Star Citizen Dev Progress Watch

[wiser3754]

Looking good!

The sarcasm level is . . . . Over 9000!!!

[dsmart]

SSOCS no longer appears in the roadmap. Which means it's not in dev. The End.

It's so weird how, for months, I had stated that neither OCS nor SSOCS (a pipedream) were going to be of any relevance. Some folks are only now noticing?

So I typed this out after skimming the most recent SCL, and the body language and non-verbal cues from the two poor bastards on camera (Jared and Todd Papy) were screaming at me so early on I started paying more attention than usual.

First of all, the first 30 seconds of this segment are so fraught with awkward tension, if someone popped a balloon I think Papy might spontaneously burst into tears. Sure they're devs and don't like the camera but.... nobody is enjoying these questions.

Three moments in particular though stood out to me, which are revelatory in a read-between-the-lines kinda way.

5m 45s; Papy tells us that SSOCS has actually tanked server framerates to.... 6fps. That is less than most clicker mobile games backends run at, which hang off cheap Chinese VPS slices. That's..... dreadful. This is only Stanton we're talking about, remember. This is nowhere near the design density of an SC system.

6m 40s; Papy's tone and turn of phrase indicates strongly that SSOCS disappointed them internally, and that they're basically hoping that they'll fix it in future. SSOCS is too compute-intensive, has cost every team a ton of work, and will cost their best guys tons more effort in future, and it might not even work right.

20m 10s; This one is more subtle. Jared is animatedly boring us about 30K errors (I left this in as a baseline for his character in the segment), when Papy suddenly mentions SSOCS.... Jared starts nodding reflexively.... and then Papy mentions Server Meshing. Watch Jared abruptly freeze stock-still and barely gulp out a silent 'myep....' and move on from it super fast with some random anecdote. Jared doesn't want to be on camera talking up Server Meshing, and didn't like Papy doing it.

These three moments alone (there are tons more) are so strongly indicative of the peril the technical team are facing. There are other gems in there, such as the 'feature team' and the 'content team' being so autistically disconnected from one another that they cause problems for each other constantly....

But generally, yeah. Based purely on nonverbals and word selection, I feel I can infer;

1) SSOCS died on its arse, and is tanking the server backend and taking the AI down with the ship.

2) SSOCS is a boat anchor tech that will absorb talent for months with no clear end in sight.

The mere mention of Server Meshing makes Jared so profoundly nervous that his limbic system took over and engaged his fight/flight/freeze reflex. Server Meshing (of which we have heard nothing but the name) doesn't sound good to Jared.

[dsmart]

Looking good!

https://thumbs.gfycat.com/BouncyGlaringBaboon-mobile.mp4

[wiser3754]

Looking good!

https://thumbs.gfycat.com/BouncyGlaringBaboon-mobile.mp4

Missiles still not working? But, but, SOCS. It was made to fix this sort of stuff!

[DemonInvestor]

@Derek

About that All About Alpha 3.8. video.
Just skimming the video a bit, as some site question, when Todd goes into the problems of their current AI and how that's steming from the backend not being able to keep up. I especially wondered about him mentioning their inability to simulate actual player impact on the server (7:29). For
1) 50 players doesn't sound like something that should bother most servers, when dozens of local host games can go into that direction.
2) I'd read some of Ron Gilbert blog posts for Thimbleweed park and how he went about to simulate a player by some very basic bot, shouldn't such a setup be enough to actually stress a server enough to catch a lot of obvious problems? I mean they've got all 'map', packet and Co. data, so doing a basic externally run bot who's "playing" should be easy enough to do right?

As always i'm not a programer or anything, but that sounded rather strange to me.

[wiser3754]

https://www.reddit.com/r/starcitizen/comments/eqh28a/star_citizen_roadmap_update_20200117/

The backers aren't too thrilled about this roadmap update.

[wiser3754]

Ah Lorville. When was this introduced? 3.3.5? And yet here we are. Funny how the coffe cup isn't mentioned.

