out.
http://gamebanana.com/skins/141515
Thereās a bunch of them on Gamebanana, and probably more elsewhere
http://gamebanana.com/skins/cats/8008
Maybe summer thematic for this summer, isnāt bad.
Humans in swimsuit and zombies with sombreros?
There were talks to add a new playermodel for āDIOSā level vips, but aside from that you have to keep into account that Mapeadores has a lot of custom content which players have to download. Every time something is added, regulars donāt really notice it since they have the rest already downloaded, but for a new player the amount of time it takes to join might be a turn off. This means that it could be driving new people away without us noticing, which is not a good idea in the long run.
Still, though, the playermodel for DIOS makes sense so it will probably done. Modelling takes time and Morell has been busy lately, working on other stuff for the servers when he was not. I think we will have plenty of fun and updates soon enough.
Me alegra leer esto Envi.
Yeah, thatās why i give the size of the model i proposed jajajaja. I can minimise size of some existing models, by decimating verticesā¦ Itās easy and fast to do.
I connected last time on a czech server, with the fastdowload not activatedā¦ I succeed to play after 45 mins of waiting, even if i have a 500Mbps connectionā¦ 
But maybe we can just swap progressively some old models for new ones?
Firstly, vertex data is not the main driving factor in filesize for most playermodels; textures are. Although reducing the number of vertices is good (Mapeadores has always used models with low vertex/poly counts to avoid client FPS lags when there are large crowds in small spaces), it wouldnāt have that much of an impact. This would be unless the model was vertex heavy, in which case it could either be modelled wrong, look weird when simplified or overly expensive; all these situations would not be a strong case for such a model being added.
For reference, I believe the default CS:GO playermodels have around 15K vertices, while the ones currently used on the servers are between 3K and 7K. Filesizes for the marines (the default human playermodels) amount to 6.6MB from textures and 1.45MB from model data (which includes vertices, skeleton, animations, UV mapping, etc.)
A progressive swapping has to deal with the fact that Mapeadoresā models have a degree of relatedness to them. In particular, all human skins are Half Life related (from either Black Mesa or HL2) so they look similar to one another, have low vertex counts and have a decent enough appearance (with shaders tweaked for CS:GO by Morell and I). Any change done here should ideally follow these rules, and shouldnāt really break the coherence in style that the models have to avoid looking out of place.
In zombie skins, there is a bit of variation (which comes from the fact that fewer Half Life related models would fit in CS:GO and from the crash issues models rigged with the old skeleton used to have), but the theme is more or less kept throughout. Things are a bit more thought out that they might seem to be from outside, and I think it would make sense to maintain those standards to make sure that:
- People with slow connections can still get to discover the server.
- People with old PCs get to play without unbearable clientside lag.
- The models have a theme so that nothing feels too out of place.
Of course, trying to follow some policy when dealing with stuff that doesnāt have to follow any pattern means things get harder. Iām not sure if the effort is worth it, but I think it is reasonable because it gives the server(s) (the same principles apply to the ZS/Prey server) a sense of being a āgame on its ownā.
1 Like
In another hand, guys have military gears and when they turn zombies, they get civilian clothes (or a short for vip class -_-)
We get a lot of remarks about ābeing supporter? Itās kinda uselessā. I do it to financially support Mapeadores, but alot of players just donāt care about thisā¦ Facts:
-Only 5% of supporters know they have reserved slot, cause you need to use the connect command in console.
-Only 5% of supporters are using tracers.
The cool fact is to being recognized easily. But 80% of colors are useless (same as others or totally ugly, you can remove them), and all the current models are just totally boring when you played CS. So a lot of people just considers that being supporter is useless.
The new supporter tag just donāt make any improvement about all of this, and is a total waste of time as given now (without giving the ability to players to choose their own when they want, but that needs to code a new plugin).
Same for DIOS Vip. After spending 1000 hours on the server, you get rewarded with a green color for the usual skinā¦ And thatās all. Iām still not yet DIOS, even if i was nolife cause of ZE, with my job/studies.
Ouch. 1000 fucking hours! You need to spend 3 hours each day during a total year to get it! 95% of players just donāt go up to DIOS, even to Diamenteā¦
And many of them give you the same amount of money than a new AAA game being supporter during this time. Logic have to get even just a reserved slot. Even if for myself i consider this a too much drawback for supporters, only if i consider this actual facts.
