Jump to content
AnnieRuru

@packetfilter

Recommended Posts

No more Work in Progress, Please report if there are any bug

 

Download: 1.0a [Complete]

plugin

 


 

What is `@packetfilter` ?

`@packetfilter` is a custom mod originate from eamod

https://github.com/zephyrus-cr/eamod/blob/master/Servers/rAthena/conf/groups.conf#L418-L422

Quote

-Activates a mode to ignore some packets on WoE, comming from other users

-This reduce the bandwith a little bit on WoE, reducing LAG.

-C = Ignore jokes and global chat messages from others.

-I = Ignores Item usage from others.

packetfilter: true

it can actually reduce lag on your live server during woe depends on how many players activate it

it totally depends on the player's side, community effort that every player should together turn on the `@packetfilter` to reduce massive lag during woe

 

this modification is actually very famous and there are members willingly to pay for it

https://rathena.org/board/topic/121200-packetfilter-battleinfo/

https://rathena.org/board/topic/123203-battleinfo-packetfilter/

https://rathena.org/board/topic/111797-packetfilter-zepyhrus-ramod/

https://rathena.org/board/topic/79995-buy-noactnodelay-partybuff-src-as-it-was-on-pro-ro-or-packetfilter/

 

 


 

this `@packetfilter` is unlike the one you are having, or that is circulating on rathena forum

 

this one has been written entirely from scratch and added some other flags that the original don't

for example,

`@packetfilter COHM EOHM`

allow block normal chat and emotion from non-related players, pets, homunculus and monsters
but this doesn't block normal chat from party/guild members

 

a full list of flags are

  •   C = ignore normal Chat, including Dancer's scream, Bard's frost joke, and pet's talk (pet start to talk when intimacy > 900)
  •   E = ignore emotion, including monster and pet emotion (but NOT pet performance)
  •   I = ignore Item use effect
  •   A = ignore normal Attack animation (based on receiving side), your own attack animation are always shown
  •   B = Buffing skill animations (based on receiving side), your own casting animation are always shown
  •   S = Status effect
  •   T = offensive single Target skill/spell animations (based on receiving side), your own casting animation are always shown
  •   G = Ground based skills (eg:storm gust), your own casting animation are always shown
  •   M = Music dance/songs from Bard/Dancer
  •   D = Direction. Recommend filter on Party/Guild type and not enemies. Highly Recommend for supportive type Class.

after the Main type, also support additional flags

  •   S = Self
  •   P = Party
  •   G = Guild
  •   B = Battleground
  •   C = Clan
  •   O = Other players, none of the above
  •   H = Homunculus/pets/elementals/mercenary and player's @summon
  •   M = Monsters

 


 

Why This plugin doesn't block Kaite or Energy Coat ?

 

2 reasons

 

No.1 - currently there is a bug with plugin,

struct packet_spawn_unit *p = (struct packet_spawn_unit*)RBUFP(buf, 0);

this line will throw error on Linux

so no choice, maybe I'll make a bug report

this bug also has been reported on Dastgir's `@noview`

 

No.2 - just set p->virtue = 0; isn't really blocking the packet

this goes against the very principle of this plugin, which is NOT send certain packet to the client

OPT3 (including OPT3_KAITE | OPT3_ENERGYCOAT | OPT3_SOULLINK) is part of the idle_unitType, unit_walkingType and spawn_unitType function

and blocking those 3 packet entirely is stupid

 

in case you don't understand, look under clif_set_unit_walking and clif_set_unit_idle function inside src\map\clif.c file

	p.GUID = g_id;
	p.GEmblemVer = status->get_emblem_id(bl);
	p.honor = (sd) ? sd->status.manner : 0;
//	p.virtue = (sc) ? sc->opt3 : 0; <-- change this line
	p.virtue = 0;
	p.isPKModeON = (sd && sd->status.karma) ? 1 : 0;
	p.sex = vd->sex;

 


 

