Repository navigation
[RIP] My research acc got banned 👯 #53
Replies: 7 comments 23 replies
|
Sorry about the account, and thanks for writing this up so carefully. Until now my client misread exactly this packet: it treated the ban as a stale token and exited with code 2, which means "recapture the session". A recapture can't lift a ban, and each one kicks the phone. Based on your post:
As you suggested, I documented only the sign rule for the ban time and gave 703 no meaning. One small thing: the post quotes two different ban times, |
|
I've used this tool a few times to test previously, no issue. Log in to the game using phone to do upgrades etc for my test acc on and off a few times, in between didn't use this tool. Suddenly on phone when I came back to do new upgrades, it pops up a message telling me login issue detected. Please access the game on a mobile device. No idea how it happens, first time see this message |
|
Tha ban is due to robotic scanning, ask AI to visualize your scanning method on an artifact and you will say: OK this must be banned |
|
I built what I call a "Hub"—a website where you can look up various information about the game, similar to how Mapscout works. Ever since then, my bots keep getting banned, one after another—three bots in three days. To be honest, I’m not sure what to do or how to get around this issue. :( |
|
The strange thing is i use 3 accounts on the same email and ip. My farm account gets banned with On this account i dont have much interaction with alliance members. Normaly all 3 accounts should be banned. I scan on all 3 accounts. |
|
any one knows what is E116 error? did i get my phone banned from the login server? it says unable to connect to the server, but the contact us button able to load fine, my other phone able to login to game, only this phone. Reinstall from play store still gives this error, previously able to use it to play the game |

Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
Login failure
ec=28: how to read the packet (ban vs token vs identity)TL;DR:
ec=28is only SFS2X's generic custom-error wrapper. The real cause is theepstring.ep = "4;banTime;type;reason;roleName"is an account/role ban, not a wrong token, endpoint, ornetwork problem. This post documents the exact packet shape, the common parsing traps, and what the
client does with it.
1. The observed failure
A cross-server Login on
APS<selfServerId>returned this error instead of the success payload/init:The request itself had been sent normally:
The TCP/SFS connection was fine; the server answered at the application layer with the role's name. This
is not a “wrong endpoint” failure.
2. The actual wire object
The Login response is a system message (
c=0,a=1), not an extension request:Concrete shape:
Typical log rendering:
Parsing trap #1 —
epis one string, not five array elementsThe brackets and semicolons in the log are the bot's
SfsObject.ToStringformatting, not JSON and notfive array elements.
epcontains one string; the semicolons are inside that string.The client does:
MessageDispather.FnLogin(BaseUtils\ProtoBufNet\MessageDispather.cs:101-121) readsGetUtfStringArray("ep")→ a one-elementstring[];ErrorCodes.GetErrorMessage(28, ep); error code 28 is the passthrough format"{0}"(
ErrorCodes.cs:63) → the whole one-element string becomeserrorMessage;LoginState.OnLoginResponsesees the;, callsOnErrorMessageSemicolonSplit, and the first fieldbecomes the ban code (
4).Parsing trap #2 —
ecis 28, not 4The wire
ecis 28. The4is the first semicolon field insideep.ErrorCodeshas a separatecode 4 (
"User {0} is banned"), but that table entry is not used for this packet because the wire code is28. Don't conflate them.
3. The semicolon fields
UIAccountBanView.ReasonStrUpdate(Lua\UI\UIAccount\UIAccountBan\View\UIAccountBanView.lua:155-176)splits the string on
;:4ERROR_BANID(Assembly-CSharp\LoginErrorCode.cs:19); the client opensUIAccountBan-92233720368555758109223372036854775806sentinel is also all-time; positive is an expiry13703<roleName>With exactly four fields, the client uses field #4 as the name; with five, field #5 is the name and #4
the reason (
:157-166). Reason602is special-cased as account deletion (:192-196).No — the client does not translate that number into a date. It only checks the sign:
-- UIAccountBanView.lua:170-175
local timeCode = tonumber(spl[2])
if timeCode < 0 then
self.ban_time_value:SetLocalText(280098) -- all-time / permanent
else
self.ban_time_value:SetText(
UITimeManager:GetInstance():TimeStampToTimeForLocal(timeCode * 1000 * 10000 / 10000000))
end
So for your packet:
The offset from long.MinValue is 7,199,998, which is suspiciously close to 7,200,000 ms (2 hours), but there is no evidence in the client for a duration/timezone meaning. Do not document it as a 2-hour ban — treat it as a server-side permanent-ban sentinel.
