Jump to content

March Cauldron Leaks Discussion


68 replies to this topic

#41 Ekson Valdez

    Volunteer Moderator

  • PipPipPipPipPipPipPip
  • 705 posts

Posted 10 March 2024 - 10:38 PM



@everyone,
please keep the aggression low and the discussion civil and constructive,
or else we will be forced to hand out some coolshots.

Thank you.



Edited by Ekson Valdez, 10 March 2024 - 10:38 PM.


#42 Ttly

    Member

  • PipPipPip
  • 80 posts

Posted 11 March 2024 - 01:24 AM

View PostLordNothing, on 10 March 2024 - 05:32 PM, said:

i wonder how many people on this board has even looked at game engine code let alone made changes to it or took on the task of writing their own. in my experience very few gamer's look behind the curtain. most of them end up parroting the usual excuses to why we cannot have x. whats really going on is the cauldron are doing the usual mix and match that all modders end up doing in every game. they got a fixed set of mechanics they can choose from and can put them together like lego, and change some variables and some effects and you get a new weapon. these are simply the tools they have available to them. sometimes the pieces don't fit and a possibility becomes a dead end. sometimes they do and you end up with something awesome. unless you can get into the underlying code you cannot add, modify or even fix the deficiencies in the existing building blocks. you cant even fix long standing engine bugs (and is that a rented blob or do they have the source). unless you actually sit down and do the modwork, how do you know what you can do?


Actually I have seen game engine codes.
On some open source indie game, think of Roblox or whatever, yeah not exactly CryEngine sure.
It really *is* like a block game where you can "mix and match" mechanics in most cases.

I don't have reason to believe that it's that "extremely complicated" especially since it (MWO) is a live-service game that should've been from the ground up made to have new content patched in routinely.
Most of the new weapons also exhibits these "mix and match" features, the plasma cannon takes the heat damage component of the flamer into a PPC projectile, beam laser is just a flamer with no heat damage, more damage, and more range, etc.

It's not like the discussion was something more drastic as "please change how rendering distance/LoD works so people can't see through objects on low enviroment settings and preaim at people like they're wallhacking" or "make post-processing look less bad", "fix loading times", etc.

View PostCurccu, on 10 March 2024 - 10:03 PM, said:

Everything Quicksilver wrote made total sense if you have been reading these forums/cauldron discord for a while. Known issues that require coding and another known issue is that this game doesn't get almost any coding time unless it is critical for this game to run.


Uhh, no?
Stating "Current code doesn't allow x mechanic in y weapon and x mechanic doesn't belong in y weapon category despite mechanics existing for it" is just absurd and only later to pull some dev statement screenshot that states that the reason actually being "didn't want it to, and there's no ammo category for it" is just silly, just say that in the first place then.
And neither of us are psychics, no one is, I can't read minds mind you.

Alright, maybe it's my fault for not being around as long enough to know state of the dev team, but having dev statements to strictly be on some public Discord channel to be buried by 100 other messages to dig through is really beyond the apathy barrier for me personally as a personal failing sure.

#43 ScrapIron Prime

    Member

  • PipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 4,880 posts
  • LocationSmack dab in the middle of Ohio

Posted 11 March 2024 - 05:21 AM

View PostTtly, on 11 March 2024 - 01:24 AM, said:


Actually I have seen game engine codes.
On some open source indie game, think of Roblox or whatever, yeah not exactly CryEngine sure.
It really *is* like a block game where you can "mix and match" mechanics in most cases.


You would think, right? But no. There's a reason CryEngine was so reviled back in the day and why its sales team went so hard after various game designers, lobbying ANYONE to adopt their core code and keep them in business. CryEngine sucks.

And here in MWO, the dev team never worked out a resource-practical way to animate melee, arm movement, squatting, or how to change from one ammo type to another with the same weapon. CryEngine likes to code you into a corner.

But I digress. If you really think its that easy to do, then head on over to that Discord server, sweet talk some folks into giving you a block or three of code, dust off your Lua coding skills, and show us all how its done!

#44 Ttly

    Member

  • PipPipPip
  • 80 posts

Posted 11 March 2024 - 06:08 AM

Alright sheesh fine, CryEngine sucks and if something hasn't been done about it within the last 11 years then it's not going to change anytime soon.