Iām sure that more players will try to reach DIOS status with changes and will spend more money being supporter during this time, if you make this changes. During this āi want to have all for freeā douchebag days, all of this is a financial miss for you.
[COLOR=#ff0000]
People with slow connections can still get to discover the server.
[/COLOR]
My opinion is that download time is the only one true argument. Now consider the true facts:
-downloading Wanderers, Rizomata and Venice Escape take more than 200 MB to download, i talk bzipped!
-you can download 100 models like the Helgast Rifleman (bzipped too) for the same amount of dataā¦ And you download them ONCE until you totally format your computer.
-Average connection speed in EU is 10 Mbpsā¦ downloading this take less than 5 mins.
Youāre disadvantaging 95% of players cause of 5% of other ones, even with a 2 Mbps connection speed youāre waiting 25 mins yeah, but you are used to ā¦
Last time i waited 45 mins before playing on another server, but that was due to no FASTDL and a shitty upload speed from the server (150 kbps). You know what? I waitedā¦ And iām not used to wait that much, with my 500 Mbps connectionā¦ I wasnāt encouraged to wait cause i will discover a new mod or a new map. I just did another thing during the download time, as anyone does when this is such longā¦ And i finally connected and played.
[COLOR=#ff0000]
People with old PCs get to play without unbearable clientside lag.
[/COLOR]
I get 300 fps on my server, when i play on the same machine and i host 25 players. I get 60 fps max (average 45 fps near 90% of time) on your server, with a Radeon R290X and an Intel Skylake I7 lol. Really, the fact is that 95% of the total lag is due to the server prediction, server upload speed, the distance between the server and the NRA of the client and distance between NRA of the client and the client, and the client prediction, that have nothing to do with model complexity.
Even if we got many dynamic lights in the same cluster vis? Just reduce the graphic settings and thatās doneā¦ CSGO is processor related more than anything else.
Only 5% of total lag is related to the graphic rendering no matter what computer you have. Models have nothing to do within it.
[COLOR=#ff0000]
The models have a theme so that nothing feels too out of place.
[/COLOR]
Yeah, Master-chief is a total mess in a ZE universe. But we are not requesting masterchief or batman modelsā¦ But not that fucking usual CS models we currently have. Even get old school HL models can enjoy us without visual optimisation. And there is alot of models that fit well (and even better) to ZE universe (Helgast Rifleman is one of many exemple).
I just see alot of good players leaving Mapeadores, coming just for 2 or 3 particular maps, or near to definitaly no more connect here, if they find a well configured EU ZE server without this total misses, that have for me not solid arguments.
As a last shot, when i see Mapeadores CSS compilation videos available here, iām just thinking āhow ZE is now such boring and tasteless excepts the gameplay, compared to those timesāā¦ (taking consideration of the difficulty to mod for CSGO or porting)
Iām sorry to be such harsh. But i really think youāre doing a bad mistake not considering this, and that makes me sad. And iām one of the few ones that have enough courage to tell you all of that, you could be surprised if you discover how many players are thinking the same things, even with a shitty computer and internet connection.
1 Like
No i think you must change the defullt of players models itās (sorry morell) Everyone hated it i gues this is good
Hi Wan,
I would have to agree with Envi with this one. Unless model fits with the thematic profile of all of other models on rotation and fits within the guidelines of size, vertex count etc.
A majority of players are one time only or short term players. Long term repeat players are rare. The server arguably has to reward returning players which it does but atleast make experience enjoyable for all players. This includes visual disabilities such as visual impairments, varying level of physical hardware of players and the least amount of waiting time when connecting. If i spent more than 15min downloading i would exit out also.
Regarding the reward for diosā¦ you shouldnāt have an end goal reward to stand out as your only incentive to play. Investing time into playing to achieve those rewards is because you physically enjoy playing zombie escape inside the community.
If your opinion is swayed by rose tinted glasses of past css on mapeadores then i believe you need to take your head out of the sand. The maps on rotation have far more mechanics, visually and aesthetically sound and are far more immersive than the general standard maps released back then.
As Envi highlighted they did not rule it out. A dios model requires a time investment that neither Morell or Envi have at the moment.
