Register Register

Author Topic: Weird Issues with 49.7 (and maybe stable 48 too) UPDATE with another issue...  (Read 413 times)

Burning Chrome

  • Lieutenant
  • *
  • Posts: 848
  • One of the Vocal Minority
I recently installed 49.7 and last night played a match with a friend.

FWLM (Marik Militia) units are tasked with punching a hole in LCAF lines so that and allied (Oriente Hussars) Field Commander can escape the Lyrans encircling and closing in on command unit.

The match consisted of level 1 units with the exception of a Mobile Command Post (MCP) and the following "players":

IS - Host:  I open an instance of 49.7 select the map and planetary conditions.  I also save games from this instance.  Custom map 33 x 51, with some restrictive terrain/ choke points but a clear path with road that slightly winds its way through the center of the map.  Weather was light rain, calm and set during daytime.

NME - Connect as a player when creating the game/campaign, later restart as a bot after initial units placed on board.  This instance controlled all 10 enemy (NME-get it?  :D) mechs.  Some were on map at the beginning of the map approaching from the north.  Others entered during following turns, most from the north, some from NW/NE.  "Nearest" is the setting for their retreat side of the map.  This bot is aggressive using custom settings but the units would retreat upon massive or combat effectiveness reducing damage was received.

NME2 - Connect as a player when creating the game/campaign, later restart as a bot after all units placed on board, many of which are hidden units.  This instance controlled all NME (14) vehicles and (10) infantry - several light tanks, APCs, motorized infantry and standard infantry platoons.  These units were placed at the start of the game, the setting was "any" and their retreat side of the map was also "Nearest".  This bot is also aggressive but more wary of damage taken.

ALLY -  Connect as a player when creating the game/campaign, restart and set as a bot on turn 1.  These 13 allied units are all vehicles moving 6/9, including APCs, Galleons, and Strikers escorting a MCP.  This bot uses the default behavior settings but it set to escape the board on the north edge.  These units will enter the south edge of the map on Turn 5. The bot chose to place most of the units on the center right side of the map in a good cluster with a few scattered of the far right to attack some previously detected NME units that were damage and left crippled.  Once these stragglers were dealt with, these units made excellent use of limited terrain options to rejoin the main column.

HOTEL - Me controlling a lance of 4 Mechs (Griffin, Hermes II (Mercury), Shadow Hawk, Ostsol in that order unless next unit was used during the movement phase to gain an advantage.  South edge of the board on turn 1.

PAPA - Friend controlling a lance of 4 Mechs (Trebuchet, Vulcan, Wolverine-M, Thunderbolt) in that order unless next unit was used during the movement phase to gain an advantage. South edge of the board on turn 1.

This game is similar to one I created on Stable 48, but we were unable to finish it due the save file becoming corrupted somehow.

Issue/Bug 1 - Firing phase and physical attack phase is set to simultaneous.  There were at least 3 times during the fire phase where I was unable to fire the Ostsol's weapons despite having LOS and targets available.  I usually go "in order/down the list of mechs" to fire weapons.  Several times, all 4 of my mechs had targets.  I'd fire the Griffin's weps first, then the Hermes II's, and then after firing all the Shadow Hawk's weapons or just those in range and selecting "done", my firing phase ended.  The Ostsol didn't get a chance to fire despite having a legitimate target and no issues/conditions preventing it from firing weapons.  This also happened during the physical attack phase were the Ostsol wasn't given the option to stomp infantry, but the Hermes II was able to kick a tank.  My friend believes this happened to his Thunderbolt as well at least once.  The issue seemed to plague the last unit in each or our lances.  If I went out of order and fired the Ostsol first, there was no issue with the remaining mechs missing their firing phase.  No settings like "skip ineligible to fire" were enabled.

Issue/Bug 2 - At least 1 instance where the Ostsol's sensor quirk failed to detect a hidden vehicle despite being in range.

Issue/Bug 3 - both the stable 48 and dev 49 games locked up about the same time.  Ally would detect a hidden NME2 unit at some point during its movement phase and then it would be Ally's turn to move, but the bot would take no further action.  Sometimes restarting Ally would clear the issue and the game would continue for a bit, but then the bot would lock hard.  Restarting the entire game from either the autosave of specific save before the restart resulted in "Error unable to load game file."

Pointless Gripe - both versions of these games had Ally with a mix of 3/3, 3/4 and 4/4 P/G with the human controlled lances with 3/3 and some pilot abilities like sniper, natural gunnery, energy spec or missile spec.  The NME mechs were 3/3, 3/4, 4/4, 4/5 with the heavier units having most of the better pilots but not all.  NME2 had a few 3/4, but was mostly 4/4, 4/5 and a few oddball 3/5s or 4/6s.  There were so many bot favored rounds when it came to RNG. 