Note: about [D]irection type packet

 

try add a ShowDebug inside clif->send

 src/map/clif.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/src/map/clif.c b/src/map/clif.c
index ab3c4422a..9e245f335 100644
--- a/src/map/clif.c
+++ b/src/map/clif.c
@@ -450,6 +450,7 @@ static bool clif_send(const void *buf, int len, struct block_list *bl, enum send
 			return true;
 	}
 
+ShowDebug("source:%s buf:0x%x send_target:%d\n", (sd != NULL)? sd->status.name : "NULL", RBUFW(buf, 0), type);
 	switch(type) {
 		case ALL_CLIENT: //All player clients.
 			iter = mapit_getallusers();

you will find 0x9c is being send repeatedly on every attack and every skill cast , getting attack etc

this direction update packet isn't just send by Shift-click, but also including all the actions above

this packet is very spam-able and I think can be safely block on party/guild/clan/bg flags

for supportive class, I also think can block all of them since supportive class doesn't need information from enemies

 

 

Spoiler

0.1 - plugin

- proof of concept

 

0.2 - plugin

- add sub-type flags (Self/Party/Guild/Bg/Other/Homun/Mob)

- add emotion filter

- fix compiler warning that doesn't display on visual studio

 

0.3 - plugin

- add missing FILTER_CLAN flag

- implement block_item type 'I' flag

- fix FILTER_OTHER not working properly when the player doesn't have a party or guild

- fix [C]hat type shouldn't able to filter self, since the client always shows you are talking anyway

- and attempt to fix the above stupid error that doesn't show on Visual Studio

 

0.4 - plugin

- if its not BL_PC|BL_PET|BL_HOM|BL_MER|BL_ELEM|BL_MOB immediately jump out of function for faster processing

- fix sd->status.clan_id

- implement block attack [A] and buff [B ] flag

- don't need to use snprintf needlessly when print out the flags

- just unset the flag if the SELF flag isn't supported

 

0.5 - plugin

- fix linux compile warning `type` not being use

- change useItemAckType into 0x1c8 ... hopefully no more nullpo ?

- allow blocking casting animation, save more bandwidth

- implement [T], [G], [D] type respectively

 

1.0 - plugin

- introduce type [D], direction packet, and existing Dance/songs move to type `[M]usic`

- no longer use nested IF-ELSE, use switch statement on the packet header for faster read speed

- fix [I.]tem type throw nullpo error on 2018 client

- add Caution as a warning that certain packet filter wont work correctly

 

Edited by AnnieRuru

Share this post


Link to post
Share on other sites

I've never seen this kinda of modification and i think its very neat one to maybe put as a forced settings via mapflag.

Maybe the ability to only see the damage you deal and recieve, the packet would still be sent anyway because I think it's bounded to "receive damage" animaiton effect, but might reduce image cluster when maps are crowded. Maybe if passing it as 0 would reduce memory? noidea

Not sure what else could we block from been area sent ^^

Gratz on the project

 

Share this post


Link to post
Share on other sites
3 hours ago, raPalooza~ said:

I've never seen this kinda of modification and i think its very neat one to maybe put as a forced settings via mapflag.

Maybe the ability to only see the damage you deal and recieve, the packet would still be sent anyway because I think it's bounded to "receive damage" animaiton effect, but might reduce image cluster when maps are crowded. Maybe if passing it as 0 would reduce memory? noidea

mf_gvg mapflag and mf_gvg_castle mapflag with `@agitstart` on will disallow showing the damage packet already ? I think they sent as 1 damage or something

 

https://github.com/HerculesWS/Hercules/blob/1354b01c7a24e9ffd9ea570cbaec668721fc2a6c/conf/map/battle/client.conf#L55-L58