- Unorth
1 Like
Iām going to address your points in the order that makes more sense, I think. First, even if people have an average 10Mb download speed in Europe, that doesnāt tell you anything about the common range in which people fall and doesnāt speak about specific countries. I did a quick check for Spain[/URL], France and [URL=āhttp://testmy.net/country/ptā]Portugal (just for the shake of proximity) with very different averages per country (a 6x difference Fr vs Pt). Drilling down further, you can see disparities even within a country, with certain areas having significantly lower download speeds. When taking things into consideration, one should start from a reasonable estimate for the worst case scenario. With those numbers at hand, I think we can take your 2Mb estimate and, in that case, filesize does matter.
Of course, if filesize does matter, what the hell is going on with maps? Well, for starters, both Rizomata and Wanderers are maps that should be about 1/3 or 1/2 the filesize they have. I donāt know why you mentioned Venice in particular, since itās a 50Mb map, and sits on a reasonable filesize. This filesize is also bound to shrink when compressed, except for the music. However, you need to understand the psychology of people. In older games, youād see a progress bar showing you how much of the download list has been processed. This doesnāt happen in CS:GO, so users only get feedback for how complete the download is. Because of that, a map download is less annoying to a player, because the download % will be shown, with something moving. For models, sounds and other assets in general, you will see bursts of names with no information whatsoever of how much time is left. I commend your patience for waiting for 45 minutes on that particular server, but when you can see people requesting Valve to let clients restrict downloads and bitching about player models on Reddit, you know other people have different opinions.
The thing here is not the amount of data you download, but rather, when you download it. To a player that has never joined a ZE server, a 15 minute download might seem overly long. After all, if that player is accustomed to most gamemodes, ZE download times will invariably seem ludricous. You want to have new players joining, because when player engagement stagnates servers invariably die. As much as one would wish for community ties to remain strong, the fact is that people have other things in their lives, work, friends, studies and so forth. The best way for a server to carry on is to be as āresilientā as possible in terms of people coming and going. Thatās not to say that the server isnāt cattered to regulars: from requests to add certain things to config suggestions, most of the stuff on the server right now has been curated by the community. Itās just that the objective is to have a community that grows and varies over time.
In that sense, I donāt understand your complaint about the new custom tag system. Mapeadores has usually few problems within the server because it is set up so that problems canāt easily arise. We decided in favor of supporter tags being decided by the community to:
- Avoid having to police tags by having people use insults or slurs (notice that we usually donāt police those ingame unless it goes out of hand, but a player using them as a tag would probably be annoying to the insulted person).
- Allow for tags for event winners, unlockable results, etc.
- Bring the community together when suggesting and adding new tags.
Also, just to point it out, it wouldnāt be hard to add custom tags to the current system. Morell was on the line but we thought this approach would cause less problems while being equally satisfying. We have been talking of another kind of text-based customization which perhaps could be fully player-controlled.
This takes us to the pointlessness of Supporter. First of all, supporter is meant to be financial aid to the server only. All the perks are visual and added only to give some extra flavour to it. However, I know that people always see these things in terms of āwhat am I getting out of my moneyā, which is perfectly reasonable. Getting stock CS players was a suggestion players themselves made when the server opened. This, alongside with coloring them, was an easy solution to dealing with the skeleton-breaking updates and the fact that supporters wanted something to distinguish themselves from the crowd. The part about the skeletons is important, because one of the reasons all of Mapeadores playermodels are rigged by either Morell or Kaem is to ensure that the skeletons rigs are updated.
Performance is what matters there. You see, as much as Iād like to agree that 95% of performance (or even 50% for the matter) has nothing to do with rendering, that doesnāt really match my experience seeing how thatād mean FPS differences on a map per map basis are negligible (they arenāt). At this point you have to consider the many possible setups people may have. Yes, client-server communication takes a good slice of the performance-cake, but for people with an old CPU or an old GPU, performance will be bounded differently. Server-side performance is also obviously important, and outdated models affect it greatly (check the one and only comment from 10months ago in the model youāre staunchly proposing). When the playermodels on the server were updated after the last skeleton swap by Valve, server response times and variance among them in high activity holds decreased. In fact, with Syou doing subtle changes when no one is looking, serverside performance has increased (by either removing issues caused by mappers on their own maps or by tweaking certain cvars to remove stuff that isnāt needed on ZE, like blood splatter prediction).
