Special Early QA testing - new Unity version

Debian is not very up to date.

[edit] Then again, neither is Ubuntu.

 

@Mishikal
The latest Debian uses Linux 4.9. This will not work with my Ryzen processor (needs 4.10 minimum).

 

I was mainly joking. I assume you were looking at debian testing, and not a fixed release (as those do get so very old so very fast).

 

The first 60 seconds I was concerned. I noticed a LOT of hard drive activity and the lag was ugly.

After the drive calmed down to 'normal'. I noticed the occasional lag when loading a scene. It wasn't horrible, but it was noticeable. Entering scenes afterwards, I didn't see any further excessive drive activity afterwards. It felt like a one time thing. Even exited and re-entered the game.

Compared to what we have going on with Unity on r59, it's not bad. Running 16.04 LTS Ubuntu. Latest set of patches. This could be doable. I'm curious what others see.

BTW I entered several scenes. A couple of towns. A few dungeons including the Rise during my testing.

 

Yes I agree. No ubuntu, I would go for manjaro (Arch Linux) bleeding edge as Fedora. OR just recompile OpenSSL on Fedora.
I will do a manual build today and see if it fixes it.

 

So recompiling openssl from source, did not help. And also I do not recommend it as it totally destroyed my fedora system
Oh well, time to reinstall my system and install Manjaro. I am done with Fedora lol.

 

A quick feedback from OsX Mojave (10.14.1) on the recent QA-release (Nov 6th): I played with default settings (=without "force-gl-core") and the game locks up at 98% scene load. It brings the Mac to a complete stop, then crashes.
Oddly, when you start on the overworld you get the usual hitch, but the game works the first 10-20 seconds, then quits rather ungracefully.

Also, after the crash the program "install command line developer tools" gets launched and tells me that the command "lldb" needs the command line developer tools. The last message also pops up with Mojave+R59 and without "-force-gl-core". High Sierra worked fine in R59 without "force-gl-core" (after the funky-gfx-glitch fix). It would be embarrassing if SOTA refuses to work on the latest Mac OsX.

Silberfarn

 

It is: -force-glcore

 

I'm currently running Ubuntu, and not experiencing the login issue. The hitching is pretty bad on start up, but smoothes out a bit after a while. The hitching seems about the same, maybe slightly worse, as Production. I just updated to Mesa 18.2.2, and am now getting graphical glitches sporadically. I'm looking into a different distro right now, but want to wait till that login bug is gone.

 

OK did some combat, crafting, tried a new basement. And ran around a few towns. This is running very smooth for me (Windows 10). Hitching is non-existant even crafting and combat. Running better than live. Seems OK for me?

User Specs:
OS: Windows 10 (10.0.0) 64bit
CPU: Intel(R) Xeon(R) CPU E5-1620 0 @ 3.60GHz (8) System RAM: 16303
GPU: NVIDIA GeForce GTX 1080 GPU RAM: 8079
SotA.QA.Win.64.760.Date.11.02.18
Area: POT_forest_metropolis_01_template/Dragomir Mori
Area Display Name: Dragomir Mori
Loc: (30.0, 38.0, 264.9)
Debug: UE9UX2ZvcmVzdF9tZXRyb3BvbGlzXzAxX3RlbXBsYXRlfERyYWdvbWlyIE1vcml8KDI5Ljk2NSwgMzcuOTk0LCAyNjQuODgpfCgwLCAtMC43MDgsIDAsIDAuNzA3KXwyNzUuMTIzNnwyMy4xNjQ0MnwxMi43NjQ=

 

OpenSSL 1.1.1 11 Sep 2018 on Manjaro (Arch) GNU/Linux the QA client works, no login issue.

 

Yeah. Pretty much what I experienced.

Seems the hard drive was getting pounded while that was happening. I was relieved when it disappeared after a couple minutes

 

Installed Manjaro last night. I'll give it another go tonight.

 

For the linux web issue, are there any more specific errors showing up in the game client log? Thanks for all the research and testing yall have been doing! I found a thread on the Unity forums that looks like it could be related to the issue we are having, and posted a question to ask for some details on a possible fix.

https://forum.unity.com/threads/ubuntu-headless-build-tls-handshake-fails.546704/

 

Hehe...you aren't sure?

Mine didn't even come close to disappearing. But it did get a little better than when it started. As I said, "The hitching seems about the same, maybe slightly worse, as Production."

 

Works with Manjaro. There's still weeds growing on my keep island lot and I get a weird tearing in the middle of the screen while moving.

[edit] Ok, the tearing is because it's using BLIT and not FLIP. It was using FLIP on Fedora, so that's a bit disappointing.

 

Yes that seems to be the problem, although I did try to reinstall curl and curl development libraries. This did not fix the problem.
The log was not showing any more errors as I posted earlier. Also the wireshark dumps I studied did not show any weird errors, it was just like the client did not initiate the TCP - TLS communications.
This is all the info I can give you.

 

check your nvidia settings, the vsync options should fix the FLIP/BLIT problem.

 

vsync is on and allow flipping is on.

 

You it on the production client also or only QA?