-
Posts
88 -
Joined
-
Last visited
-
Days Won
1
Content Type
Profiles
Forums
Downloads
Jobs Available
Server Database
Third-Party Services
Top Guides
Store
Crowdfunding
Posts posted by Digos
-
-
First, congratulations! This is an amazing work. I see a lot of potential on it.
Are you planning to implement the AI for 3rd jobs ?
If you authorize, I can make a patch file from your git to easily apply it on the latest rAthena repository.
I hope you keep the project going and thanks for share it.
-
Hello @Tokei
I don't know the exact reason, but the GRF encryption feature is not working on Wine versions after 2.7, I was using it to protect my palettes.
After the above wine version, the client shows a black palette for head and body, the same effect as if the file was missing.
Tested on older client (2016) and a new one (2018).Can you to help me on this issue? Thank you!
-
On 3/20/2019 at 9:51 PM, TrueNoir said:
Does anyone know what commands work on the tipbox i know there's /n for line breaks but is there any other commands that work with it cause it seems to overlay the lettering ontop of each making it unreadable and i'm trying to figure out how to fix it.
Edit: nvm figured it out it just works weird as hell.
Can you tell us how you fixed it?
-
Confirmed to work on 2016-12-28aRagexeRE. Thanks for the fix!Oops game crashed after a while
-
This change proposed by Feilor should increase the performance in all queries related to items stolen since will not be necessary lookup the charlog table with a 'like' clause (it's so more slow than a query with =).
I would suggest also add a guild_id field in picklog since we already have an enum with a value that means the item was moved to the guild storage but currently is very hard to find out what guild was.
If the changes are made should be considered that currently the picklog table can be quite big in running servers so everyone must be careful on 'alter table' queries that will take much time to finish.
-
1
-
-
I've successfully compiled on macOS High Sierra with the following steps:
- Install the latest Xcode from Apple App Store
- Install homebrew https://brew.sh
-
Install the following packages using homebrew on Terminal
- brew install mysql
- brew install pcre
-
Go to the server directory on Terminal and type
- ./configure CC=clang CXX=clang++ --with-mysql=/usr/local/opt/mysql/bin/mysql_config --with-pcre=/usr/local/opt/pcre/
- make clean && make server
-
Hello,
After the rathena changed to c++11 support I can't successful compile the server on macOS (High Sierra).
Even using gcc-6/g++-6 from homebrew the compilation fails.
The error is the following:
... LD char-server Undefined symbols for architecture x86_64: "std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >::find_last_not_of(char, unsigned long) const", referenced from: YAML::ScanScalar[abi:cxx11](YAML::Stream&, YAML::ScanScalarParams&) in yaml-cpp.a(scanscalar.o) ...
The undefined symbols error repeats many times.
So I ask if someone have successfully compiled on macOS can tell how this was done?
Thanks!
-
Some time ago I was experiencing map-server crashes caused by the usage of "instance_announce -1," command (with the -1 param).
Also it's a good practice to debug the map-server using gdb to troubleshoot crashes. Or at least inspect the coredump with gdb.
How to run gdb:
https://www.youtube.com/watch?v=USPvePv1uzE
https://www.youtube.com/watch?v=mlfz6c9frSU
-
Yes you can set a mapflag via script. It certainly is more flexible but if it's easier it's a point of view as it need to be invoked from a running script.
Quote*setmapflag "<map name>",<flag>{,<zone>{,<type>}};
This command marks a specified map with the given map flag, which will alter the
behavior of the map. A full list of mapflags is located in 'src/map/script_constants.h' with
the 'mf_' prefix, and documentation can be found in 'doc/mapflags.txt'.The map flags alter the behavior of the map regarding teleporting (mf_nomemo,
mf_noteleport, mf_nowarp, mf_nogo), storing location when disconnected
(mf_nosave), dead branch usage (mf_nobranch), penalties upon death
(mf_nopenalty, mf_nozenypenalty), PVP behavior (mf_pvp, mf_pvp_noparty,
mf_pvp_noguild), WoE behavior (mf_gvg,mf_gvg_noparty), ability to use
skills or open up trade deals (mf_notrade, mf_novending, mf_noskill, mf_noicewall),
current weather effects (mf_snow, mf_fog, mf_sakura, mf_leaves, mf_rain, mf_clouds,
mf_fireworks) and whether night will be in effect on this map (mf_nightenabled).The optional parameter 'zone' is used to set the zone for restricted mapflags.
For the 'skill_damage' mapflag, 'zone' functions as 'value' (-100 to 100000) and
'type' can be:
1: damage against players
2: damage against mobs
3: damage against bosses
4: damage against other
5: caster typeTo disable all kind of commands, you should use mf_nocommand
-
1
-
-
I've made a pull request to fix this error
https://github.com/rathena/rathena/pull/2461
You can check the changes and apply to your repo https://github.com/rathena/rathena/pull/2461/files
Also is recommended to use a C++11 compiler as it don't raise errors for that syntax
-
A very similar problem was solved in this post: https://github.com/rathena/rathena/issues/2127
-
I think getd / setd are the commands that you are looking for
setd ".@" + .@var$ + "123$", "Poporing is cool"; mes getd ".@" + .@var$ + "123$";
-
I also had issue like this before by using a broken text editor that mess the encoding of text areas not modified like the Akkarin's comment above.