Quote

// When set to true, the damage field in packets sent from woe maps will be set
// to -1, making it impossible for GMs, Bots and Hexed clients to know the
// actual damage caused by attacks. (Note 1)
hide_woe_damage: true

search hide_woe_damage in clif.c, the packet indeed altered to become if (skill is multiple hit)? number of hits : 0;

	if( battle_config.hide_woe_damage && map_flag_gvg2(src->m) ) {
		p.damage = damage?div:0;
		p.leftDamage = damage2?div:0;
	} else {
		p.damage = damage;
		p.leftDamage = damage2;
	}

 

and yes, this modification is indeed REMOVE certain packet send to reduce server bandwidth

this can only stays as plugin, and could never get into our emulator, despite how popular it is

Share this post


Link to post
Share on other sites

@AnnieRuru

 

I knew about the woe hiding the damage, but I was wondering if hiding everyones damage but yours would reduce bandwidth and memory.

 

Example:

If tou have 20 player hiting a worldboss for 123456789(damages are high nowadays)

 

That wouldbe 20 x 12356789 packets sent with that value per player...

 

If it shows only your (self) damage there would be 1 x 123456789 damage packet and 19 x -1 for the damage animation.

(Hope its clear)

Do you think It would improve performance?

 

Share this post


Link to post
Share on other sites

actually normal hit a monster send multiple packets

 

let's put a ShowDebug on clif->send function

 src/map/clif.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/map/clif.c b/src/map/clif.c
index 54c9869c3..38ac2ea1e 100644
--- a/src/map/clif.c
+++ b/src/map/clif.c
@@ -449,7 +449,7 @@ static bool clif_send(const void *buf, int len, struct block_list *bl, enum send
 		else if (type == AREA_WOS || type == BG_WOS || type == BG_AREA_WOS)
 			return true;
 	}
-
+ShowDebug("source:%s buf:%d send_target:%d\n", (sd != NULL)? sd->status.name : "NULL", RBUFW(buf, 0), type);
 	switch(type) {
 		case ALL_CLIENT: //All player clients.
 			iter = mapit_getallusers();

`@monster poporing` and hit it once, and here is what I get

[Debug]: source:NULL buf:2248 send_target:2
[Debug]: source:AnnieRuru buf:156 send_target:2
[Debug]: source:NULL buf:2557 send_target:2
[Debug]: source:AnnieRuru buf:2248 send_target:2
[Debug]: source:NULL buf:156 send_target:2
[Debug]: source:AnnieRuru buf:2248 send_target:2
[Debug]: source:NULL buf:156 send_target:2
[Debug]: source:AnnieRuru buf:2248 send_target:2

analyze the packet

2248 hex 0x8C8 is ZC_NOTIFY_ACT2 came from clif_damage function

/// Sends a 'damage' packet (src performs action on dst)
/// 008a <src ID>.L <dst ID>.L <server tick>.L <src speed>.L <dst speed>.L <damage>.W <div>.W <type>.B <damage2>.W (ZC_NOTIFY_ACT)
/// 02e1 <src ID>.L <dst ID>.L <server tick>.L <src speed>.L <dst speed>.L <damage>.L <div>.W <type>.B <damage2>.L (ZC_NOTIFY_ACT2)
/// 08c8 <src ID>.L <dst ID>.L <server tick>.L <src speed>.L <dst speed>.L <damage>.L <IsSPDamage>.B <div>.W <type>.B <damage2>.L (ZC_NOTIFY_ACT2)
/// type: @see enum battle_dmg_type
///     for BDT_NORMAL: [ damage: total damage, div: amount of hits, damage2: assassin dual-wield damage ]
static int clif_damage(struct block_list *src, struct block_list *dst, int sdelay, int ddelay, int64 in_damage, short div, enum battle_dmg_type type, int64 in_damage2)

156 hex 0x9c is ZC_CHANGE_DIRECTION came from clif_changed_dir function

/// Updates body and head direction of an object (ZC_CHANGE_DIRECTION).
/// 009c <id>.L <head dir>.W <dir>.B
/// head dir:
///     0 = straight
///     1 = turned CW
///     2 = turned CCW
/// dir: @see enum unit_dir
static void clif_changed_dir(struct block_list *bl, enum send_target target)