Anyway are the MASC acc/decel nerfs even warranted?
It's pretty much a blanket nerf that seems to just be after one mech, Executioner.
Ugh fine, there are other MASC Assaults too like Koloss, Spirit Bear, Juggernaut, the MASC Zeus, etc.
But the most seen MASC Assault is still the Exe regardless.

For the other weight brackets, on heavy the Hellfire has pretty good acc/decel already so it's not even affected that much.
I guess the Skoll, MASC Thunderbolt, and MASC Cataphract might feel noticably less agile.
Also the change to Mk1 MASC, light MASC being buffed to have as slow of a fill-rate as other class is a buff to Fleas and the accel/decel nerf doesn't really matter. Does it really need a buff in this regard?

+200% thrust to Highlander and +100% to Vindicator, the most interesting part of the patch, unsure how this would actually look in the game. Highlanders and Vindicators flying up at Spider velocity? Might be disappointing with the jump height.

Highlander gets its quirks consolidated into general quirks instead of weapon category quirks, less pigeonholed into certain builds I guess.

And why did VND-1R and 1X lost its PPC velocity quirk for laser duration? Seems counterintuitive to the +thrust% quirk change and unwarranted.
The 1AA buffs seems decent though.

-10% heat on large laser, cool, more reason for people to use it over the noobtrap that is ERLL.
-More laser duration on BLC: Made my eye roll a bit, why not just increase the weight to 10 ton and give it the same stats as 2 large lasers already? At least it'll be a good way to spend 1 energy hardpoint if your mech doesn't have much of those.
-LXPL +12% heat, -5% range, well it's going to be less obnoxious for sure.

-ASP on clan mechs, some guy already mentioned how some clan mechs (omnis) don't even have cockpit slots, not much real reason to use it just for TAG either

Edited by Ttly, 11 March 2024 - 06:57 AM.


#45 KursedVixen

    Member

  • PipPipPipPipPipPipPipPipPip
  • The Wolf
  • The Wolf
  • 3,240 posts
  • LocationLook at my Arctic Wolf. Closer... Closer...

Posted 11 March 2024 - 06:08 AM

have we Forgotten the plasma cannon is a long range FLAMETHROWER as it not only causes kinetic damage but also increases a mechs heat when hit?

#46 Ttly

    Member

  • PipPipPip
  • 80 posts

Posted 11 March 2024 - 06:14 AM

View PostKursedVixen, on 11 March 2024 - 06:08 AM, said:

have we Forgotten the plasma cannon is a long range FLAMETHROWER as it not only causes kinetic damage but also increases a mechs heat when hit?


The heat damage is just extra frills, it's very minimum that it might as well not exist as it is.
Also the DPH for Plasma is somehow even worse than cERPPCs, the former #1 worst DPH weapon in the game (not counting splash).
Only thing it has going is slightly higher DPS if you mount 2Plasma for the same weight thanks to the faster cooldown.

C-ERPPC (2slot 6ton)
DPS: 3 with splash, 2 without splash
DPH: 1.03 with splash, 0.69 without splash

Plasma Cannon: (2slot 3ton)
DPS: 1.85 with splash, 1.28 without splash
DPH: 0.93 with splash, 0.64 without splash

Thunderbolt missiles are an even worse joke, they seem to go for center mass with less spread than LRMs, notably heavier than LRMs, but less velocity than direct LRMs, and only 200 damage/ton ammo? Some people cope that you can just take +velocity quirks and enjoy its more pinpoint damage without as bad of a velocity but why use it instead of even LRMs or SSRMs then?

You can check it yourself here:
https://docs.google....f4zY/edit#gid=0

Edited by Ttly, 11 March 2024 - 06:47 AM.


#47 1453 R

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Bridesmaid
  • Bridesmaid
  • 5,549 posts

Posted 11 March 2024 - 07:54 AM

...DPH isn't the point of plasma cannons, though? They're a light, compact LPPC-equivalent for the Clan side, which was seriously hurting for lack of LPPC equivalent. PPCs have bad DPH/DPS, yes, because their advantages lie in other areas. If you want deeps, do autocannons.