This means that, in order to keep things smooth, all playermodels would need to be routinely updated when Valve decides to go bananas. So custom playermodels add an extra layer of work to be done, with the models currently on the server already taking a toll (protip: the basic human models use the same model with the same skeleton and skins for a reason; it makes the fixing process faster). Now, if the focus is getting money for the server, it makes sense to go out of our way to create more playermodels. But since the server isnāt āprofitableā by any stretch of the imagination, community-building makes more sense. Thatās why a playermodel for vip Dios was being considered at the time (and considered still, itās in the queue after the !tag system). If we were to start adding models for supporter, thatād be incentivizing donating at the expense of new players being able to join and the server being as smooth as possible as quick as possible when Valve pulls an outdate. A Vip Dios model, as something that anyone can try to go for, serves as an engagement loop. Some players that try to get there will probably chip in and contribute financially to the server at some point, hopefully.
Finally, if people are leaving Mapeadores because there are no custom playermodels, Iād think there are more serious issues at hand, probably config. related. If they are leaving for servers that break Valveās server guidelines (using knife and weapon skins plugins), then there is nothing we can do about it because we are not going to get into that attitude (even with how poor Valveās handling of community servers has been). If it is config related, we should be able to fine tune things without losing the serverās identity. With respect to CS:S, however, Iām not sure how thatād go: Mapeadores in CS:S used only one playermodel for humans (a model that Morell ported, coincidently, and decided against to have more models) and the setup was very much alike. If the problem is that CS:S had more interesting maps, I (as always) invite everybody to start mapping and contributing and Iāll be happy to help with any doubt. If the problem is that CS:S had a harder config, Iād probably have to disagree (though Iād admit that the molotovs are a game-changer which might need to be controlled further). If the problem is that CS:S had more playermodels in general, Iād agree, but I donāt see how that makes the mod more enjoyable.
As always, we are open to suggestions (just as people should be open to counter arguments). If everything goes smoothly, a new playermodel should be done somewhat soon, but that always depends on what is needed.
EDIT: BANG BANG WALL OF TEXT
1 Like
Thanks for the wall
This is the only way to give real opinions. And problem here is that the benefits and counter-backs are depending on how precise our estimation is about downloading volume and time, and what can accept players. Iām aware about near all facts you gave.
More than use the speed testing websites (i have concrete argument about why), you can read the Akamai Q1 2016 report, page 59 for Europe countries. (see annex 1 for details). Not all people are as patient as me, true. I think the psychological accepted download time is 5 mins for anyone.
Combined with this 4 Mbps target (95% to 98% of players as Akamai reports), that correspond to 1.2 Gygabits, or 150 Mbytes of download size. Taking alot of considerations in the worst scenario (see annex 2) that lefts approximately 55 Mbytes for models. 22 models like the Rifleman. 8 like the nanosuit like models (HD model with 3 different skins). This worst scenario still respect the server philosophy.
The real problem is about optimised and easily maintainable models. But valve will never more make a major update like the September update, if you know what are the historical reasons (competitive complains: animation glitches, shitty hitboxes during some animations too). And the overload is not related to custom models, even classic servers was as much impacted (see annex 3).
The rifleman perfectly works on my server, like 3 other models i already installed. This models had no impact on performances, and need no port, work or optimisation. Ever for mates that have really old CPUs and GPUs. Yeah, i know i donāt have 60 players on it, but 25 player shows already what impact you can estimate on a fully server.
About custom tags, i already gave my opinion and what are the best ways to integrate it, with the problems that adds. The only problem i found is thatās itās worth it as itās now. I was talking about this cause it seems for me thatās new models is the best return on investment in any aspect of the server and the community (not only financial, but also for the community and for the players, even the totally new ones discovering ZE).
Annexes:
(1)
Percentage of connection above 8 Mbps:
France - 81%.
Spain - 90%
Portugal - 89%
A lot of countries like belgium, uk, norway, etc (and you have alot of players from here too) get a 95% rate.
And weāre talking about above 8 Mbpsā¦ Above 4 Mbps is already targetting more than 95% of the players or near 98%.
And itās drastically growing fast this last times!
(2)
The download volume for a totally new player needs to include map file, sounds, overlays, models and effects. Considering the guy have no chance and start with a 80 Mbytes (once bzipped) map, and 15 Mbytes of anything else than the map and models for volume size.
(3)
The overload of this update was really related to:
-hitbox calculation, as spherical algorithm just give a huge overload compared to cube one, for server and client prediction.
