Ashes of Creation

Secrets

ResetEra Staff Member
1,485
1,308
Not sure I would have jumped on the UE5 bandwagon mid development. Massive undertaking with a ton of new bugs and features. Kiss those production milestones goodbye...and RIP to your QA teams quality of life.

Maybe focus on releasing the game…
It's slightly a bit more work than a version bump, from what I heard.

Definitely not as massive as going as Unreal 3 -> Unreal 4, at any rate. And based on that video alone, it seems they already did a majority of lifting.
 

Daidraco

Ahn'Qiraj Raider
5,837
3,583
65d 16h 15m
It's slightly a bit more work than a version bump, from what I heard.

Definitely not as massive as going as Unreal 3 -> Unreal 4, at any rate. And based on that video alone, it seems they already did a majority of lifting.
You might know the possibility of this? But could they have been planning ahead for a "just in case" or future expansion and that made the upgrade a lot easier?
 

Mahes

Blackwing Lair Raider
3,142
1,876
44d 13h 59m
To my understanding, this was discussed for months both on the forums and within Intrepid as to whether or not they were going to make the move. The one thing stated was that no move would be made if it appeared to actually impede the time process for releasing this game. As it was explained, the process of actually moving the game will add a few months for the move but it comes with a lot of advantages that will shorten the work time post move for the rest of the process and even beyond release. I think in the long run it was the right decision, despite the few months delay. Now hopefully we start to see progress to the game itself this next year.
 

mkopec

<Gold Donor>
23,579
34,804
181d 21h 30m
Honestly if this move helps improve development, than this pause will pay for itself many times over and even after release with patch and content cycles.

Its what hampered Destiny 2 for over 2 yrs (or 1? cant remember). when only one dev at a time was allowed to fuck with files and then wait 24 hrs to compile all the shit. Like even to move one rock. Imagine the planning that had to go into even the most basic of patches.
 
  • 1Like
Reactions: Rhanyn

Gravel

Mr. Poopybutthole
28,115
76,495
204d 4h 30m
Honestly if this move helps improve development, than this pause will pay for itself many times over and even after release with patch and content cycles.

Its what hampered Destiny 2 for over 2 yrs (or 1? cant remember). when only one dev at a time was allowed to fuck with files and then wait 24 hrs to compile all the shit. Like even to move one rock. Imagine the planning that had to go into even the most basic of patches.
Pretty sure it was both. Destiny 1 eventually fixed it, but Destiny 2 was built using some of the same code, and so it ended up being in it at release. Not sure if they resolved it again.
 

Daidraco

Ahn'Qiraj Raider
5,837
3,583
65d 16h 15m
Pretty sure it was both. Destiny 1 eventually fixed it, but Destiny 2 was built using some of the same code, and so it ended up being in it at release. Not sure if they resolved it again.
Wait, was that how its worked all this time? If so, then I have to give devs so many more props in this avenue than I actually thought. That shit would drive me insane.
 

Warrik

<Donor>
1,191
1,203
6d 10h 31m
Honestly if this move helps improve development, than this pause will pay for itself many times over and even after release with patch and content cycles.

Its what hampered Destiny 2 for over 2 yrs (or 1? cant remember). when only one dev at a time was allowed to fuck with files and then wait 24 hrs to compile all the shit. Like even to move one rock. Imagine the planning that had to go into even the most basic of patches.

This is close to accurate in regards to engine upgrades (can't speak specifically to Destiny). Even taking a minor version drop can be a big deal. Typically though, we lock the branch or stream, and one guy will do the merge and resolve issues, CI will run on the farm, etc...I've seen it go smooth and not so smooth. I do know major version bumps are usually far more intense, and cant imagine UE4-5 any different.

That said, if they can (have) pulled it off, then its a win for sure. In my experience though, its usually increases scope quite a bit. Depending on what cadence they run their sprints at, you could lose a few to resolving issues, stabilization, refactor, etc.

Don't get me wrong, its cool to see them on UE5, just not sure overall the schedule impact and I would be shocked if it was minimal.

wait 24 hrs to compile all the shit.

Most places have overnight CI that runs and does this which sounds like what your referring to. To compile binaries for respawn (3 types) took us about an hour, but thats because we had broken up monolithic CI jobs into smaller independent tasks. I typically like to have CI that runs and compiles at least one type with each data change to try and smoke out a bad commit. For engine upgrades ( or visual studio, which can be a pita as well ), we usually like to prepare to support the Code team post integration with some custom automated tests and jobs.