#48 Duke Falcon

    Member

  • PipPipPipPipPipPipPip
  • Trinary Star Captain
  • Trinary Star Captain
  • 937 posts
  • LocationHungary

Posted 11 March 2024 - 08:42 AM

Anyone else feels that "continous" laser is like a clan X-pulse? Well, not exactly but kind of...

View PostTtly, on 11 March 2024 - 06:14 AM, said:


The heat damage is just extra frills, it's very minimum that it might as well not exist as it is.
Also the DPH for Plasma is somehow even worse than cERPPCs, the former #1 worst DPH weapon in the game (not counting splash).
Only thing it has going is slightly higher DPS if you mount 2Plasma for the same weight thanks to the faster cooldown.

C-ERPPC (2slot 6ton)
DPS: 3 with splash, 2 without splash
DPH: 1.03 with splash, 0.69 without splash

Plasma Cannon: (2slot 3ton)
DPS: 1.85 with splash, 1.28 without splash
DPH: 0.93 with splash, 0.64 without splash

Thunderbolt missiles are an even worse joke, they seem to go for center mass with less spread than LRMs, notably heavier than LRMs, but less velocity than direct LRMs, and only 200 damage/ton ammo? Some people cope that you can just take +velocity quirks and enjoy its more pinpoint damage without as bad of a velocity but why use it instead of even LRMs or SSRMs then?

You can check it yourself here:
https://docs.google....f4zY/edit#gid=0

The heat damage is negligible. Plasma is trully but a cLPPC. And as such it could\may work pretty well if boated.
Proto ACs not looks so much "persuasive" currently. I think MGs would remain preferred on most builds instead of PrACs.

Thunderbolts...
Well, ladies and gentlemen, there's one in every family, so to say...
Sorry, no pun intended!

#49 Ttly

    Member

  • PipPipPip
  • 80 posts

Posted 11 March 2024 - 09:06 AM

View Post1453 R, on 11 March 2024 - 07:54 AM, said:

...DPH isn't the point of plasma cannons, though? They're a light, compact LPPC-equivalent for the Clan side, which was seriously hurting for lack of LPPC equivalent. PPCs have bad DPH/DPS, yes, because their advantages lie in other areas. If you want deeps, do autocannons.


Yeah alright, fine, let's just see what's the DPS/DPH of LPPCs then, oh 1.83 DPS that doesn't have splash damage and 1.22 DPH. And not every mech has the hardpoint/tonnage for an autocannon.
Look, I wish plasma is good/decent but it's not even a good LPPC equivalent.

Let's just take I dunno, Mist Lynx-C for our hypothetical 2PC loadout. Should be good for it right? ECM, jump jets, a decently agile 25t light, some quirks that synergize with this even.
https://mwo.nav-alph...=65d6d0c4_MLX-C
MPLs+AMS ammo as placeholder for same slot and tonnage use.

How many DPS would we have? 4.12 with splash? But shouldn't matter because it's a hit and run/poptart weapon right.
DPH? 2.06
ATO? With ECM we have 4 (7) cDHS, let's go with this loadout for now.
Everytime we fire our 2PCs we'd be generating 12.6 heat, and with 3.15s cooldown we would've dissipated 6.93 of this heat by the time we can fire again for a heat gain of 5.67. So that's 8.9 ATO? Each of those alphas doing 13 damage each time for a total of let's just round it up to 9 ATO, 117 total damage assuming 100% accuracy and nonstop alphaing which is unrealistic I'd admit. Also does 9 alphas within 28.3 seconds of constant firing.
But oh wait! I forgot to take into account jump jet heat! Aaargh, I can't be bothered to crunch the number for that so let's just forget that for now as this number only gets worse anyway.
Oh right, did you also know that movement causes heat?

It's a pretty far cry to LPPCs you ask me.

Yeah alright should we try this with the Nova instead or the Viper maybe?
Yeah sure, let's go with Nova-A, same deal right? Poptart medium with PPC quirks, 3PC loadout.
https://mwo.nav-alph...=3f8de3db_NVA-A

Total DPS, 6.18 with splash, doing 19.5 damage per alpha.
ATO, 12.58, not counting JJ heat, pretty decent, a total of 234 damage on 12 alphas before overheat, again assuming 100% accuracy and nonstop firing. Also reaches overheat within 40 seconds of constant firing.

