NWNX Speech plugin!

[beeatch]

Can anyone explain to me how the speech plugins work?
Like can a player on server a tell/party with a player on server b?
If so how does player a know about player b?
Can this system be used on say 4 mods running on three LINUX servers all on different ports behind a router?

Lastly is there a step by step somewhere on how to install configure it for LINUX? I can't seem to piece it all together but if it does what it sounds like then I really want to try it.

Thanks
Aaron

[dguntner]

I'd love to help ya, but my server is running Linux.

--Dave

[Dazzle]

Woo. Found a small bug. CListener::getType() seems to be missing from the CNWNXSpeech::OnRequest() function.

[Lanthar D'Alton]

My alpha version for TSS 3.0 is located at the following link:

http://lanthar.dyndsl.com/downloads/lds_talus_speech_30a_win.rar

Jeroen, feel free to grab that file and put it on your nwnx downloads site.

This version includes support for /o emotes and commands and stuff. The list of commands in the readme is out of date, but hey, it's an alpha... The list of commands in server is available by saying *help* in talk as usual... or you could say /o help with this version.

Note that I will suggest you first launch the exe listener to find which interface you should use, and then paste the dll into your nwn dir only afterwords.

Some tips:
-Speech floods by players will spike your cpu from the normal 0 to 2% or the setup module to about 6% as I limit the list control in the dialog to 200 lines... and the deletes are slow. I will be making the limit configurable in a later version, and I will probably replace the list control with a richedit window of color coded text instead, since deletes should be much faster. Minimizing the eavesdrop dialog will completely prevent this spiking from occuring.
-Closing the eavesdrop dialog is completely safe. If you close it, you cannot get it back without reloading your server... but the listener will continue to run in the background whether the dialog is present or not.
-Normal speech uses about 1.5 milliseconds of cpu time from saying it to the server parsing it. This include the time used in actually parsing what was said inside the server.
-The listener's dll cpu time was not effectively measured... but during floods with the window minimized, the cpu usage was 0 to 2% for nwserver.exe running the speech setup module (mostly on the 0% for that time, so that 2% could have been other stuff happening).

If you already have TSS 2.0 installed, and have added your own commands, back them up before importing 3.0!

Read the readme file before you do anything!

[dguntner]

Does this include the DMFI 1.07 merge I sent you?

--Dave

[Lanthar D'Alton]

Actually, I seem to have never received it. You sure you sent it to the right address? Also did you rar it? The account I gave shows 6 mb free so... *shrug*

It is certainly worth noting that the above link is NOT a final erf or release, though forthcoming releases will overlay that one fully, so it should be safe to import and start using as long as you don't start merging your own older commands into it.

lanthardalton@icqmail.com

-Lanthar

[dguntner]

Well, that's weird. I watched my mail log when I sent it, and icqmail.com said accepted for delivery. I zipped it, BTW.

I've just sent it again. From my logs:

Feb 22 15:39:37 janet postfix/smtp[29967]: F2EDEFA1F1:
to=<lanthardalton@icqmail.com>,
relay=mx1.icq.mail2world.com[216.163.188.54], delay=11, status=sent
(250 Mail accepted)

As you can see there, it *claims* the message was accepted. Again. Let me know if you get it this time.

--Dave

[Lanthar D'Alton]

That rar included the wrong version of the dll... I've uploaded an updated copy. Stupid dev studio crashed and never saved my changed output directory... hate that. Anyway, the copy there now has the right version of the dll, with date/time: 2/22/2005 6:17PM

http://lanthar.dyndsl.com/downloads/lds_talus_speech_30a_win.rar

-Lanthar

[Lanthar D'Alton]

1: dguntner : got it this time.
2: beeatch : If you are using the version I posted a link to just now, I'm not sure the dll is multi-server safe now... I'll have to check on that. the exe is... to use that copy though you need to edit lds_spch_runonce and change the line at the top that says:
#include "lds_speech_dll"
to
#include "lds_speech_db"

Then rebuild the test module. The same holds true of the erf after importing, in case you are using the erf imported stuff with your mods.

Then just launch the Talus Listener.exe on each server you are running.

-Lanthar

[Lanthar D'Alton]

Sigh. Okay, found an exception due to a change in the code that checks for config file which I forgot to recheck... I've yanked the file off server again. I've fixed that problem, but while I'm at it I'm going to make the dll handle multiple servers.

-Lanthar

[beeatch]

Lanthar,
I'm not using any version right now. I'm not sure what to download =(
If this will let me use tells and party chat across servers then I'm all for installing and testing it. I'm just not positive what all it brings to me. How can an a person on one server see other people on other servers?

Can it run on a Linux server with multiple instances of nwn running?

The reply you just made said to launch the Talus Listenter.exe cept that won't work cause all my servers are Linux. I take it your version is strictly windows?

I thought I read something about a linux speach port. Too many threads...

[Lanthar D'Alton]

Okay, link is rehooked, there is a new copy that should work great for multi-server machines too now. There is a related change inside the lds_speech_dll and lds_speech_db files, which means if you downloaded earlier and imported the erf, you need to download this one and reimport and recompile. The module included has also been updated.

Most importantly, if you tried to run this with earlier copy, you need to delete nwnx_talus.ini from your nwn directory. This is rather important, as that file is invalid and will cause further problems for you if you don't delete it this first time with the new copy of the dll.

Sorry for all the trouble here. Guess that's what I get for testing with a debug mode copy and then releasing the release mode copy.

-Lanthar

[Mr-XXS]

Hello, it doesn't work on my server.
When i use the old version of the nwnx-speech plugin, it's quite perfect, and the Talus .exe works too. But when i want to use the nwnx-speech 3.0 dll plugin, the config popup appears, i put the same interface number and some log file name... And then, nothing happens, nothing is logged, nothing appears on the listener window.

[Lanthar D'Alton]

Important Notes:
-don't check the use database check... not sure that works right
-ensure that you import that updated erf. It has several important changes that must be in your module.
-Most importantly, the module name must be sent to the dll by your module load script... Have a look at the Talus Speech Setup module's script to see what I mean. There should be a function call named lds_ModName() in lds_speech_db and lds_speech_dll (interchangeable includes). This must be called when your module loads, or else the dll won't have the name of the module it is attached to. (I think I'll change it so it grabs all if it doesn't have the module name).
-If you have troubles, instant message me. icq is the best choice as it queues messages if I'm offline.

-Lanthar

[Lanthar D'Alton]

Anyone care to report in about the alpha?

-Lanthar