2557 hex 0x9fd is unit_walkingType (src\map\packet_struct.h) ... couldn't find where this came from ... I rather don't touch this at the moment

 

this means, whenever I attack something, it sends an attack animation and also updates the client to face the attacking source enemy

 

so if we block ZC_CHANGE_DIRECTION for the monster, this will cause rogue's backstab to not function properly

 

13 hours ago, raPalooza~ said:

Do you think It would improve performance?

the answer is YES, block ZC_NOTIFY_ACT2

here's the plugin `@filterattack`

https://github.com/AnnieRuru/Release/blob/master/plugins/packetfilter/filterattack.c

 

after tested it, `@filterattack 63` looks like everyone is standing still facing each other (remember ZC_CHANGE_DIRECTION packet is still sent)

 

 

Share this post


Link to post
Share on other sites
10 hours ago, AnnieRuru said:

the answer is YES, block ZC_NOTIFY_ACT2

here's the plugin `@filterattack`

https://github.com/AnnieRuru/Release/blob/master/plugins/packetfilter/filterattack.c

after tested it, `@filterattack 63` looks like everyone is standing still facing each other (remember ZC_CHANGE_DIRECTION packet is still sent)

 

 


Nice analysis for my random question xD. Thanks for all the informations ^^.

I think you can reverse the development process and see where the "playable" treshold stands at.
For example blocking all the packet and then enabling back, while doing some kinda of perfomance test to see the hit on memory.
You would probably need to setup a test environment, eg, 1,4,8,20 players in the area.

But that's a lot of work xD And afterall the project already stants on a very awesome state so congratz again.

Share this post


Link to post
Share on other sites

0.2 - plugin

- add sub-type flags (Self/Party/Guild/Bg/Other/Homun/Mob)

- add emotion filter

- fix compiler warning that doesn't display on visual studio

 

actually there are a lot of people interested in this topic...

@xJhay

 

let me just copy paste from a PM

 


 

the reason I started this topic because I received the `@packetfilter` for rathena use, and asked me to convert into plugin

and upon reading it I found its very inconvenient and thus I want to rewrite it

that version I receive is block based on skill ID

 

but my plan is to block based on flags, not every skill ID entry

if block by skill type listed in skill_db,conf will make the plugin future proof

	SkillType: {                                (bool, defaults to "Passive")
		Passive: true/false                     (boolean, defaults to false)
		Enemy: true/false                       (boolean, defaults to false)
		Place: true/false                       (boolean, defaults to false)
		Self: true/false                        (boolean, defaults to false)
		Friend: true/false                      (boolean, defaults to false)
		Trap: true/false                        (boolean, defaults to false)
	}
	SkillInfo: {                                (bool, defaults to "None")
		Quest: true/false                       (boolean, defaults to false)
		NPC: true/false                         (boolean, defaults to false)
		Wedding: true/false                     (boolean, defaults to false)
		Spirit: true/false                      (boolean, defaults to false)
		Guild: true/false                       (boolean, defaults to false)
		Song: true/false                        (boolean, defaults to false)
		Ensemble: true/false                    (boolean, defaults to false)

 

PS: just like card seller A-Z, before I made that script every member posted on the board manually insert each card ID into each npc,
until I made that SQL version which rip from SQL database, that old script still works today, make it future proof

 


 

another thing to note about that version is because it block base on skill ID only, it doesn't make use of the flag of each packet

 

on the battlefield, or I mean during woe, there are a lot of packets being sent even if its unnecessary information

for example, your guild and another guild A attacking defending guild B,

and your guild and guild A are not on good terms and rushing to destroy the emperium

my point is, guild A attacking guild B are unnecessary in my opinion, and I don't need those packet informations

but I still need to know my party/guild are under attack

 

