devtrackers.gg
8 months
ago -
Coffee Stain Studios
-
Direct link
Transcript (by Youtube)
| 0s | hey what's up I'm snot I'm the community |
|---|---|
| 3s | manager for coffee stain Studios the |
| 5s | only Community manager for I know that |
| 7s | okay maybe this is a good time to |
| 8s | address this uh I know that some people |
| 10s | might be asking like didn't you guys |
| 11s | hire like a third Community manager like |
| 13s | where where's Ela what happened to ELA |
| 14s | so Ela moved back to Turkey uh and then |
| 17s | jce left and now it's just me the the |
| 20s | only one left I'm the community manager |
| 22s | for C |
| 24s | studios in today's video I wanted to |
| 26s | give a quick state of Deb for up toate I |
| 29s | also wanted to give a quick PSA if |
| 32s | you're playing on experimental on up |
| 33s | toate and have run into an issue with a |
| 36s | cheeky copper node that's producing |
| 38s | limestone in the western canyons and I |
| 41s | want to talk a little bit about a topic |
| 43s | that comes up every now and again |
| 44s | regarding the download size for our |
| 46s | patches so lots to talk about there are |
| 49s | time stamps for these topics if you want |
| 51s | to skip ahead so let's get into it so |
| 54s | first and foremost let's do a real quick |
| 56s | recap of what's happened since our last |
| 58s | video and be sure to grab WP a pen and |
| 61s | paper and take notes cuz this is all |
| 63s | going to be in the test all right and |
| 65s | I'm going to mention a whole bunch of |
| 66s | different versions of the game and it |
| 69s | might be confusing but let's go all |
| 72s | right so since coming back from vacation |
| 74s | we made a build including a bunch of |
| 76s | performance improvements and Bug fixes |
| 78s | we had been working on for update 8 and |
| 80s | release it on our experimental Branch |
| 83s | this was update 0.8.1 actually before we |
| 86s | progress further with the video uh let's |
| 88s | just quickly explain the way we version |
| 90s | of the game because it might make things |
| 92s | clearer so when we post patch notes you |
| 95s | might notice that we post this |
| 97s | information along with the update this |
| 99s | is our public version of the game we |
| 101s | also include something called the Chang |
| 103s | list number which is our internal number |
| 105s | to keep track of the build so this |
| 107s | change list is also what you're seeing |
| 109s | at the top left corner of the game this |
| 111s | is so that we can see in videos or |
| 113s | screenshots what exact build people are |
| 115s | playing the public version follows a |
| 117s | somewhat loose version of semantic |
| 120s | versioning uh you don't really need to |
| 122s | know too much about what that is other |
| 124s | than it's a sort of standard for for |
| 126s | numbering software releases by denoting |
| 129s | what type of release uh each version is |
| 132s | each number corresponds to a major minor |
| 135s | and Patch changes for the game and since |
| 138s | we're in early axis and we needed a bit |
| 140s | more granularity uh the way we do this |
| 143s | is by somewhat following semantic |
| 145s | versioning but shifting everything one |
| 147s | step to the right so when we release the |
| 149s | initial version of update 8 we denote it |
| 151s | with version 0.8.0 point0 and when we do |
| 155s | hot fixes and Bug fixes to this version |
| 157s | We increment the patching number of that |
| 159s | version to 0.8.0 point1 and if we hot |
| 163s | fix again it'll be |
| 166s | 0.8.2 and so on uh and when we make |
| 169s | bigger changes to updates or new |
| 171s | features uh we increment the minor |
| 174s | version so the version we released after |
| 177s | vacation including dlss and the |
| 179s | performance Improvement |
| 180s | was |
| 182s | 0.8.1 point0 and then if we make hot |
| 186s | fixes we again increment that patching |
| 188s | number of the version to be |
| 190s | 0.8.1 point1 and so on and so forth |
| 194s | we're not always like super consistent |
| 196s | with the way we do this public version |
| 198s | but this is roughly how we treat it all |
| 201s | right confusing yes it is all right cool |
| 204s | back to the topic at hand so since the |
| 206s | release of 0.8.1 we've seen significant |
| 209s | improvements in perform performance |
| 210s | across the board for people and we are |
| 213s | super stoked about that so the plan was |
| 215s | to release one patch on experimental |
| 217s | that's one patch uh just to like include |
| 220s | those performance fixes and then focus |
| 222s | on working on |
| 224s | 0.8.2 which will include upgrading the |
| 226s | engine again to Unreal Engine 5.2 uh |
| 229s | which will resolve a bunch of issues |
| 231s | with update 8 however with the release |
| 233s | of 0.8.1 we ended up needing to do a lot |
| 235s | of hot fixing as initially dedicated |
| 238s | servers on Linux wouldn't boot uh people |
| 240s | were crashing when joining dedicated |