-inverse kinematic calculation now used for the weapons, as it needs to use geometrical matrix solving systems.
-the new animation networking (now continually synchronizing animation state instead of periodically latching).
i guess about what you and Envi are doing, is vis cluster optimisation for maps, and as you said cvars tweaking for reducing prediction load not required.
To chip in a bit from someone who is currently (and actively) maintaining a bit the back end of the server, mapping and doing whatever else I canā¦
Thereās a reason why I tell active mappers to map less of a mess. Itās really not that hard. I donāt understand how one can manage to compile a map so absurdly large and NOT compile in full in the end making their map look ugly. This is all the mapper at fault. Venice is 30mb (rounded up) compressed while Wanderers is 99mb (rounded) and Rizomata at 79mb (rounded). I donāt see your argument regarding venice. The other two are fairly big for what they really offer in the end. Perhaps they didnāt compress the music to a fair amount? Still, it ends up being the mapper at fault.
Having joined oversea servers Iād rather not download so many models. I have an SSD and I have to navigate and deal with a massive clusterfuck of crap in Hammer as a mapper. Most GO players seemingly install on their SSD in hopes they can get the most out of performance for this game. You need to be wary of it and tread carefully as Enviolinador posted above. Not many enjoy custom models as much as you and I do unfortunately.
Not taking in factor of how the game handles these files after they have downloaded. Which can take a few seconds more to unzip, place or rearrange files. Again it all wildly depends. A simple 11kb .nav file can be downloaded instantly and take around 1~6 seconds to set (from your ādownloadsā cache folder to setting it into your āmapsā folder) that I usually see.
Iād have been pissed and chewed out whoever was managing the server. I donāt think I have the patience for that.
I doubt it. To help me find the issue at hand if itās coming from the server please give more details than saying āI get less fps in only this server!ā because really, most of the complaints I read regarding it is mostly bullshit thanks to sites like Reddit. Please give me the map name, amount of players, average fps you think youāre getting and running āperfvisualbenchmarkā in console with all the message prints it gives you. It will lock your mouse for a minute or two before it finishes. After that paste it to pastebin.com or something of the likes and PM or send it on Steam to me, Iāll try to take it from there. From what people tell me 40~60 is going to be an average for a full 64 slot server. Itās just the way Source is going to handle it. Thatās the way the game is. Itās handling it damn well too compared to older Source titles. Another factor is background processes or people mindlessly running Skype, IRC, AV or any browser with media playing in the background and it can all add up.
If you think gun fire dynamic lights are causing lag you can type āmuzzleflash_light 0ā in console.
Not as much as you think anymore. It still is but not as significant as older Source titles.
My computer would disagree with that percentage.
Itās been happening since large maps in CS:S have been released. Itās quite normal.
Some configuration for maps are being changed from time to time. If you have suggestions please do let me know through Steam or PM. Such as ze_arctic_escape or ze_alien_mountain_escape have more zombies spawning and can catch up easily to CTās if theyāre not shooting leading to less boring rounds.
The polycount for this model[/URL] is almost the same as default CT/T models. To compare, the model [URL=āhttp://i.imgur.com/HWCmX4m.pngā]in the server is far less significant. The impact can be minor to severe depending on the players setup. Also to point out and I could be wrong here is the current released model for this is using the older outdated skeleton which can lead to floating weapons, weapon switching looking awkward at times and in general if we were to accept this model it would still have to be recompiled in the end to be rigged for the newer skeleton. Weāre trying to avoid high poly excessive models. What we also look into are models that are tintable in color as well with our setup. This model is straight black and would look like crud in any color a player might have.
I hope this sheds some bit of light that ānothing is happeningā towards the server. Some little things are coming to hopefully gain a bit of performance such as reduced death effects and the likes. It should come āsoonā. No exact date for when. As Enviolinador said, we are open to suggestions but you have to be clear.
1 Like
Venice is not an argument in my explanation. I get the pattern of the last 3 maps you put in the server to get closer to reality. I think itās a good pattern, as average file size of the Mapeadores maplist is 59.77 MB. I take in account maps that you removed, so now it should be higher, as newer maps tends to be more and more heavier, cause there are more and more huge and fullfilled with custom materials.