/// 02e1 <src ID>.L <dst ID>.L <server tick>.L <src speed>.L <dst speed>.L <damage>.L <div>.W <type>.B <damage2>.L (ZC_NOTIFY_ACT2)

and I can use <src ID> and <dst ID> and check are they the same guild ? then send packet

is <dst ID> my guild mates ? then send packet

the rest are not important

 

that version simply block by skill ID and from my understanding, I wouldn't know who is actually healing me when enable `@packetfilter`

there are some healer who are good in healing others and some just standing there acting like a doll waiting to get hit

 

my idea is make additional field so only filter out unneeded information

when playing as crusader, my job is to defend my party, and I want all information packets send that are threatening my party

when playing as assassin, I don't have to take care of guild member, I only want to assassinate the healer or dd who are actively killing the most of my guildmates etc

yeah, make it highly configurable

 


 

anyway read the additional information I already made this part a reality,

but so far only chat and emotions are implement, skills has to do ... maybe tomorrow ...

Share this post


Link to post
Share on other sites

PLEASE GIVE ME YOUR COMPILER

I want my visual studio 2019 also display error like yours

btw I'm guessing that's CentOS, this happened before on my OnPCUseSkillEvent

 

0.3 - plugin

- add missing FILTER_CLAN flag

- implement block_item type 'I' flag

- fix FILTER_OTHER not working properly when the player doesn't have a party or guild

- fix [C]hat type shouldn't able to filter self, since the client always shows you are talking anyway

- and attempt to fix the above stupid error that doesn't show on Visual Studio

Share this post


Link to post
Share on other sites

@AnnieRuru i not sure is this can works, but you can try use clang in visual studio. it probably will show more errors.

or use WSL or WSL2 with linux kernel and tools inside.

 

 

Share this post


Link to post
Share on other sites
On 11/6/2020 at 10:57 PM, 4144 said:

@AnnieRuru i not sure is this can works, but you can try use clang in visual studio. it probably will show more errors.

or use WSL or WSL2 with linux kernel and tools inside. 

https://devblogs.microsoft.com/cppblog/clang-llvm-support-in-visual-studio/

and maybe this one ?
https://devblogs.microsoft.com/dotnet/debug-your-net-core-apps-in-wsl-2-with-visual-studio/

 

visual_studio_with_clang.png

 

... maybe next time ...

 


 

0.4 - plugin

- if its not BL_PC|BL_PET|BL_HOM|BL_MER|BL_ELEM|BL_MOB immediately jump out of function for faster processing

- fix sd->status.clan_id

- implement block attack [A] and buff [B ] flag

- don't need to use snprintf needlessly when print out the flags

- just unset the flag if the SELF flag isn't supported

 

On 10/26/2020 at 8:07 AM, freezing1 said:

no kaute/kaipe pls :(

Kaite.png

Kaite is indeed very interesting

I'm 100% sure I've block the casting animation for Kaahi Kaizel Kaupe ... and of course Kaite

but I think what you guys want is the effect of the Kaite to turn off ?

that should be another kind of source edit plugin

 

EDIT: wait, let me put ShowDebug on it again .... editing ....

 

it send 2 types of packet, 0x9cb and 0x983

0x9cb is clif_skill_nodamage

0x983 is clif_status_change_sub

means has to read p.index to know the the skill ID ....

 

EDIT2: 0x229 ... WOW don't block this packet

if the player is hiding/cloaking, enable this packetfilter will show her !!

needs to be cautious about this packet...

