MQ2EQBC Question

Now it's my turn to be defensive. The new version works just fine, you just need to configure it correctly. Here is the lazy man's way to configure it to work with multiple computers
Code:
MQ2EQBCS.exe -p 2112 -i 0.0.0.0
. Have fun!
Well, to be fair... once again I saw it wig out, like I mentioned to you a while back. Local server, 2 local clients, both AFK in guild lobby (nothing special, no netbots/netheal, no macros, etc.). EQBC Interface running at 5 sec update interval for stats (yours). After around a day or so, all clients got disconnect and server was spamming constant data to the window. I had to end up killing it with TaskMan & reconnect the clients.

Not saying it's not good work, don't get me wrong. I just mean that we probably need to go the route of a more intensive grouped beta test here with volunteers, instead of pushing it live now.

htw
 
Can't we just make this new Python version of EQBCS detect the network interface address the same as the old version? I admit I never used the built in MMOLoader version, all I ever did was run MQ2EQBCS.exe, and then connect my computers to the local network address. I never messed with any ini's or ran it with additional command lines. I actually just have the exe on my desktop that I run, it never creates any ini. It was pretty straight forward, but I knew it's user friendliness could be improved.

From my eyes, to improve EQBCS it needs console commands showing what IP's it's listening to, maybe allow it to issue commands to all connected computers, and maybe a help readme in the actual console. It needs to be more user-friendly and have more features than the old version.

I do have to give major props to Xeniaz for all the work he has put into re-writing it in Python, it must of been one hell of a job.
 
It doesn't really detect. The old one listens by default on all interfaces. The new one was defaulting to 127.0.0.1 (localhost). I just changed our copy to default to 0.0.0.0 (all) without any -i provided, so that will match the old one. It does not use an INI except for if you use a login token (password), using MQ2EQBCS.ini, as in:

[Settings]
Password=MyPassword

So if I were to push a new one from our svn now, it would act the same way now as the old one, for port/interface defaults and behavior.

To me we have more immediate issues to address:

1) Why can person A use old done listening on e.g. 10.0.0.10 and it works fine for his other local computers, but the new one does not? (assuming it listens on the same port as old one). We need to completely test this, with connecting from other computers on the LAN, to ones outside your WAN with PFW, and make sure multiple configurations work.

2) Why does it eventually choke and get jacked up. I have no idea as I haven't tried looking into why it would do this, xeniaz would be more familiar with the boost libraries and python code and why this happens, once he is able to duplicate it.

3) Even if logging to file, have it display what interface and port it's on when it starts up (to be fair to xeniaz, the old one wouldn't do it either if logging, you had to pick console or log, not both). (EDIT: It already did this via the default logging.ini, as far as the port. I added the interface display as well. - htw)

htw
 
Last edited:
Here is an example of what it logs/prints if ran from console now, with no switches (just as mq2eqbcs.exe):

[2014-05-03 03:26:23] EQBCS version 1.3.0 by Xeniaz (MMOBugs.com)
[2014-05-03 03:26:23] Waiting for connections on all interfaces port 2112...

Here is an example of me providing an interface (mq2eqbcs -i 10.0.0.2):

[2014-05-03 03:31:48] EQBCS version 1.3.0 by Xeniaz (MMOBugs.com)
[2014-05-03 03:31:48] Waiting for connections on interface 10.0.0.2 port 2112...

And providing a port and password (mq2eqbcs -p 10000 -s MyPassword):

[2014-05-03 03:32:51] EQBCS version 1.3.0 by Xeniaz (MMOBugs.com)
[2014-05-03 03:32:51] Waiting for connections on all interfaces port 10000...


I tested LAN connections, and all is fine. I didn't test outside via NAT, but there is no reason at all there would be any issue as long as firewall and PFW is set up properly, since LAN works fine.

This covers all what we have been talking about, except:

1) Why it sometimes eventually wigs out, as I mentioned in a previous post.


Anyone wants to try it out, I have attached it here:
 
To add to this, I wanted to mention why this is even going on, for those who wonder.

Historically (for years), MQ2EQBCS tends to choke eventually and stop working properly. The more data you would hand to it (such as more clients, or clients + netbots), the quicker that would happen.

We wanted to move away from the old win32 winsock'ish nature of that and provide something more robust, and potentially even adding more back to things like netbots (which was partially data raped to help with the issue).

The buffer increase alleviates some of the oddness, but not all of the issue. Xeniaz took it upon himself to volunteer to do this project and get everyone a meaner, better performing, MQ2EQBCS.

Anyway, that's the purpose of this. There was no intention to put more work on anyone, and in fact, that's why we are even talking here, to get it to where it is 100% transparent for anyone using the old one, except 1 difference: It works better.

