Minkata Update Build #186 2020.07.01

[rarified]

Minkata now has a new client program version, plus some re-built Python files.

The updates to the client are fixes to (hopefully) address the problem of crashing when leaving the game client.

Please try going into Minkata, playing around a little bit, and then exiting, perhaps from locations you left and had the problem before.

Also, try the /startlog command to begin a chat log to see if the time/date has returned to the beginning of each line.

Thanks
_R

[rarified]

Alas, the /startlog chat log file is still missing the timestamps. At least for me.

_R

[Treehugger]

Just played around a bit, around five exits and and no crash logs so far, well done!

[dgelessus]

I got this crash when exiting the game:

Code: Select all

UruLive.Live.1.918 - Minkata.186 - External.Release
Exception type: Access violation
Call stack (14 levels):
 0x76CA6359
  0x00F1C139
   0x00EEC7AE
    0x00EEF8DE
     0x00EF0FF1
      0x00EF043B
       0x00D9A2A2
        0x00CA8890
         0x00DC41A1
          0x00DC3D40
           0x00D97E56
            0x00EEBE7E
             0x00EEA803
              0x00D97FB3

I was testing on Windows 10 1909 (x64), on a laptop with Intel HD Graphics, in fullscreen mode. In the game I roughly did this:

• started in Relto
• linked to Nexus
• linked to Watcher's Pub, tested chat log (no timestamps, unsurprisingly)
• linked to Chiso Preniv
• linked to Veelay Tsavahn, walked around the entire outer path, up the inner stairs, read all the books, walked back down
• linked back to Chiso Preniv using the book in Veelay Tsavahn
• linked to new GoMe pub (there was one other person there for a short while), played around with the kitchen sink and some of the doors
• linked back to Relto
• exited the game

No idea what part of that might be relevant.

Surprisingly, when I tried similar steps under Wine (5.0-3ubuntu1) running on Linux (Kubuntu 20.04) on the same machine, I wasn't able to reproduce the crash, even though this Wine setup often gave me crashes on exit before.

Unrelated to the crash, I noticed that with this build and the last one (and possibly earlier ones, not sure) the Minkata launcher's update check takes much longer than usual. Every time I run the launcher, I get two "Checking for updates..." progress bars - the first one only takes a couple of seconds, and the second one takes a minute or two and a lot of hard drive activity. With MOULa and older Minkata versions, the launcher normally only did the first (quick) update check, and the second check was only done when any files actually needed an update. The latest Minkata versions always do both update checks, even when my installation is already fully up-to-date and no new files are downloaded after the update check.

[rarified]

Thanks for the crash report, I'll look at it shortly.

I've heard about the dual update scanning and need to look at the logs myself to try to understand what's happening. It may be happening to me, but being local to the server, the short update scan might seem to be a flicker on the screen rather than something I'd notice. I'm working on a new build now, so it will be a little while.

_R

[Hoikas]

The patcher works in two stages. First, it updates itself. This is the "fast" stage. Next, it updates the game itself... This is the slow part. The patcher in use during the slow stage is the original Cyan patcher, and it's pretty slow. There is a mitigation that can be done though... Make sure the "ThinExternal" manifest contains only client files. If I remember correctly, the pacher in use here will request the full "External" manifest if any file listed in "ThinExternal" changes - hence the old "delete TOS.txt" trick. If the "ThinExternal" manifest contains all ~3gigs of the game, that will take awhile to process.

[rarified]

Actually, three stages now, with the dependency installer update.

But yes, ThinExternal is a significant subset of the External manifest. I have included the fan ages under development in ThinExternal as they are likely to change frequently and I felt it would be better in testing to pre-load rather defer loading to the inter-age updater for those files.

Looking at the client patcher logs will reveal why it always falls into a full External scan, I hope.

_R

[dgelessus]

Looking at the client patcher logs will reveal why it always falls into a full External scan, I hope.

This is the UruLive*.Live.log that is generated when I start the launcher, let it check for updates, then quit immediately once I get to the login screen. Doesn't seem all that useful unfortunately - it only shows that it's updating the launcher itself and the dependencies. After that it says nothing for two minutes while it checks for updates for everything.

Code: Select all

Log Opened
07/02/20 10:33:51 [UruPlayer] Info Patcher: UruLive.Live.1.918 - Minkata.186 - External.Release
07/02/20 10:33:52 [UruPlayer] Debug plSelfPatcher::IIssueManifestRequest: Issuing manifest request 'ExternalPatcher'.
07/02/20 10:33:52 [UruPlayer] Debug plSelfPatcher::IRun: Worker thread is now blocked on 'ExternalPatcher'.
07/02/20 10:33:53 [UruPlayer] Debug plSelfPatcher::IEnqueueFile: Enqueueing hash check of 'Client\External\UruLauncher.exe.gz'
07/02/20 10:33:53 [UruPlayer] Debug plSelfPatcher::IIssueManifestRequest: Issuing manifest request 'DependencyPatcher'.
07/02/20 10:33:53 [UruPlayer] Debug plSelfPatcher::IRun: Worker thread is now blocked on 'DependencyPatcher'.
07/02/20 10:33:54 [UruPlayer] Debug plSelfPatcher::IEnqueueFile: Enqueueing hash check of 'Client\External\oalinst.exe.gz'
07/02/20 10:33:54 [UruPlayer] Debug plSelfPatcher::IEnqueueFile: Enqueueing hash check of 'Client\External\vcredist_x86.exe.gz'
07/02/20 10:33:54 [UruPlayer] Debug plSelfPatcher::IEnqueueFile: Enqueueing hash check of 'Client\External\dxwebsetup.exe.gz'
07/02/20 10:33:54 [UruPlayer] Debug plSelfPatcher::IEnqueueFile: Enqueueing hash check of 'Client\External\PhysX_Setup.exe.gz'
07/02/20 10:33:54 [UruPlayer] Debug plSelfPatcher::ICheckAndRequest: File 'UruLauncher.exe' appears to be up-to-date.
07/02/20 10:33:54 [UruPlayer] Debug plSelfPatcher::ICheckAndRequest: File 'oalinst.exe' appears to be up-to-date.
07/02/20 10:33:55 [UruPlayer] Debug plSelfPatcher::ICheckAndRequest: File 'vcredist_x86.exe' appears to be up-to-date.
07/02/20 10:33:55 [UruPlayer] Debug plSelfPatcher::ICheckAndRequest: File 'dxwebsetup.exe' appears to be up-to-date.
07/02/20 10:33:56 [UruPlayer] Debug plSelfPatcher::ICheckAndRequest: File 'PhysX_Setup.exe' appears to be up-to-date.
07/02/20 10:33:56 [UruPlayer] Debug plSelfPatcher::IRun: No work in queue, exiting.
Log Opened
07/02/20 10:36:20 [UruLive] Info Client: UruLive.Live.1.918 - Minkata.186 - External.Release

None of the other log files are modified if I only run the launcher (not even patcher.0.log, strangely).

[hhhenry]

For months now, whenever I login, at the top I get:

<!--#set var="TITLE" value - "Object not found"
--><!--#include virtual -"include/top.html" -->

UruLive.Live.1.918 ...etc
No bad behavior, just the message

[Mac_Fife]

Yes, it looks like the status message file is missing.