 

EDIT3: I think I figure it out, it is OPT3, means just set OPT3 into 0 will do

Edited by AnnieRuru

Share this post


Link to post
Share on other sites
On 11/6/2020 at 7:59 PM, AnnieRuru said:

PLEASE GIVE ME YOUR COMPILER

I want my visual studio 2019 also display error like yours

btw I'm guessing that's CentOS, this happened before on my OnPCUseSkillEvent

 

0.3 - plugin

- add missing FILTER_CLAN flag

- implement block_item type 'I' flag

- fix FILTER_OTHER not working properly when the player doesn't have a party or guild

- fix [C]hat type shouldn't able to filter self, since the client always shows you are talking anyway

- and attempt to fix the above stupid error that doesn't show on Visual Studio

Tried the latest Ms. @AnnieRuru and getting error on @packetfilter IOHM

 

image.png.a3285b9142500d7010fdba876e36c1af.png

Share this post


Link to post
Share on other sites
12 hours ago, AnnieRuru said:

https://devblogs.microsoft.com/cppblog/clang-llvm-support-in-visual-studio/

and maybe this one ?
https://devblogs.microsoft.com/dotnet/debug-your-net-core-apps-in-wsl-2-with-visual-studio/

 

visual_studio_with_clang.png

 

... maybe next time ...

 


 

0.4 - plugin

- if its not BL_PC|BL_PET|BL_HOM|BL_MER|BL_ELEM|BL_MOB immediately jump out of function for faster processing

- fix sd->status.clan_id

- implement block attack [A] and buff [B ] flag

- don't need to use snprintf needlessly when print out the flags

- just unset the flag if the SELF flag isn't supported

 

Kaite.png

Kaite is indeed very interesting

I'm 100% sure I've block the casting animation for Kaahi Kaizel Kaupe ... and of course Kaite

but I think what you guys want is the effect of the Kaite to turn off ?

that should be another kind of source edit plugin

 

EDIT: wait, let me put ShowDebug on it again .... editing ....

 

it send 2 types of packet, 0x9cb and 0x983

0x9cb is clif_skill_nodamage

0x983 is clif_status_change_sub

means has to read p.index to know the the skill ID ....

 

EDIT2: 0x229 ... WOW don't block this packet

if the player is hiding/cloaking, enable this packetfilter will show her !!

needs to be cautious about this packet...

 

EDIT3: I think I figure it out, it is OPT3, means just set OPT3 into 0 will do

thank you for your dedication and for helping us, Annie. 

Share this post


Link to post
Share on other sites

0.5 - plugin

- fix linux compile warning `type` not being use

- change useItemAckType into 0x1c8 ... hopefully no more nullpo ?

- allow blocking casting animation, save more bandwidth

- implement [T], [G], [D] type respectively

 

also, I split the original Land Protector and bard/dancer song out

Land Protector is under [G] type, Apple of Idun is under [D] type

Basilica is under [B.] type

 

type [S.] still has some bug, if being cast directly in front of the player, it doesn't block correctly

.. very close to finish

 

@Kuroyama, try again

Share this post


Link to post
Share on other sites
3 hours ago, AnnieRuru said:

0.5 - plugin

- fix linux compile warning `type` not being use

- change useItemAckType into 0x1c8 ... hopefully no more nullpo ?

- allow blocking casting animation, save more bandwidth

- implement [T], [G], [D] type respectively

 

also, I split the original Land Protector and bard/dancer song out

Land Protector is under [G] type, Apple of Idun is under [D] type

Basilica is under [B.] type

 

type [S.] still has some bug, if being cast directly in front of the player, it doesn't block correctly

.. very close to finish

 

@Kuroyama, try again

I did try it now Ms. @AnnieRuru. Here's the new error:

 

On windows compiler:

image.png.9d91cd547aef3febebeedea8b2395838.png

 

On CentOS/Linux:

Spoiler


image.png.87ff2e8078b5f0a087acec5bd40b917f.png

 

image.png.fbb209c7b91b4a99ace4093d1c6a2353.png

 

image.png.4f55c67bffd5b5699ab4b82f5601190c.png

 

image.png.15c7526510d67698439ab08fd9901576.png

 

 

Share this post


Link to post
Share on other sites
28 minutes ago, Kuroyama said:

I did try it now Ms. @AnnieRuru. Here's the new error:

 

On windows compiler:

image.png.9d91cd547aef3febebeedea8b2395838.png

 