Not every game studio has a Build Engineering team, or DevOps, etc. Most studios have someone in QA or on the Dev team that runs this stuff on the farm (or their local machine), which makes them nimble, but costs them in speed and efficiency when it comes to CI. (it has a lot more cons than pros). For those studios, they usually do this manually, i.e. Dev merges code, resolves conflicts (which can be time consuming), submits, and then manually compiles locally. If you aren't leveraging distributed build your compiles can take quite some time. I think a full build for Fortnite took us 7 hours. Full build for Apex took about 2-3 hours. New World took 5-6 hours. Thats everything: client, server, audio render, shaders and textures (cache is faster, fresh render of all assets much longer), etc. UE cook time was a pain point for Fortnite (in 2017. I cannot speak to any improvements or changes made since then).
 
Last edited:
  • 2Like
Reactions: Flobee and wilkxus

Secrets

ResetEra Staff Member
1,485
1,308
You might know the possibility of this? But could they have been planning ahead for a "just in case" or future expansion and that made the upgrade a lot easier?
I've worked on an Unreal 3, and an Unreal project 4 and was responsible alongside the lead for doing version bumps with BeyondCompare. It took us a solid week every version bump. A major (ie; 4.2x) took us maybe 2 weeks to a month.

It would make sense if they put a lead or sr/staff engineer in a dark basement and just had them complete the Unreal 5 version upgrade branched while merging in the 'development' branch.

Unreal 3 to 4 wasn't compatible due to the major changes in UnrealScript. But 4 to 5 is a glorified version bump with better replication, and improved rendering. Fundamentals were mostly the same from the gameplay programming / non-rendering engine code side. Improved, if anything.

Warrick is 100% spot on here with the build farm stuff. I was super privileged to have a lead programmer in my first programming gig who had a TeamCity setup and we had Ansible scripts that just automatically managed nightly / on-demand builds. Next company had just a Incredibuild farm with some Jenkins scripts, and it was far slower until we had my threadripper attached.
 
  • 2Like
Reactions: TJT and wilkxus

Qhue

Trump's Staff
6,927
3,807
40d 20h 16m
Maybe they should focus on getting something actually finished enough to be played. I am sensing a strong Wildstar vibe at this point wherein they will do a ton of work to release a complete MMO experience that then just dies on the vine.
 

Nirgon

YOU HAVE NO POWER HERE
8,897
13,709
73d 5h 35m
Not sure I would have jumped on the UE5 bandwagon mid development. Massive undertaking with a ton of new bugs and features. Kiss those production milestones goodbye...and RIP to your QA teams quality of life.

Maybe focus on releasing the game…

Smells like the version jump was needed to resolve something they couldn't figure out or work around otherwise?
 

ToeMissile

Pronouns: zie/zhem/zer
<Gold Donor>
1,660
780
26d 7h 1m
Ridiculous amount of customization.
Same tech/tool will/is apparently used by the devs for NPC design and customization.
 

Mahes

Blackwing Lair Raider
3,142
1,876
44d 13h 59m
I like how they said that players will also get to use it for Housing/Castle NPC's as well. Will certainly add a flavor to areas of interest when all the NPC's are large breasted woman or retard looking outcasts.
 
  • 1Worf
Reactions: Lasch

Rhanyn

<Bronze Donator>
792
999
111d 5h 58m
What's the almost certainly inaccurate projected ETA for when this is coming out these days?
 

Mahes

Blackwing Lair Raider
3,142
1,876
44d 13h 59m
I would say 2 years until Alpha 2 at the very least. The director keeps referencing what will be in Alpha 2 but of course never mentions a target date. Alpha one had about 10% of what they are claiming is going to be in the game and that started 1 year ago. I am guessing that the next reveals will involve actual class designs based on his latest brief hypothetical discussions of the assassin. He also brought up the combat revamp which seemed like it was pretty much done in design and was getting programmed in.

I feel like a lot more actual information is going to come around mid-summer early fall, as it seemed Steven(Director) was chomping on the bit not to accidentally say something that was not ready for release.
 

Xerge

<Donor>
1,092
981
63d 7h 28m
Id say less than 2 years but more than 1. They have an internal date but every time I pry they get hush and I get the stink eye.
 
Last edited:
  • 1Like
Reactions: Daidraco