[N0mad]

I don't usually watch these, but this was absolute drivel. Are they just trolling their supporters? are they even making a game any more?

Although I feel a bit sorry for the concept artist - his life spent trying to please Chris must be absolute hell. It's a little insight into the CRoberts micromanaging technique: force them to keep iterating over designs again and again until Chris seems happy with the result. No wonder they never actually release anything.

[wiser3754]

2020 has a had a very surprising start, thanks in no small part to China and the world's dependence on the world's biggest factory.

I'm sure all of us here have noticed the massive rout in the stock market, the global panic over COVID-19 and the crash in the oil price. All of this feels familiar to to the 2008 GFC. However, one game company didn't exist back then. Know who that is?

Anyway as we always suggest going by CIG's financials that the cashflow will someday dry up and CRoberts pulls a miracle out his ass and continues to add and break to Star Citizen, is this really going to be the straw the breaks the camel's back when backers lose their jobs and investors are too terrified to invest in something deemed too risky?

I wasn't looking at you, Elon Musk.

[jwh1701]

2020 has a had a very surprising start, thanks in no small part to China and the world's dependence on the world's biggest factory.

I'm sure all of us here have noticed the massive rout in the stock market, the global panic over COVID-19 and the crash in the oil price. All of this feels familiar to to the 2008 GFC. However, one game company didn't exist back then. Know who that is?

Anyway as we always suggest going by CIG's financials that the cashflow will someday dry up and CRoberts pulls a miracle out his ass and continues to add and break to Star Citizen, is this really going to be the straw the breaks the camel's back when backers lose their jobs and investors are too terrified to invest in something deemed too risky?

I wasn't looking at you, Elon Musk.

Sadly even before the corona the feds were pumping billions into the repo market and the djit, rut on my tradeview were showing a lot of weakness. I wonder if this could affect the bottom line for CI and I suspect it definitely could and just like the gov they could use this.

[wiser3754]

For all those who've been wondering where update 3.9 is, look no further.

TheEradicator, whom I believe is a senior software engineer living in Taiwan breaks down  the elephant in the room. Spoiler, it's server-client desync. He believes it may be related to the bastardized form of SSOCS and this blocker has been reeking havoc with the evocaties for a while now.

[dsmart]

@Derek

About that All About Alpha 3.8. video.
Just skimming the video a bit, as some site question, when Todd goes into the problems of their current AI and how that's steming from the backend not being able to keep up. I especially wondered about him mentioning their inability to simulate actual player impact on the server (7:29). For
1) 50 players doesn't sound like something that should bother most servers, when dozens of local host games can go into that direction.
2) I'd read some of Ron Gilbert blog posts for Thimbleweed park and how he went about to simulate a player by some very basic bot, shouldn't such a setup be enough to actually stress a server enough to catch a lot of obvious problems? I mean they've got all 'map', packet and Co. data, so doing a basic externally run bot who's "playing" should be easy enough to do right?

As always i'm not a programer or anything, but that sounded rather strange to me.

It's all rubbish - and lies. Not even going to waste my time explaining it. Most of us who have developed multiplayer games know how to use bots to simulate users.

[dsmart]

*moved*

Ewww. Well maybe this didn't come as a surprise. Having to work remotely on a AAAA title on a Cryengine derivative is sure to cause problems.

After posting this and coming back I just noticed that this post is in the wrong thread.

[dsmart]

ROTFLMAO!!

Roadmap Roundup | April 17th 2020

[wiser3754]

A knick-knack shop is being added in New Babbage. This reeks of microtransactions.

Priority is being put on server meshing for Q2. That buzzword was introduced back, when? 2015? Didn't a whale visiting CIG in Austin sell out when he relised this tech was way too far fetched to be achieved?

Server to client network refactor delayed to 4.1. Seeing in spectrum one post underneath the headline post was very disappointed in this announcement because he's sick of players desyncing and jittering all over the place.

Crusader and port orrison delayed till the end of the year. No doubt being uses for CitizenCon. Stanton still remains unfinished.

340 million in development and this is how far they've come.