On CentOS/Linux:

  Reveal hidden contents

 

image.png.87ff2e8078b5f0a087acec5bd40b917f.png

 

image.png.fbb209c7b91b4a99ace4093d1c6a2353.png

 

image.png.4f55c67bffd5b5699ab4b82f5601190c.png

 

image.png.15c7526510d67698439ab08fd9901576.png

 

 

then you have to tell which client version you are using,
this was tested on 2019 RE client

 

seems like the packet is entirely different

Share this post


Link to post
Share on other sites

I found this in src\map\packet_struct.h

struct PACKET_ZC_USE_ITEM_ACK {
	int16 packetType;
	int16 index;
#if PACKETVER_MAIN_NUM >= 20181121 || PACKETVER_RE_NUM >= 20180704 || PACKETVER_ZERO_NUM >= 20181114
	uint32 itemId;
	uint32 AID;
#elif PACKETVER >= 3
	uint16 itemId;
	uint32 AID;
#endif
	int16 amount;
	uint8 result;
} __attribute__((packed));

so client version 20181121 below uses `RBUFL(buf, 6)`, anything above use `RBUFL(buf, 8)`

 

1.0 - plugin

- introduce type [D], direction packet, and existing Dance/songs move to type `[M]usic`

- no longer use nested IF-ELSE, use switch statement on the packet header for faster read speed

- fix [I.]tem type throw nullpo error on 2018 client

- add Caution as a warning that certain packet filter wont work correctly

 

and I guess this is finally over

time to mention everyone that has PM me about this project

@raPalooza~

@Azhura

@freezing1

@xJhay

@Kuroyama

@fiction

@Skyzone

@cusco ? <-- from discord

 

Share this post


Link to post
Share on other sites

if you want change some fields in packet, better use pointer to struct

PACKET_ZC_USE_ITEM_ACK

 

Share this post


Link to post
Share on other sites
On 11/10/2020 at 4:31 AM, AnnieRuru said:

I found this in src\map\packet_struct.h

struct PACKET_ZC_USE_ITEM_ACK { int16 packetType; int16 index; #if PACKETVER_MAIN_NUM >= 20181121 || PACKETVER_RE_NUM >= 20180704 || PACKETVER_ZERO_NUM >= 20181114 uint32 itemId; uint32 AID; #elif PACKETVER >= 3 uint16 itemId; uint32 AID; #endif int16 amount; uint8 result; } __attribute__((packed));


struct PACKET_ZC_USE_ITEM_ACK {
	int16 packetType;
	int16 index;
#if PACKETVER_MAIN_NUM >= 20181121 || PACKETVER_RE_NUM >= 20180704 || PACKETVER_ZERO_NUM >= 20181114
	uint32 itemId;
	uint32 AID;
#elif PACKETVER >= 3
	uint16 itemId;
	uint32 AID;
#endif
	int16 amount;
	uint8 result;
} __attribute__((packed));

so client version 20181121 below uses `RBUFL(buf, 6)`, anything above use `RBUFL(buf, 8)`

 

1.0 - plugin

- introduce type [D], direction packet, and existing Dance/songs move to type `[M]usic`

- no longer use nested IF-ELSE, use switch statement on the packet header for faster read speed

- fix [I.]tem type throw nullpo error on 2018 client

- add Caution as a warning that certain packet filter wont work correctly

 

and I guess this is finally over

time to mention everyone that has PM me about this project

@raPalooza~

@Azhura

@freezing1

@xJhay

@Kuroyama

@fiction

@Skyzone

@cusco ? <-- from discord

 

 

On 11/10/2020 at 12:57 PM, 4144 said:

if you want change some fields in packet, better use pointer to struct

PACKET_ZC_USE_ITEM_ACK


PACKET_ZC_USE_ITEM_ACK

 

 

@AnnieRuru @4144

Is it possible to include the announcements. map announce and global announce?

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...

×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use.