I created a Powershell as well and my anti-virus maxed out real-time scanning, you'll probably need to make an exclusion for the powershell in antivirus and be more selective on your chat filters if you raid, otherwise your triggers will be very laggy using either solution. I also use filebeat to send my logs to an ELK stack and I have a bunch of raid dashboards with kibana, so I may have too many things reading the logfile too.
I have my EQ directory excluded from AV scans. Not the very safest solution, but security is always a trade off with convenience You might try pointing everything at the new file, just to make sure the hit to EQ is minimized. I'm curious about the lag/latency of your ELK/kibana solution. I'm not very familiar with them, but I've been toying with the idea of creating a client that would read log files from multiple characters and display selected output in a custom chat window overlay. I'd like to get things like raid and guild chat, maybe even all chat/log info, moved to a separate screen.
I was thinking of something like this as well, I would love to free up some space on my screen and move this text to other monitors. Is the Gina developer still around? I was looking for the source code but can't find it in the usual places.
Gima hasn't played in years and hasn't been actively developing it for some time but is maintaining it.
1 Patch broke both ingame audio triggers and external ones, we just didn't know about the external part until May started.
In 8 Days it will be working again! Can the devs let us know if the log change is going to be permanent? I would hate for a guy who doesn't even play the game to fix this just to have it break if things are changed back. It's already going to require a bit of effort to get everyone updated.
It depends on what you want to send to logstash. You can drop lines at the filebeat client and only send chat related text to logstash and it would be super fast. I collect everything non-chat related with no chat filters and have a bunch of different groks setup to tag the inputs, it's fast enough that it only has a few seconds of delay, but not fast enough to be reactive. Kibana has a 5 second minimum refresh anyways and that's usually about the delay I see. It's kinda of cool, you can change your time intervals to check out different raid encounters in the past without having to ingest a bunch of log files. The filebeat client runs as a service so I never have to worry about it, once I start logging files in EQ they automatically get sent and ingested. You can get the same and better results from Gamparse, I was just playing around and wanted to see what was possible. I created some visualizations and got mostly what I needed, but never came back around to figure out how to tag duplicate inputs from multiple clients. My ultimate goal was to try and make a web based parser that could take inputs from multiple people to get around the range logging issue. I use timelion to get a crude semi-realtime damage meter, so I can tell when damage is spiking etc. Just a sample, I have like 40 different visualizations created for all kinds of different points of data. Timelion(This has a legend, I just hid it) Heat Maps(Character casting what spells how many times) Damage Distribution(Hovering shows actual numbers and percentages)
This concept is absurdy awesome. Joining log files to get real parses, in almost-real time, sounds like a game changer...
I have been wanting to write a script that would take logs from several people, uploaded to a AWS VM, then remove values that are duplicate and then combine and display the overall numbers. This would give you a more complete picture of the over all damage from everyone in one parse.
Yep, that would be ideal, but with timestamps being based on the client computer, I think matching up duplicates would be very difficult, if not impossible. But even just parsing each file separately, and then only taking the highest values for each toon (from across all the logs) would give us a much better picture than we get now. (I don't think we ever see "over reporting", just "under reporting").
I was going to have the script sync to an NTP server, and as it captured the logs the time stamps could be synced to say US CST. It would basically do what the scripts are doing currently to fix gina but normalize across time zones. Than upload the files to a central location to be parsed. I just need some free time now
That might work if you're only working with your box toons, but if you wanted to combine logs from other people (like for a raid), then everybody's clocks might be a little bit different. Let's say my comp's clock is off from yours by 1 second. When I cast a spell, the message might appear in both of our logs, but since the timestamps don't match, how will you know whether it is a duplicate or simply a message that was only in my log and not yours (due to range)? A simple spell cast might be an oversimplification, but you can definitely see how the problem compounds with things like quad-hits and dots and such. The problem would also worsen if clocks were more out of sync. I think you could definitely combine all the logs into one file, but I think you will need to add information to each line, indicating which log it came from. I don't see any way to accurately remove dupes.
I think what you would want to do is just take the personal parse from each parser. You know your own parse is 99.9% accurate for you. So just send that. Of course, then you would need everyone to be doing it.
A guy in Return of the Exiled made a program that uploaded everyone's parse to a server. All players had a client version that would upload to the server, then the server would return a parse of everyones DPS. iirc he uploaded the client & server version to the guilds old website before i think retiring from EQ, and we never fully adopted it.
Yeah, if the client and server were done to work together, I'm sure the clock differences would be taken into account via NNTP or w/e. That's a great idea for GamParse, isn't it? What do you think, Beimeith?
I agree. If you only have 10 logs for a 50-man raid, 10 people would be nearly accurate. For each of the remaining people, I think you would want to only take their highest parse (out of the ten logs they appear in). And then be sure to tell everybody that if they want their DPS to be recorded as high as it can be, they need to submit their logs!
https://forums.daybreakgames.com/eq...date-bugs-4-18-18.248662/page-14#post-3662793 So, this means any change in the GINA code should take this into account. Unless they don't change it and we are forced to do without GINA until the May patch (edit: I realize now it will work beginning May 10). And no, changing your PC clock is not the answer. Too many important things depend on accurate dates. I recommend against changing your clock.
You all realize that if DBG fixes this, you all need to be thankful? As in give them no BS about other things for at least a 24 hour period after? In terms of a business, GINA has to be patched, not EQ.
Gina will work again on the 10th, the patch will happen mid May. So we do nothing and it will work next Thursday.