Maybe it's better off going with 2LPL or 2cERPPCs for better burst damage.

View PostDuke Falcon, on 11 March 2024 - 08:42 AM, said:

Anyone else feels that "continous" laser is like a clan X-pulse? Well, not exactly but kind of...

The heat damage is negligible. Plasma is trully but a cLPPC. And as such it could\may work pretty well if boated.
Proto ACs not looks so much "persuasive" currently. I think MGs would remain preferred on most builds instead of PrACs.

Thunderbolts...
Well, ladies and gentlemen, there's one in every family, so to say...
Sorry, no pun intended!



Beam laser works like flamers, 5 DPS, 1.66 DPH (Better than LXPL post-nerf coming this patch also), 3 HPS (dubious because flamer ramp mechanic), 600m OpRange (pretty good).
Kinda hard to compare in reality because LXPL benefits from cooldown and laserduration quirks/skills. While this thing has to deal with flamer ramp mechanic.

Flamer ramp mechanic works in that you have a meter that fills up after firing it for 5 seconds, this meter is shared across ALL mounted flamers. So if you have 1 flamer in weapon group 1 and 3 flamers in weapon group 2, ALL of the meters will fill up even if you only fire one flamer.
This meter fills up after 5 seconds of constant firing, begins ramping down AFTER 5 seconds of not firing and resets the delay everytime it fires, and fully empties after 5 seconds of ramping down from full meter. You can fire with full meter but the weapon gains exponential heat multiplier, unclear how much.

Just not benefiting from laser duration and cooldown already puts it down, but dealing with this on top? Kind of a headache to measure actually. Might be more obnoxious than LXPLs once it gets added since people would see it having 600m range and use it to snipe all over the place (only to realize it's less efficient than just using HLLs).

As for the L/PACs, they're really a bunch of whatevers even to me, PAC8 seems strikingly bad though.

AP/MGauss and the SBGauss are potentially the broken ones.
The SBGauss spread is narrow enough (it's tighter than LB2) that you can just use it like you would a gauss rifle at medium range except with superior cooldown.

AP/MGauss weighs the same as cHMG/IS MG with medium laser range and the damage of AC2s, extremely good for hit and run on paper.

Edited by Ttly, 11 March 2024 - 09:15 AM.


#50 Tiy0s

    Staff

  • Developer
  • Developer
  • 112 posts
  • LocationEdo, Turtle Bay

Posted 11 March 2024 - 09:15 AM

Current code doesn't support ammo for energy weapons.

But also, even if it did, we were not interested in doing it for Plasma Cannons due to how we implemented them.

Hope that clears things up.

#51 Quicksilver Aberration

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • The Nightmare
  • The Nightmare
  • 11,757 posts
  • LocationKansas City, MO

Posted 11 March 2024 - 10:04 AM

View PostTtly, on 11 March 2024 - 01:24 AM, said:


Actually I have seen game engine codes.
On some open source indie game, think of Roblox or whatever, yeah not exactly CryEngine sure.
It really *is* like a block game where you can "mix and match" mechanics in most cases.

I don't have reason to believe that it's that "extremely complicated" especially since it (MWO) is a live-service game that should've been from the ground up made to have new content patched in routinely.
Most of the new weapons also exhibits these "mix and match" features, the plasma cannon takes the heat damage component of the flamer into a PPC projectile, beam laser is just a flamer with no heat damage, more damage, and more range, etc.

It's not like the discussion was something more drastic as "please change how rendering distance/LoD works so people can't see through objects on low enviroment settings and preaim at people like they're wallhacking" or "make post-processing look less bad", "fix loading times", etc.

View PostScrapIron Prime, on 11 March 2024 - 05:21 AM, said:

You would think, right? But no. There's a reason CryEngine was so reviled back in the day and why its sales team went so hard after various game designers, lobbying ANYONE to adopt their core code and keep them in business. CryEngine sucks.