For sure, Wanderer get 77.3 MB of musicā¦ With 15 Mb for Bahamutās one compressed in MP3 VBR 224 kbpsā¦ The music is 9 min long for a fight that take 3 minsā¦ LOL. Near all other musics are perfectly encoded (VBR 128 kbps, and itās the reasonnable limit to donāt get the music noisy, cause of their choices) and their duration are ok. Thatās a question about what the mapper want: music or not music? As they really adds to the atmosphere, but so many players use the !stopmusic commandā¦ There is a mid way (add music only on relevant situations).
For Rizomata, 86.5 MB is custom materials. The guy who made the port did no replacement of CSS textures to CSGO onesā¦ [ahemā¦]
Some other like Frost, are pretty well optimised (i think, i need your advice). The problem then is the source engine and his brush count limitation, that requires Luff to abuse modelsā¦ And then create a huge amount of VHV filesā¦
Also CSGO vanilla textures, are useless when you want to create another atmosphere than an urban one.
Half of the size of Frost is due to the VHV and custom textures.
Yeah i know. You have to consider writing cycles on the SSD clusters. In fact, as you download them once itās okā¦ The game just read the files, or process them in ram.
The tip is to have 2 folders and switch the name for the mappers.
Itās painful if you do it manually, but you can create a batch script to do it easily.
As it, i donāt pollude my materials and models browser in Hammer, to try to mostly use CSGO vanilla stuff.
The engine first download the map and decompress the whole bsp in ram, then store the files in the hard drive. So your nav file example is irrelevant.
Takin your consideration with a relevant file, just benchmark the decompression of zh3fix3.mdl.bz2 (models\player\mapeadores\morell\zh). Itās totally irrelevant, even with a low end and old processor, that took under 500 ms or even fewerā¦ (if someone with a shitty proc can benchmark it?).
Thatās another thing with huge size filesā¦ And thatās why for sure, we canāt allow to get heavy filelsize models.
Thatās a really great idea!!! I will dot it! With different population size on server (i hope to get 30 and 60 players benchmarks). I will send you videos of the benchmark, to get the whole context (like if there was massive rekt with alot of mollys for the worst case).
I donāt have lag. Or iām the least player to suffer it, thanks to high end computer and connection. My client lag is under 7 ms, and i get 90 ms ping on Mapea (when you switched to the new server, i was more between 45 ms. Weirdā¦)
The chokes are for me the major problem. We compared with some friends with a low connection speed, and we have exactly the same amount of choke (currently an 50% average). Maybe this could be one of the major cause of the lag complains?
Itās your feeling with your own computer.
I started ZE with a Radeon HD 6750 (mid-end GC released in 2011) and a Core 2 Duo E7300 (mid-end proc released in 2008).
I switched my GC to the Radeon R290X. One month after that, i switched my proc to an i3 -4150, and bought an SSD. 3 month after i switched to a i7-4790.
So i get a view of what ZE is with a low, mid, and high computer.
The performance boost is real in solo Source Engine games and competitive CSGO (only 10 players) when i changed my graphic card. The boost was low on full ZE servers. When i changed the proc (twice), the boost was irrelevant on solo and competitive, but was here with full ZE servers. Thatās why iām still conviced of that (as i talk about 5% of perf in the case of a full ZE server).
Yeah, was a bad argument from me. As the custom tag. Sorry for that.
The fact is more than players are getting bored with maps, and then try to find servers that are currently running which one they love.
For me itās perfect! In general, all players (or 99.5%) are thinking that Mapeadores have the best config of all ZE server. The only point where rare players can disagree is the zombie spawn process (spawn in middle of the crowd, when some servers TP them to the starting point). But for me, the Mapeadores one is the best way, as mappers can handle the infect grow.
I really donāt know if itās new skeleton based. I need to learn more about rigging, and need to decompile it. Maybe you can try? But i never saw bad animation on it. As for the others i have. I understood what the problem are, and itās true that at the moment, models that fit your requirements are really rare. I think the polycount have less drawback than you think (within a reasonnable polycountā¦)
About the server models:
-Where are they stored? I want to decompile them if you agree, to learn more about all of that. -i have a bad walking animation problem with them, but only on myself, using thirdp. (iām using the FBI one). Really weird.
I try to be clear and the most accurate possible. You know how itās hard.
And iām and will ever be impressed by the work and the knowledge of the Mapeadores team, really, thanks for all you are already doing!
TL : DR pls?
2 much writing 4 me