Thanks Xeniaz for taking the time after all these years to work on this old thing.

htw
 
Htw, I am getting error running your attached exe.

----------
Traceback (most recent call last):
File "EQBCS.py", line 318, in <module>
File "EQBCS.py", line 304, in main
File "logging\config.pyo", line 70, in fileConfig
File "logging\config.pyo", line 106, in _create_formatters
File "ConfigParser.pyo", line 607, in get
ConfigParser.NoSectionError: No section: 'formatters'
 
Htw, I am getting error running your attached exe.

----------
Traceback (most recent call last):
File "EQBCS.py", line 318, in <module>
File "EQBCS.py", line 304, in main
File "logging\config.pyo", line 70, in fileConfig
File "logging\config.pyo", line 106, in _create_formatters
File "ConfigParser.pyo", line 607, in get
ConfigParser.NoSectionError: No section: 'formatters'
That's what it does when it doesn't have logging.ini, or something is wrong with logging.ini - so far anyway. hahaha. logging.ini needs to be in the same location as the mq2eqbcs.exe

htw
 
Any way for it to generate the ini file if one doesnt exist?
 
Any way for it to generate the ini file if one doesnt exist?
Well, I mentioned it in Irc last night, we'll see what is available. I thought about using defaults if one doesn't exist (basically, console).

Htw
 
I will be adding in code to generate the ini file if it doesn't already exist. I may be a little slow getting these changes done though. Haven't had a day off of work in over a week, and not sure when my next one will be. Just wanted to post back that I see the suggestions and I will work on implementing them asap. Also, I do welcome any feedback / suggestions people out there may have.
 
Here is new DLL with change. It will be in next release.

this eqbc is no good anymore, on test, it left me no choice but to replace it , but the new one after a cpl characters are joined it starts disconnecting. is there something i can do with my options to change it , or is there something wrong with the newest (test) release of eqbc now?

will be a min , i have to redo all my eq stuff , but ill report if a complete reinstall changes anything
 
Last edited:
this eqbc is no good anymore, on test, it left me no choice but to replace it , but the new one after a cpl characters are joined it starts disconnecting. is there something i can do with my options to change it , or is there something wrong with the newest (test) release of eqbc now?

will be a min , i have to redo all my eq stuff , but ill report if a complete reinstall changes anything
That change wasn't added to test. I just added it, and uploaded to server.

You can grab and drop in with Download MQ2EQBCS for Test Server

htw
 
thank you so much, i wish i had just waited a little bit longer before i went all stupid and screwed up all my stuff like a dumb dumb.
 
if it possible to make the live version of eqbc and eqbc interface compatible with test please? eqbcinterface keeps on disconnecting with the version available in test compile, but live version crashes the test loader
 
if it possible to make the live version of eqbc and eqbc interface compatible with test please? eqbcinterface keeps on disconnecting with the version available in test compile, but live version crashes the test loader
Sure, I'll get to it tonight or tomorrow night. You should be able to (for now) grab the 2 exe's from the live release zip, and drop them in, and don't patch them with loader on test, and the plugin from test will work fine with those.

htw
 
I noticed in the MMOBugs patch there are now two EQBCS files. A "Classic" and a "Xeniaz" version.

Could someone explain the difference between these two and which one is recommended by the devs?
 
I noticed in the MMOBugs patch there are now two EQBCS files. A "Classic" and a "Xeniaz" version.

Could someone explain the difference between these two and which one is recommended by the devs?

Classic is the old version, small file size, has always been stable for me, so it's what I have been using. It did have a disconnect problem that people were having trouble with from it, but we found out it was a plugin issue and should be resolved.

Xeniaz is Xeniaz new version he has been working on with new features, a little larger in size and written in Python. Some people get it flagged as a virus due to how it's packed. I had a little problem with this lagging my network every 30 seconds when it pinged connected users if you had a lot of characters connected.

Both should do the same thing, it's down to personal preference.
 
Also, if you use EQBC Interface, make sure if you use the old eqbc server, that you use the old eqbc interface (by ieatacid), and if using xeniaz eqbc server, use his new eqbc interface.

htw
 
new memory and hd are here i will install fresh os and eq/mq2 and will try out eqbcs and see if i get random disconnects. Never heard back HTW, the random disconnect was from the 1.4.2 version (xeniaz's) but if i ran an older one (i have a 1.2.1 version) no random disconnects, that was before this patch and mq2 update, once i get home i will test out the compile and see if i disconnect. Thank you both for looking at this issue, i know that all of us that bot many accounts at once thank you for your continued support of plugins/compiles/macros/uifiles again thank you.