If you would like to pursue custom licensing, please visit this page and fill out the form on the right side of the page. After this, you will be contacted about how to proceed.
I’ve sent you a private message about this matter, partially because it’s unrelated to the issue at hand but also because I need some information from you to see where your submission is in the process.
Yea I’ve done that. Could you ask them to answer their email?
After running 6 instances of the project provided for ~15 minutes, 3 of the 6 instances have already frozen. With a fairly consistent repro, I’ll be entering a JIRA report for this with the information provided about ’ project specifically, as not to cause confusion. There will be a link to this post on the JIRA report and the engineers looking into this are aware of the possible different natures of all these freezes. I’ll update here when the ticket is in.
Unfortunately 2 of those days were the weekend. But I do wish this had been escalated earlier.
Matthew, I’m concerned here. I thought this was being escalated 4 days ago when you said “Due to this, I’ll be escalating this issue to someone more specialized in this area.”
I’d also like to stress like the others, we can’t make you(unreal engine) money until we launch the game… which is waiting to launch solely because of this bug. Can we umm… escalate this to a code emergency real life zombie breakout that needs to be contained and exterminated ASAP?
I’ve entered a JIRA report with ’ project as the reproduction case. The report can be found here: UE-52611
As previously mentioned, the report is mainly focused around the issue in the project that provided, as we know exactly what the cause is in that project. If you have any questions in the meantime while this is reviewed, please let us know here.
, your JIRA report says “The issue in this particular project is caused by the TickGroup being set to TG_PostUpdateWork on one of the actors. If this is not set, the issue does not occur.”
Just going to let this be known… I posted about this issue first, sent you our project, you straight up ignored said project, you took someone elses project and made it known in your report that it was TG_PostUpdateWork but our project DOES NOT have any TG reassigned. So now you are categorizing multiple projects under 1 issue that may seemingly be unrelated. Sorry but this is concerning. I would assume to make an educated guess and troubleshoot properly, you would try and find as many similarities as possible.
I feel like my thread has been hijacked, and preferential treatment is being made here.
Hello dodi,
I was not even aware that you and were working on the same project. Other than your most recent two posts, I had assumed that you had abandoned the post and had not provided any information.
Rest assured that if the issue ends up being different and the fix provided for this does not solve your problem, we will fix it. As I have explained to in a private message after they aired their concerns, the aim was to get a reproduction and we have done that. ’ project has produced a JIRA because not only did it provide a very quick and concise reproduction with small project, it even pinpoints the issue that is occurring and why it is occurring.
Also, as I told , we try to avoid getting a user’s project, especially ones this close to release, unless we absolutely need to for reproduction purposes. As you all seemed to be collaborating and reporting as this as the same issue, even though the ways of getting to it were different, I limited myself to working on one at a time.
I’ll be downloading the project provided now to try to reproduce the issue.
Thank you Matthew. You only need to package the project as is, and connect to the dedicated server. Within 2 hours, probably sooner (often it’s only 30-45 minutes for me), you’ll freeze.
I’ve tried removing all particles to no effect, it seems to me it might be one of the actors in the level (there are only 4 besides the character) are causing the issue as I’ve tested an empty map and it will only freeze when the actors are present (bp_goal, bp_jumppad, bp_powerups, possibly BP_ball though I don’t think the ball is at fault).
However, we don’t change any tick groups. We had tried moving a lot of particle components to TG_PostPhysics once, but yielded no results.
The call stack is ALWAYS the same no matter what combination of actors & particles produce the freeze.
Hijacking was definitely the opposite of my intent here. One of the primary reasons I worked so long and hard to get the repro case setup for Epic is because of all of us here. The hope was that with a fresh project and simple repro steps the engineers at Epic would be able to more efficiently diagnose the root of a problem that is possibly the cause for all of our problems.
I wasn’t looking for preferrential treatment, just trying to help speed things along for everyone’s sake, and I can only work with the code that is in front of me.
I hope debugging goes well on your end as well, and am very sorry my solution was not able to help.
Thanks ! Sorry that’s my fault for not informing we’re working together. Thank you sincerely for confirming you are testing out our project as well. It’s brought some hope. This bug has been morale crushing, we’re going on like 2+ months of this. We’re heavily invested in this game, so many people are waiting on an engine fix. We’ve had to postpone major events with partners over this.
you are 100% fine bud, the direction I was going with my message wasn’t meant to be directed to anyone specifically by any means, the direction I was attempting to go with is that it was appearing as if we we’re getting left behind and sole focus shifted to yours ( this is based off pm’s and the JIRA description). If we’re all having issues, we all need help. Group effort is great! We can’t really afford this bug hunt to be narrowly sighted in when it appears its cause may be slightly different (based off the JIRA report matthew posted and our project not having TG reassigned.)
I’m from a science background, where typically it’s most common to take a large sample pool and compare similarities. That being said I’d also like to get us all in a discord or something away from this answerhubs cluttered nightmare to discuss similarities hopefully more efficiently.
We will test this as soon as possible!
We will try this too (dodi, and me). It seems low level enough to be a solution for all of us.
Looking through the code, I assume it will be cleaned up once this is confirmed to fix the issue?
Hey all, my team have also been experiencing random freezes in PIE and standalone and packaged games, I have applied this fix to our engine source and it seems to have fixed it from what I can tell so far. Also using 4.18.
It also seems to fix our freeze (dodi, and mine).
However, how much is that a “hack” that will have bad effects in other situations?
I mean, I’m glad the freeze seems to be gone, but that looks more like stopping stalling in general?
The actual cause of the stall isn’t removed with that or?
I think its a temporary fix for now. We’ve started receiving the following in the log when streaming a level:
[2017.11.24-10.49.15:367][836]LogStreaming: Error: Couldn't find file for package /Script/FunctionalTesting requested by async loading code. NameToLoad: /Script/FunctionalTesting
[2017.11.24-10.49.15:368][836]LogStreaming: Error: This will hitch streaming because it ends up searching the disk instead of finding the file in the pak file.
[2017.11.24-10.49.15:369][836]LogStreaming: Error: Found 1 dependent packages...
[2017.11.24-10.49.15:369][836]LogStreaming: Error: /Game/Maps/Build/IntroLevels/Sector4/CubePlacementPart3
We’ve never had this error before, although everything still loads, there is a minor hitch when streaming this particular level in a packaged game. Other than that, this fix seems to have improved level transition loading performance and overall game performance, infact our game has never performed as well due to this hack, bar the above hitch.
I’ve had several devs report they are still freezing at least from switching screen mode and resolution. I don’t have any reports of freezing during gameplay yet. Not a lot of testing though because of the holiday, and I haven’t had a chance yet to personally test it myself.
Hello, we were having this issue with a project we ported over from 4.16 to 4.18. We applied the patch from the github link and so far so good. We will have to conduct more testing, but it was “freezing” rather frequently and after an hour of running matches it hasn’t frozen once. Thanks everyone that assisted in helping to pinpoint this issue and for getting the fix out.
We have not had a freeze since we implemented the engine change. At least 30 hours so far. I believe this has done the trick.
e; Thank you Matthew & Epic Engineers!