And here in MWO, the dev team never worked out a resource-practical way to animate melee, arm movement, squatting, or how to change from one ammo type to another with the same weapon. CryEngine likes to code you into a corner.

But I digress. If you really think its that easy to do, then head on over to that Discord server, sweet talk some folks into giving you a block or three of code, dust off your Lua coding skills, and show us all how its done!

I was going to add to LordNothing's post but it has nothing to do with even being able to read the engine code, it has to do with just understanding of typical software development lifecycles. Scopes increase, features get added that you never intended for. I'm not trying to white knight for PGI's management, but for the devs, this was their first mechwarrior game. Preparing for all potential changes that could come down the road would be impossible, especially if you don't have lessons learned from a prior iteration under your belt. Migrations in place are painful (looking at you MWO UI 2.0) and take a lot of time, so they have to have buy-in for management. Supporting ammo for energy weapons hadn't really come up until now so if I were them I would've thought the decision was honestly the right one, sometimes given time constraints the working solution is better than spending months on the "perfect" solution.

There is also a difference between an engine and the code you build on top of it. Engines like UE are really flexible, the code you write on top of it however may not be, and that's I think where the fundamental differences are.



On a separate note, I think the plasma cannon has too low of DPH as well considering it sacrifices range and upfront damage just to get faster cooldown (typically it is the reverse, the more DPS oriented weapon has slightly better DPH). 2 Plasma vs an ERPPC the ERPPC just seems like the better weapon. You get more damage for almost the same heat, better range, less slots, you just have to deal with a slower recycle time. It still looks dope and the Cauldron is expecting to have to do some balance passes, just like the whole software lifecycle stuff, sometimes it's better to just dump err on the side of caution and just dump all the new stuff and see where things land.

Edited by Quicksilver Aberration, 11 March 2024 - 10:11 AM.


#52 feeWAIVER

    Member

  • PipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 1,724 posts

Posted 11 March 2024 - 10:07 AM

I'm personally excited for the magshots.. I already have some ideas for those.
The plasma cannons look cool too.

#53 ScrapIron Prime

    Member

  • PipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 4,880 posts
  • LocationSmack dab in the middle of Ohio

Posted 11 March 2024 - 01:27 PM

View PostfeeWAIVER, on 11 March 2024 - 10:07 AM, said:

I'm personally excited for the magshots.. I already have some ideas for those.
The plasma cannons look cool too.


Oh yes. A Flea or a Crusader with 8 of them... Posted Image

Better yet... Blackjack Arrow with 6 and two LPL. Posted Image

Edited by ScrapIron Prime, 11 March 2024 - 01:35 PM.


#54 LordNothing

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 17,060 posts

Posted 11 March 2024 - 05:39 PM

View PostQuicksilver Aberration, on 11 March 2024 - 10:04 AM, said:

I was going to add to LordNothing's post but it has nothing to do with even being able to read the engine code, it has to do with just understanding of typical software development lifecycles. Scopes increase, features get added that you never intended for. I'm not trying to white knight for PGI's management, but for the devs, this was their first mechwarrior game. Preparing for all potential changes that could come down the road would be impossible, especially if you don't have lessons learned from a prior iteration under your belt. Migrations in place are painful (looking at you MWO UI 2.0) and take a lot of time, so they have to have buy-in for management. Supporting ammo for energy weapons hadn't really come up until now so if I were them I would've thought the decision was honestly the right one, sometimes given time constraints the working solution is better than spending months on the "perfect" solution.

There is also a difference between an engine and the code you build on top of it. Engines like UE are really flexible, the code you write on top of it however may not be, and that's I think where the fundamental differences are.