Bots need several 10s?  No problem, have an S-load of 10s, 11s and 12s.

Humans need 4s?  HAH! GET BENT! Nothing but 2s and 3s consistently. 

I know its RNG and it's been beaten to death...but there were more rounds of that utter BS than not unfortunately...enough for my even my normally mellow/laid-back friend to be like, "Seriously?  Again!?! Did you mess with the settings and give the bots and modifier!?!  I needed 3s and still missed...again!?!"

In the nearly 2 decades we've played Megamek, we've noted some serious shenanigans revolving around GRF-1S Griffins.  ???   :)

Thoughts?  Similar experiences?

« Last Edit: 27 May 2022, 13:55:10 by Burning Chrome »
"Matchup of the century: desire to play vs. resentment"

dgorsman

  • Captain
  • *
  • Posts: 1749
Re: Weird Issues with 49.7 (and maybe stable 48 too)
« Reply #1 on: 23 May 2022, 10:24:04 »
I believe there are issues with simultaneous play, at least with Princess involved.  There's a pull request to remove that feature on the tracker.

I just put the rolls down to horrible luck, with any remaining target number issues to a bit of streaky-ness.  I wait a few hours to a day or two and things appear (emphasis here) to be more even, and I'm a little less PO'd.  Try to play longer games i.e. 30+ rounds, and include the fatigue modifier (adds +1 penalty to all shots and piloting checks after round 14 or so).  That tends to even things out more.

I don't think there are any sensor quirks which emulate active probes, only the Eagle Eyes SPA.  In theory the quirk just affects sensor ranges when double blind and sensor contact rules are in play, although I will admit I've not seen any difference in the shown numbers.  It may be affecting the values internally, or not at all, as the process is (thankfully) heavily automated.

Princess has some 'interesting' ideas around certain Mechs.  With Hellhound/Conjurers and Hunchback IICs, I've seen it run up the heat through shutdown levels on a regular basis with those two designs.  Favorite time was it opening up full-ultra, both cannons, along with the lasers... on a vehicle crew.  Followed by shutting down.
Think about it.  It's what we do.
- The Society

Thunder LRMs: the gift that keeps on giving.  They're the glitter of the BattleTech universe.

Burning Chrome

  • Lieutenant
  • *
  • Posts: 848
  • One of the Vocal Minority
Re: Weird Issues with 49.7 (and maybe stable 48 too)
« Reply #2 on: 23 May 2022, 11:13:47 »
Thanks for the reminder!  Both the Hermes II and Vulcan had "Eagle Eyes" as pilot skills and those also failed to detect any hidden vehicles despite being in range.  All hidden vees were discovered by blundering into or past them.

My Ostsol ended a turn 2 hexes away from a hidden Hetzer 2 hexes away and didn't detect it.  Not until I ended up right behind it and the hidden infantry in the same hex fired at me did I detect the Hetzer.  I thought the special sensor perk had a three hex range and would detect hidden vees and mechs.
"Matchup of the century: desire to play vs. resentment"

dgorsman

  • Captain
  • *
  • Posts: 1749
Re: Weird Issues with 49.7 (and maybe stable 48 too)
« Reply #3 on: 23 May 2022, 12:40:40 »
I *think* active probe detection of hidden units requires LOS, but I could be wrong.  Also only checks at the end of movement by default, there's an option to have it check during movement which was only enabled very recently.  It's quite dicey to detect hidden units even with the full range of the active probe.
Think about it.  It's what we do.
- The Society

Thunder LRMs: the gift that keeps on giving.  They're the glitter of the BattleTech universe.

NickAragua

  • Master Sergeant
  • *
  • Posts: 367
Re: Weird Issues with 49.7 (and maybe stable 48 too)
« Reply #4 on: 23 May 2022, 15:13:09 »
Simultaneous "anything" is pretty wonky and almost impossible to troubleshoot - once it gets into an error state, loading a save doesn't really tell us how it got there and it's impossible to predict what makes it get into that error state; I don't think there's anybody left on the dev team with the desire to support it at this point.

Active probes require LOS to detect hidden units within range. The "Improved Sensors" quirk does indeed have a 3-hex range. Eagle eyes is basically useless in MegaMek by itself, as it has a 1-hex range, and is pre-empted by point blank shots.

As for RNG, well, we actually put in special code to detect the exact user names you guys used so that we can give the bot to-hit bonuses as appropriate. The solution is to switch your user names until the next release, when we'll update the list and you'll have to switch user names again. [/troll]

Seriously though, RNG is just built-in java RNG. Bad luck happens sometimes. Sorry it didn't work out.