| 242s | servers uh pipeline flow indicators |
| 244s | broke crab hatchers and creatures |
| 246s | spawning broke there were a few issues |
| 248s | and and we we had to take care of that |
| 250s | so we fixed most of those blockers |
| 253s | for now but there's still like a few |
| 255s | problems lingering with the current |
| 256s | build uh has to do with science and and |
| 258s | whatnot uh but we really need to focus |
| 263s | on stabilizing the the 0.8.2 build um so |
| 268s | at this point we won't be able to go |
| 269s | back back and fix up issues related to |
| 272s | 0.8.1 anymore because it's significantly |
| 274s | slowing down progress on the stable |
| 276s | release of update 8 as a whole so we've |
| 279s | made a lot of progress on updating the |
| 280s | engine 5 by 2 there's still a couple of |
| 283s | problems |
| 295s | left and we sort of need all hands on |
| 298s | deck to be able to stabilize |
| 300s | 0.8.2 uh cuz there's a whole lot of work |
| 303s | needed left so for now there'll be no |
| 305s | more patches on 0.8.1 so we can focus |
| 309s | all our efforts on |
| 311s | 0.8.2 uh it'll be a quicker road towards |
| 314s | a stable release of update 8 but quite |
| 317s | frustrating though for you folks playing |
| 319s | on experimental right now running into |
| 320s | the current issues until we can get |
| 322s | around to releasing uh 0.8.2 I love how |
| 325s | I said releasing with a z there that |
| 328s | that's free it's not an optimal solution |
| 331s | but I hope this explanation sort of |
| 333s | makes sense and that you can sort of |
| 334s | bear with us just just a little bit more |
| 337s | um we really need to make sure that the |
| 340s | stable release for update is as good as |
| 342s | it can be |
| 344s | and that's just how it is right now as |
| 346s | to when you can expect to see 0.8.2 and |
| 349s | the stable release of update 8 uh for |
| 351s | that matter it's still hard to say we |
| 353s | don't really have a deadline we're |
| 354s | working towards as we think it's more |
| 356s | important that the update gets like the |
| 358s | necessary work done and we don't want to |
| 360s | like overextend ourselves uh given it's |
| 363s | so hard to like kind of estimate this |
| 364s | kind of work so at the moment we think |
| 367s | that we need probably another month |
| 369s | before we can release 0.8.2 on |
| 371s | experimental and then we probably need |
| 373s | like a little bit more time to fix it up |
| 375s | any lingering issues uh but then we |
| 378s | should be good to go upgrading the |
| 380s | engine to unre Engine 5 has been such a |
| 382s | huge undertaking and even though we |
| 384s | started working on this upgrade since |
| 385s | the start of the Year we're still seeing |
| 387s | sort of the Fallout from it so it's |
| 390s | going to be worth it in the long run I |
| 392s | believe but geez we were expecting that |
| 395s | it would take longer than we were |
| 397s | expecting but it took even longer than |
| 400s | we were expecting that we didn't expect |
| 402s | it to |
| 403s | take but you're all playing Starfield |
| 406s | and Ballers Gate 3 and Texas Chainsaw |
| 409s | Massacre and armor core 6 and mind |
| 411s | sweeper and ringfit Adventure so |
| 414s | hopefully you won't like miss us for a |
| 416s | bit more next up I wanted to do a little |
| 418s | PSA regarding a known issue on the |
| 420s | experimental Branch regarding one of the |
| 422s | copper nodes the desert Canyons that's |
| 425s | producing Limestone for some reason can |
| 427s | you believe that can you believe the |
| 428s | audacity of that node we don't know |
| 431s | exactly what happened but for some |
| 433s | reason that node just started producing |
| 435s | Limestone uh it might be a migration |
| 437s | issue like we don't really know at this |
| 439s | time but uh also quickly to point out |
| 441s | like another node that's also been |
| 443s | affected is a limestone node in the |
| 444s | western swamp so there might be some |
| 446s | kind of like trickery happening there I |
| 448s | don't know so if you already had minor |
| 450s | placed on this node in desert Canyons |
| 452s | it's it's going to start producing |
| 453s | Limestone that's bad our solution to |
| 455s | this problem is that we're going to |
| 457s | replace these nodes uh what's good to |
| 460s | know here is if you're playing on |
| 462s | experimental and you've got miners |
| 464s | acting up because of this issue you can |
| 466s | fix this by just rebuilding the minor on |
| 469s | the Node it's supposed to be sitting |
| 471s | under uh folks that haven't played on |
| 473s | experimental won't notice any difference |
| 476s | uh because once the update goes out on |
| 478s | stable um it'll all be solved so like |
| 480s | this will only affect people that played |
| 482s | up to date on experimental and have had |
| 484s | miners on those nodes so that's it |
| 488s | that's it so lastly I want to talk about |
| 490s | the patching system used by Steam and |
