RawTech

Assuming your proxifier targets are appropriately setup yes. Though it sounds like *.runescape.* may not valid here. Just launched the client on my machine and the official osrs client process looks to be called JagexLauncher.exe perhaps try targeting this instead? I would stick with the native proxy configuration on the osbot client here to simplify things.

Not hugely. Some form of X server is required to launch the client heedlessly it seems. I have changed the entry point to use debug mode to get around the tailing of the logs from the filesystem.

Proxifier will still send your *.runescape.* traffic to your proxy, just via nordvpn. You can verify this by creating a proxifier rule for a web browser and then visiting a site like http://ifconfig.so/ Try it twice with nordvpn enabled and disabled, to verify your results.

69.420 credit please

Grafana is a great visualisation tool, it supports mysql as a data source, and above all, is free and open source. If you know your way around SQL you can create some very data rich visualisations. Obviously you'll need to capture all of this data in a meaningful way.

No question.. Git is the best practice. If you're creating an entirely new script with no intention to merge the 2 together then make a new git repo. If you're creating a new version of an existing script, like a beta or alpha version, then make a new branch within the same repo. Googling git best practices you get a lot of branching strategies, commit tagging and naming conventions etc. But those are only really useful when you run a large software shop with multiple developers. If you're working alone, just find out what works for you. Github offer free private repos now and if you're looking to roll with your own selfhosted gitlab is really nice but potentially a bit overkill for a few repos.

I typically get one headless osbot client per GB of ram on linux systems. CPU is not really a bottleneck providing the clients have the lowcpu flag. Not using webwalking reduces the ram requirements by half, so 500MB per client. All from my own experience using custom scripts.

"Fixed trading not offering items in resizable mode." Nice one, cheers Patrick

In your module configuration within the project structure make sure that the scope for osbot is set to "provided". This will exclude it from the compiled jar.

Epic, thanks Patrick.

That's worked a treat. Thanks @Token

Hi, When running the client in resizable mode, is there any way to programmatically set the client dimensions, either through the script itself or the CLI arguments? I'm wanting to set the client viewport to 720p or 1080p automatically when I launch the client/script. Thanks

Thanks for all the hard work @Patrick !

Restarting the client doesn't seem to help. worlds.hop() is just returning false in the cases where it doesn't work. The client opens the world switcher tab but doesn't attempt to scroll to the desired world, if the desired world is in view, it will click on it.

For certain scripts you need to hop to a specific world. I guess this would be a workaround if you didn't need to do that. I just noticed this same issue when I got home from work today, hopefully its something the developers can resolve in good time. @Patrick edit: From my brief testing, it doesn't look like the problem specifically affects certain worlds, seems to be broken across the board.

Great project! How long do your accounts typically last if you're flagging them at 12 hours runtime? Enough time to mule a profit I assume?

Once a new version of the client is released, you are unable to use any previous ones, so yes you should update your libraries. Now technically you can use a script that has been compiled with an older version of the client but there may have been some breaking method signature changes, so best to update it.

Hello, I have a script that takes periodic screenshots for debugging purposes. When running in nointerface mode the screenshots seems to be showing a banner ad sometimes covering useful information I would like to see in the chat box. Regular, non nointerface mode and there are no advertisements as expected. I would have thought this would be expected behaviour for free accounts but not VIP+? Can someone clear this up if its intentional? Thanks

I've gone with the following solution for now.. Created a custom entrypoint script which I later intend to add support for environment variables to pass stuff in dynamically. Dockerfile: FROM java:8 RUN mkdir -p /root/OSBot/Data RUN mkdir -p /root/OSBot/Scripts RUN wget https://osbot.org/mvc/get -O /root/OSBot/osbot.jar COPY map.bin /root/OSBot/Data COPY script.jar /root/OSBot/Scripts COPY entrypoint.sh / ENTRYPOINT ["/entrypoint.sh"] entrypoint.sh https://pastebin.com/raw/nJ4qt3dz (Posted on pastebin as cloudflare thinks its malicious and wont let me submit) It effectively forks by tailing the stdout (/prod/<pid>/fd/1) of the new java process, the only problem with this approach is that the logs aren't properly passed to the docker daemon and the kill commands are not handled gracefully, but it works. This keeps the image size down to 700MB rather than 2GB that consol/centos-xfce-vnc uses. If I ever find a more elegant solution I will be sure to update this thread. Thanks for your input.

I considered forking, but I would like to avoid it if possible since it plays havoc with stdout and enabling docker to detect zombie/killed processes. But if it works, then it works. And docker because I use it a lot at work and I appreciate the portability, can run it anywhere without leaving a mess on the file system etc. Plus it makes proxying easier if you have a docker "bot" network with a proxy or two for outbound traffic, gives me peace of mind that nothing can escape out of the systems native internet connection & potentially leak IP addresses. I'll post something to the bug reports forum Re the x server.

Hello, ( first post on the forums \o/ ) Long story short, I'm trying to run osbot in a docker container. When the jar is ran, the launcher starts a new process which is the bot client itself. This causes the container to exit since the original process stopped. I can get around this by running with the -debug flag, which doesn't launch a new process, however this behaviour is not desirable (would rather not run with debug on). Additionally I'm fairly confident these 2 arguments aren't supposed to work together since I get an x server warning when both are passed (-script nointerface & -debug 5005 output at bottom of post) Dockerfile: FROM java:8 RUN mkdir -p /root/OSBot/Data RUN mkdir -p /root/OSBot/Scripts RUN wget https://osbot.org/mvc/get -O /root/OSBot/osbot.jar COPY map.bin /root/OSBot/Data COPY script.jar /root/OSBot/Scripts ENTRYPOINT ["java", "-jar", "/root/OSBot/osbot.jar", "-allow", "nointerface,lowcpu", "-login", "REDACTED"] Output when running: [root@docker-standalone osbot]# docker run --rm osbot -bot REDACTED -script Script:1 Starting OSBot with no interfaces, inherits lowresource Starting OSBot in low cpu mode Starting local script with name: Script [DEBUG][11/14 06:00:39 PM]: Injected 2 field list filters [DEBUG][11/14 06:00:39 PM]: Injected 2 field filters [DEBUG][11/14 06:00:39 PM]: Injected 2 method list filters [DEBUG][11/14 06:00:39 PM]: Injected 3 method filters Starting OSBot in low resource mode [root@docker-standalone osbot]# As you can see it immediately exits once the client attempts to start. Ideally I would not like to put in a hack to keep the original process running as I would like the container to exit cleanly once my script calls stop(). Bonus points if I can get a pointer to map.zip / map.bin on an osbot webserver somewhere such I can automatically fetch the latest in the Dockerfile. Any help appreciated! Cheers Snippet when running -debug and -allow nointerface