Burning Chrome

  • Lieutenant
  • *
  • Posts: 848
  • One of the Vocal Minority
Re: Weird Issues with 49.7 (and maybe stable 48 too)
« Reply #5 on: 23 May 2022, 22:09:33 »
Simultaneous weapons fire and physical attacks causes issues?  Wasn’t aware of that…

Is it better to not use those options?
"Matchup of the century: desire to play vs. resentment"

Hammer

  • Numerorum Malleo
  • Global Moderator
  • Major
  • *
  • Posts: 3514
    • MegaMek Website
Re: Weird Issues with 49.7 (and maybe stable 48 too)
« Reply #6 on: 23 May 2022, 22:18:15 »
Simultaneous weapons fire and physical attacks causes issues?  Wasn’t aware of that…

Is it better to not use those options?

Yep. They are unofficial and princess doesn’t support them. Plus the the 49.x releases they are causing issues with Multi play
MegaMek Projects Wiki
Bug Trackers
MegaMek Tracker
MekHQ Tracker
MegaMekLab Tracker
New Units and RAT's aren't added until after the 2 month release moratorium is passed.


Windchild

  • Master Sergeant
  • *
  • Posts: 290
Re: Weird Issues with 49.7 (and maybe stable 48 too)
« Reply #7 on: 24 May 2022, 13:05:29 »
I believe there are issues with simultaneous play, at least with Princess involved.  There's a pull request to remove that feature on the tracker.
To clarify, we are only removing sim movement and sim deployment... both of which have major issues.

Hammer covers the others, although it's just sim fire that's causing issues as far as I've seen.
MegaMek Developer - MekHQ, Personnel, and Internal Focus
MegaMek Discord
MegaMek Projects Wiki
Bug Trackers
MegaMek Tracker
MekHQ Tracker
MegaMekLab Tracker

Burning Chrome

  • Lieutenant
  • *
  • Posts: 848
  • One of the Vocal Minority
So...what I thought might have been an issue related to the new dev version but alas its affecting stable 48 too... :-\

I can no longer open any saved games.

"Error.  Unable to open game file"

Which is really "inconvenient" as I tend to make the games before hand, save and play when on the weekend when players are available.

I'm glad I found this out before I set to work on the next mission in our campaign.

Thoughts?  Advice?  Fixes?

Thanks in advance!
"Matchup of the century: desire to play vs. resentment"

Hammer

  • Numerorum Malleo
  • Global Moderator
  • Major
  • *
  • Posts: 3514
    • MegaMek Website
So...what I thought might have been an issue related to the new dev version but alas its affecting stable 48 too... :-\

I can no longer open any saved games.

"Error.  Unable to open game file"

Which is really "inconvenient" as I tend to make the games before hand, save and play when on the weekend when players are available.

I'm glad I found this out before I set to work on the next mission in our campaign.

Thoughts?  Advice?  Fixes?

Thanks in advance!

We'd need to see logs to troubleshoot.
MegaMek Projects Wiki
Bug Trackers
MegaMek Tracker
MekHQ Tracker
MegaMekLab Tracker
New Units and RAT's aren't added until after the 2 month release moratorium is passed.


Burning Chrome

  • Lieutenant
  • *
  • Posts: 848
  • One of the Vocal Minority
We'd need to see logs to troubleshoot.

Would love to help...is there a guide on where and how to pull and submit logs?

I believe it has something to do with the new version of JAVA...never had this issue before with JAVA8 and earlier.
« Last Edit: 27 May 2022, 15:46:02 by Burning Chrome »
"Matchup of the century: desire to play vs. resentment"

Windchild

  • Master Sergeant
  • *
  • Posts: 290
That sounds like using Java 17, which is completely unsupported and breaks save games
MegaMek Developer - MekHQ, Personnel, and Internal Focus
MegaMek Discord
MegaMek Projects Wiki
Bug Trackers
MegaMek Tracker
MekHQ Tracker
MegaMekLab Tracker

Burning Chrome

  • Lieutenant
  • *
  • Posts: 848
  • One of the Vocal Minority
That sounds like using Java 17, which is completely unsupported and breaks save games

I used the links provided for OpenJDK11U-jdk_x64_windows_hotspot_11.0.15_10.  Followed installation instructions and got it to work.

Uninstalled the above and reinstalled Java 8 - no issues with saved games on Stable 48.

Guess I'll be using that until I try the new version again.

It would be nice if the updated mechlist and icons/images from dev49 were used in stable 48.  Ah well...

Thanks for the awesomeness that is MegaMek.
« Last Edit: 27 May 2022, 17:48:27 by Burning Chrome »
"Matchup of the century: desire to play vs. resentment"