| 492s | epic because often when we release hot |
| 494s | fixes and patches to satisfactory uh |
| 497s | people tend to comment that like the |
| 498s | download sizes are quite large for |
| 500s | seemingly small changes and I don't know |
| 502s | if we've ever explain what's going on so |
| 505s | what's important to know is that we |
| 506s | don't handle distribution of our game |
| 509s | like at all |
| 510s | uh that's all handled by each respective |
| 512s | platform so that's both in terms of like |
| 514s | you know sales and that kind of stuff |
| 517s | but also like how the game is downloaded |
| 518s | Steam and epic uses Delta encoding when |
| 520s | patching games this is how they've done |
| 522s | it since the dawn of time essentially so |
| 524s | it's not that they've suddenly changeed |
| 526s | the system or anything like that uh and |
| 528s | then it's like it's not because of that |
| 529s | change you're seeing bigger patches like |
| 531s | it's it's nothing like that it's always |
| 532s | been like that uh it's probably just |
| 534s | circumstantial if you notice that the |
| 536s | patches are big or anything like that uh |
| 538s | anyways the the way Delta encoding works |
| 541s | is that each platform encodes and |
| 543s | manages like how the build is |
| 544s | distributed over the internet and it |
| 546s | attempts to look at what version you |
| 548s | have installed and what version is to be |
| 550s | downloaded and they only send you the |
| 552s | delta or the difference between your |
| 555s | local version and the latest version and |
| 557s | then locally it will sort of recreate |
| 559s | the build using the data from the Delta |
| 562s | uh this system isn't always perfect as |
| 564s | it sometimes can mess up when you when |
| 566s | it recreates the build locally this is |
| 568s | why it's so important to verify your |
| 569s | game files if you run into any weird |
| 572s | issues with the game so using Delta |
| 574s | encoding reduces the download bandwidth |
| 576s | but it also means that it might take |
| 577s | longer to unpack and recreate the build |
| 579s | locally if it's not installed on an SSD |
| 582s | for instance so we don't manage the |
| 584s | patches that you download it's all |
| 586s | handled by the respective platforms that |
| 588s | being said there are things game |
| 591s | developers can do to improve how their |
| 593s | games are handled by that patching |
| 595s | system to get the most efficient |
| 597s | downloading of initial like the initial |
| 599s | game install and patching of updates and |
| 602s | we've taken most of the recommended |
| 604s | steps but there are a few things we can |
| 606s | do to improve download download size but |
| 609s | these systems are kind of like a |
| 610s | blackbox so sort of hard to know for |
| 612s | sure in the future we have plans for the |
| 614s | game to take advantage of a new runtime |
| 617s | loader that should optimize how the game |
| 619s | is handled when loading resources and |
| 621s | it's possible that we can better adjust |
| 623s | how the game is packaged uh to make the |
| 626s | Delta encoding handled by each platform |
| 628s | to be a bit more efficient but yeah it's |
| 631s | hard to say like how much better it will |
| 633s | actually get at this point one thing I |
| 635s | also want to point out is that it's a |
| 637s | bit unfair uh for folks to make |
| 639s | assumptions as to how much the binaries |
| 642s | should change just based on what you see |
| 643s | in patch notes and in the game uh a |
| 646s | single line in the patch nodes doesn't |
| 648s | imply that it's a small fix under the |
| 650s | hood and seemingly small fixes can |
| 652s | sometimes boil down to like quite |
| 654s | extensive changes I also often see the |
| 656s | argument that we're just changing like a |
| 658s | few lines of code and that should just |
| 659s | be like a couple of kilobytes of change |
| 662s | and yes that is true for source code but |
| 665s | source code is then compiled into |
| 666s | machine code and then it's cooked and |
| 668s | packaged with the rest of the resources |
| 670s | the game uses in a quite complex |
| 673s | packaging process so to optimize how the |
| 675s | game takes advantage of those resources |
| 677s | during a load so it's not quite that |
| 680s | simple and with that being said I think |
| 682s | that's it so thanks everyone for |
| 683s | watching this video I'm so glad you took |
| 685s | some time off your busy schedule playing |
| 687s | Starfield to watch this video |
| 690s | that means a lot to me that really does |
| 693s | that is of course unless you got the |
| 694s | standard |
| 696s | version stand the standard |
| 700s | version video games am I right all right |
| 704s | anyway take care everybody |
| 712s | bye-bye first and I keep so I did this I |
| 717s | did this whole thing a lot in the last |
| 718s | video and people are commenting like |
| 720s | crazy about that so when I try to |
| 723s | not do that |