For documentation, the safe rule is:
banTime < 0 -> permanent
banTime == 9223372036854775806 -> permanent (all-time)
banTime >= 0 -> expiry timestamp (client formats it)
4. What
703displays703is not in the client’s reason map. The map only contains:705operate_notice_10004706operate_notice_10003727operate_notice_10002602operate_notice_10005703falls back to705, so the game displays:“Data Error” is the unknown-code fallback, not a statement that the real reason is a data error. The
public locale for the same build (
locale/23609/en.bin, gzip-compressed table) also containsoperate_notice_10006= “Illegal Activities”, but there is no 703 mapping in this client build. The exactserver-side meaning of 703 requires support or a newer client mapping — don't guess it.
The
operate_notice_10001wrapper is:5. What the client does with it
MessageDispather.FnLogin(BaseUtils\ProtoBufNet\MessageDispather.cs:101-121):ec+ep→ErrorCodes.GetErrorMessage(28, ep)→errorMessage; callsOnLoginError.LoginMessage.CSHandleResponse→onLoginResponse(Assembly-CSharp\LoginMessage.cs:175-251).LoginState.OnLoginResponse(:95-210):errorMessagecontains;→OnErrorMessageSemicolonSplit;textbecomes"4"; passes(text, rest)toLoadingErrorState.LoadingErrorState(:125-142):text == "4"→ opensUIAccountBan; an emptytextshowsE100085.UIAccountBanView(:155-198) renders role, server, ban time, reason, and the support/exit buttons.No
initpush arrives; the client never enters the game.6.
ec=28: telling the common cases apartepvalue"4;<banTime>;<type>;<reason>;<role>"UIAccountBan; permanent if ban time is negative/all-time"E011"opt=refreshonce"E005"ta, package, platform, app/buildLoginErrorCode/UIUtilsmappingec=28alone is not diagnosable. Always capturefull={...}(or the pcap) so theepfield is visible.7. Why this is not an endpoint or token problem
account, so the gateway/zone/port are correct.
ep[0] = 4.E011would be the token case,and it is not present here.
same ban and ToS risk.
8. Reproducing / capturing it
cross-server login failed ... full={ep=[...], ec=28}.LWDEBUG_DUMP_LOGIN=1(logs the full Login object) andLWDEBUG_DUMP_LOGIN_BODY=<path>(writes the encoded request body).full=log is the decoded object.
tests/LastWar.Tests/SessionErrorsTests.cs; the classifier(
SessionErrors.ClassifyLoginEp) handles the one-element string and a pre-split array.9. What to do if you see this
epstarts with4;— that is the ban path.UIAccountBan/ “Data Error” for an unknownreason like 703.
zone/token); do not reuse the banned role's token.
LOGIN FAILED: ec=28withoutep, capture the full object first —ec=28alone couldbe
E011,E005, a ban, or another custom error.10. References
Assembly-CSharp\LoginErrorCode.cs:19—ERROR_BANID = "4".BaseUtils\ProtoBufNet\MessageDispather.cs:101-121—ec/ep→errorMessage.BaseUtils\ProtoBufNet\ErrorCodes.cs:63— code 28 ="{0}".Assembly-CSharp\LoginState.cs:95-210,LoadingErrorState.cs:125-142— ban path /UIAccountBan.Lua\UI\UIAccount\UIAccountBan\View\UIAccountBanView.lua:114-119,155-198— reason map and field split.https://lastwar-cdn.akamaized.net/hotupdate/locale/23609/en.bin(
SHA-256 14820C8A3F666CE48FED20C64E868A62015CD83E20FC123C47877A030A795088; decompressedSHA-256 0EF60A76A37D3C5EC783199C3A53ABA54A5E87D134372A5812241F71292C9810).All ids, tokens, role names, and server values in this post are placeholders. This is a protocol
analysis for diagnosing
ec=28;.No — the client does not translate that number into a date. It only checks the sign:
-- UIAccountBanView.lua:170-175
local timeCode = tonumber(spl[2])
if timeCode < 0 then
self.ban_time_value:SetLocalText(280098) -- all-time / permanent
else
self.ban_time_value:SetText(
UITimeManager:GetInstance():TimeStampToTimeForLocal(timeCode * 1000 * 10000 / 10000000))
end
So for your packet:
The offset from long.MinValue is 7,199,998, which is suspiciously close to 7,200,000 ms (2 hours), but there is no evidence in the client for a duration/timezone meaning. Do not document it as a 2-hour ban — treat it as a server-side permanent-ban sentinel.
For documentation, the safe rule is:
banTime < 0 -> permanent
banTime == 9223372036854775806 -> permanent (all-time)
banTime >= 0 -> expiry timestamp (client formats it)
All reactions