Multi Player bugs and crashes - After the 18th of August 2014

[Thunderbrd]

I did encounter an OOS upon reloading after an OOS that had to do with movement routing or a difference in the order of the unit move and a road being built along the route. Rare problem so far but I'll be watching for it.

We were able to easily resolve the OOS on reload (which is very odd in itself) by saving again, going out to main menu, and reloading the new save. I have to wonder if somehow one of us going to desktop and the other going to main menu might've caused the immediate OOS. I'll be watching for more of this kind of thing.

But OOS's are so rare that we've only had about one each night of rather extended play.

 

[Toffer90]

But OOS's are so rare that we've only had about one each night of rather extended play.

What era, on what map size? Giant late classical had very frequent OOS from what I remember.

 

[Thunderbrd]

We're in Classical now and it hasn't worsened (yet).

 

[sintri]

Seems essentially oos only when one comp stalls while processing things, though not all momentary freezes cause oos, pretty much every oos will follow a momentary freeze. Seems from the oos logs the comp that doesn't freeze is just processing things during the duration when the other comp's stalled so when the stalled comp resumes play the states are mismatched.

Almost at the end of classical and oos every 3-10 turns. There's a huge difference between the turn processing times of both comps as well. Would probably be less of an issue if there was a dedicated server processing all the ai actions as opposed to each computer having to do it separately.

 

[Thunderbrd]

Seems essentially oos only when one comp stalls while processing things, though not all momentary freezes cause oos, pretty much every oos will follow a momentary freeze. Seems from the oos logs the comp that doesn't freeze is just processing things during the duration when the other comp's stalled so when the stalled comp resumes play the states are mismatched.

Almost at the end of classical and oos every 3-10 turns. There's a huge difference between the turn processing times of both comps as well. Would probably be less of an issue if there was a dedicated server processing all the ai actions as opposed to each computer having to do it separately.

I'm currently evaluating the OOS point that you're talking about that starts taking place at the end of classical. Though I had it resolved but it's a little deeper apparently.

 

[Folket]

Did anyone ever look at why autosaves did not work in hotseat?

 

[Thunderbrd]

NVM... I believe I've misdiagnosed an issue... I'll bring it up elsewhere.

 

[Folket]

I have had the question about hotseat autosaves for three years. Do anyone even read the mp bug thread? IT would be Nice to Know why i have been ignored.


Skickat från min iPhone med Tapatalk

 

[Sparth]

I have had the question about hotseat autosaves for three years. Do anyone even read the mp bug thread? IT would be Nice to Know why i have been ignored.

If no one answered then probably no one knows this bug and no one know how to fix it.

 

[Folket]

If no one answered then probably no one knows this bug and no one know how to fix it.

I'm uncertain what "no one knows the bug" means here. They should know about it since I have reported.

All I would like is someone saying something like: "We do not plan to support hotseat.", "Hotseat autosave works for me." or "We added the bug to our bugtracker but it has very low priority."

I cn't tell if anyone had seen my post when there is just silence.

 

[Thunderbrd]

I'm uncertain what "no one knows the bug" means here. They should know about it since I have reported.

All I would like is someone saying something like: "We do not plan to support hotseat.", "Hotseat autosave works for me." or "We added the bug to our bugtracker but it has very low priority."

I cn't tell if anyone had seen my post when there is just silence.

I was hoping it would clear itself up because I have no idea how to even look into fixing it. Was also hoping that if it didn't, another modder with more familiarity with save and load procedures might've been able to find the issue and resolve it.

From what limited knowledge I have of the matter, it was likely a problem with a particular game you had. But if it's more prevalent and still continues to be an issue on the SVN version, then I can't fathom how a hotseat game in particular over any other type of game would be having a specific problem with this. But if you gave us a savegame it will give us an opportunity to at least look into it... though understand that it would be low priority because there are larger problems with more critical areas.

Are you having a problem with saving and loading not working at all or just autosaves on hotseat?

 

[Folket]

A autosave from a hotseat game has never worked for me. While single player autosaves and hotseat normal saves work fine.

 

[Thunderbrd]

K. It'll take a long while before looking into that but it's officially a known problem to eventually resolve.

 

[Folket]

Thanks.

 

[sintri]

OOS seems to occur more recent in the svns in the past 2 weeks. Used to have stretches where we'd never run into oos, past 2 weeks oos pretty much 10-20 turns in.

 

[Thunderbrd]

There's some issues with Naval movement that can be a real pain. I've been trying to narrow in on the problem but it's one that can seem to be due to updates when in fact it has more to do with how the game is faring and what's happening in it. The problem is somewhere in pathing and THAT is a HUGE problem because pathing is a major b***h to even read through, let alone try to find an OOS bug in.

 

[Folket]

Sounds like a challenge.