Usually I edit all the Korean text using the Windows Latin 1 encoding that works fine. If you like to see the Korean characters, I suggest you to use the Korean Windows DOS encoding.
Note that if you open the text as Windows Latin 1 and reinterpret it as Korean Windows DOS it should work without need to convert the text, convertion that could mess the text encoding.
-
Hello,
I know that there's an issue open at Nemo github regarding the patch Ignore Resource Errors which is not working in some clients.
Currently I'm using 2016-02-03 and the same as explained in the Nemo Issue #155 happens: There's no error when patching on Nemo. But the errors in the game are not ignored.
My question is if someone know a newer (2016 or 2017) hexed that this patch is working properly.
Thanks
-
I've heard about this issue in some topics at harmony board and here
https://rathena.org/board/topic/96124-harmony-version-3310/
https://rathena.org/board/topic/86520-server-hardwarespecs/
http://foro.warning-ro.net/index.php?/topic/329-harmony-fps-lag/
I mean a lot of players in the same area like more than 60 in a gvg match.
@xlaws27
Try to monitor your server resources usage with a more detailed application like atop.
Also consider to use a monitoring tool at client side like rcx to view the FPS and the latency
-
there reports that harmony causes a heavy graphical lag when there's many players in the same area.
-
As I remember there's a diff made by Shakto that allows it. Also, the command not work anymore, instead, there's an icon that you can create party and guild too.
If you don't have the diff, you can easily use a hex editor and apply the following search and replace because the Shakto's diff was based on my topic:
https://rathena.org/board/topic/81884-hex-to-allow-space-in-guild-name/
-
1
-
-
Não to falando do usuário e senha do banco de dados. É uma conta mesmo do servidor, como se fosse uma conta de player.. Vc tem que inserir essa conta aí na tabela login com o campo sexo marcado como S
-
1
-
-
parceiro, de acordo com suas configurações, faltou inserir o usuário ragnarok com senha ragnarok na tabela login (não esqueça de deixar o sexo dessa conta como S)
-
1
-
-
You could also use the atoi function http://rathena.org/wiki/Atoi
-
Hello,
Someone knows if two or more login_servers can use the same login db with no issues?
Thanks
-
The Goal
The proxy protocol comes to provide high availability to the RO servers by placing the server behind Load Balancers but preserving the ability of keep the user real ip address.
By using load balancers it's possible to protect against DDoS attacks, create custom hops to improve the route, even create a distributed network over the globe and many others possibilities.
How It Works
The Proxy Protocol was designed to chain proxies / reverse-proxies without losing the client information.
A proxy will use its own IP stack to get connected on remote servers. Because of this, we lose the initial TCP connection information like source and destination IP and port when a proxy in involved in an architecture.
A few workarounds exist, but the problem of these workarounds is that they are either protocol related or require architecture changes, preventing scalability.
That’s where the proxy-protocol comes in:- it is protocol agnostic (can work with any layer 7 protocols, even when encrypted).
- it does not require any infrastructure changes
- nat-ing firewalls have no impact it
- it is scalable
The is only one condition: both endpoints of the connection MUST be compatible with proxy protocol. This could be either proxies, reverse-proxies, load-balancers, WAF, application servers, etc….
Proxy-protocol ready softwaresThe list below summarizes which software have already implemented the proxy protocol:
Elastic Load Balancing, since July 2013, AWS’ Load-Balancer
exaproxy, since 1.0.0, forward and reverse proxy
exim, since 4.83, client side only, SMTP MTA
gunicorn, since 0.15.0, python HTTP server
haproxy, since 1.5-dev3, reverse-proxy load-balancer
nginx, since 1.5.12 in HTTP server client side only, Web server, HTTP + Mail reverve-proxy
postfix, since 2.10, SMTP MTA
stud, since the first release, SSL offloader
stunnel, since 4.45, SSL offloader
Proxy protocol documentationThe description of the protocol by Willy, HAProxy developer: http://haproxy.1wt.eu/download/1.5/doc/proxy-protocol.txt
And a few articles speaking about the subject:
http://blog.haproxy.com/2012/06/30/efficient-smtp-relay-infrastructure-with-postfix-and-load-balancers/
http://blog.haproxy.com/2012/06/05/preserve-source-ip-address-despite-reverse-proxies/What do you think about it?
-
Hello Stolao,
Your events are very good.
I noticed some erros and I'm reporting it.
In Event_Management:OnJoinEvent
1 - When an event require items and the event is configured to store all the equips, the join command fails because the delitem is actually place after the storeall command. So when delitem runs, the inventory is empty.
2 - else if ( $@Cost[0] && $@Cost[1] )
If the event is configured to require an item, it will trigger the above condition.
But if the player have the item in the inventory, it will simply hang the screen.
Because the event require an item, ok, but the nested condition is if the player don't have the item, so, it will not trigger in the above situation and the options will hang.
The correct sould be:
else if ( $@Cost[0] && $@Cost[1] && (countitem($@Cost[0]) < $@Cost[1]) ) {
-
Purchasing from the control panel is not a good idea, because the shop ingame would not be refreshed on purchase.
I think that is a good idea. The vending would be stored in a table, when the sell is done, the item will be deleted from that table.The player shop in the game will fetch the items from that table and will refresh the purchase list in an interval of N seconds. If someone buy an item that is not in the vending table, the sell is cancelled.By this way will be possible to shop an item from the control panel.
New World RO
in Other Graphics Showcase
Posted
Fantastic!