On a separate note, I think the plasma cannon has too low of DPH as well considering it sacrifices range and upfront damage just to get faster cooldown (typically it is the reverse, the more DPS oriented weapon has slightly better DPH). 2 Plasma vs an ERPPC the ERPPC just seems like the better weapon. You get more damage for almost the same heat, better range, less slots, you just have to deal with a slower recycle time. It still looks dope and the Cauldron is expecting to have to do some balance passes, just like the whole software lifecycle stuff, sometimes it's better to just dump err on the side of caution and just dump all the new stuff and see where things land.


i was mostly speaking in general terms (and not aimed at anyone in particular mind you). im still convinced that major parts of the engine are immutable binary blobs with no source. an unfortunate side effect of using proprietary closed source 3rd party blobs. this is my explanation for why some long standing bugs still exist, eg the fullscreen bug. these existed way back when pgi still did most of the work. its either that or they abandoned development on it. id like to think its a case of being unable to do something rather than choosing not to.

game logic would totally have source though, since it was written by pgi coders to their spec. if the cauldron had access to that, you would see some really interesting things. im sure they either have or could find programming talent. though im not sure its really necessary. we can have an assortment of new weapons without it. there are things id have liked to seen but were limited by whatever inbaked incompatabilities there are between mechanics.

#55 Quicksilver Aberration

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • The Nightmare
  • The Nightmare
  • 11,757 posts
  • LocationKansas City, MO

Posted 11 March 2024 - 06:21 PM

Oh I know you weren't talking to me, I was just gonna add on top of your comment.

"Just" is a dirty word in software engineering as it's often used by people without technical on contextual knowledge to know how involved a change is, it's a pet peeve of mine passed on by teams I've had the pleasure of working with.

Edited by Quicksilver Aberration, 11 March 2024 - 06:21 PM.


#56 LordNothing

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 17,060 posts

Posted 11 March 2024 - 09:19 PM

View PostScrapIron Prime, on 11 March 2024 - 05:21 AM, said:

dust off your Lua coding skills, and show us all how its done!


are the weapon classes defined in lua or are they inherited from a blob? because im pretty good with lua. asking for a friend.

#57 ScrapIron Prime

    Member

  • PipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 4,880 posts
  • LocationSmack dab in the middle of Ohio

Posted 12 March 2024 - 05:17 AM

View PostLordNothing, on 11 March 2024 - 09:19 PM, said:

are the weapon classes defined in lua or are they inherited from a blob? because im pretty good with lua. asking for a friend.


That's a question for the current developers. From the outside looking in, I would think everything is defined in tables. Blobs are mostly for handling of lots of data and input at once, so... rendering.

For example, our calculations of where the edge of an object that blocks your shot is handled by by bounding boxes in a blob after pulling the data from the 3D model. And we know how precise those are. Loves me them invisible walls!

#58 Curccu

    Member

  • PipPipPipPipPipPipPipPipPip
  • Liquid Metal
  • Liquid Metal
  • 4,623 posts

Posted 12 March 2024 - 05:54 AM

View PostScrapIron Prime, on 11 March 2024 - 01:27 PM, said:

Better yet... Blackjack Arrow with 6 and two LPL. Posted Image

Two snubbies and 6x magshots Posted Image ?

#59 Hawk819

    Member

  • PipPipPipPipPipPipPipPip
  • The Tip of the Spear
  • The Tip of the Spear
  • 1,677 posts
  • Location666 Werewolf Lane. Transylvania, Romania Ph#: Transylvania 6-5000

Posted 12 March 2024 - 11:20 AM

Other than what I read in the ilClan Series, I haven't found one single build (other than the Shadow Hawk) that carries the T-Bolt.

The Silver Bullet on the other hand is readily available: MAD-9D Marauder. There's one slight problem - Clan ER Lg Laser. One in each arm. Medium X-Pulse, nada problema.

Hopefully, we'll get this as a reward.

Here's a thought: Let us have cross tech for a change. There's a nifty idea.

As for the Mag Shot. No dice. Not a single build lorewise exists. Zip. Zero. Nada.

Plasma and PACs, oh yea, there're plenty of those for the Clans. And as luck would have it. I've got those variants already with different weapons to substitute the missing ones. Like the Summoner HH. Looking forward to that. You betcha.

I'm going to make a Mad Dog CC: x2 PAC/8's w/ x2 ammo, and Four Streak 4's with x4 tons ammo. Should be sweet for brawling. We'll see.

#60 Adam Sandler

    Member

  • Pip
  • 13 posts
  • LocationProudly living in Jade Falcon territory

Posted 12 March 2024 - 11:26 AM

I have seen the beam laser and it is beautiful





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users