Show more
Jess Birch boosted

1. Calculate a "double" 1e-300/1e300 and keep the result in the x87 80-bit register because optimisations are on
2. Test if it's zero - it isn't, it's 1e-600
3. Let's use it, but first it has to be saved to a 64 bit value in RAM
4. Surprise, it's zero now

I'm starting to wonder if revenue figures are hiding an eroding (unit sales) market. About a decade ago (also well into a console cycle), example release day deliveries (from major retailers): Fallout 3 Collector's (lunchbox tin) Edition: £35. Halo: Reach & 2100 "free" MSPs: £38.
Today, top sellers for digital downloads on the front page of PSN: Sekiro: £60 (~$80). BF V Deluxe: £80. The Division 2 (std ed): £60.
Cumulative UK inflation in that time: 25%. Wages have not matched inflation.

There is always a layer of "grass is always greener" going on too. It's natural to gripe about the tools you use rather than the ones you've not. But something about the breathless reveal of which engine a project is using or writing off an entire publisher as "using the wrong engine so their teams obviously can't perform" doesn't feel productive (or true to the complexity of tech decisions).

Something I've noticed happening:
Team releases a game & whispers reach games press about gripes with engine tech.
Press compare the missing features (in a codebase possibly forked several years earlier) with current (or even "coming soon") features in other engines. "Engine X is bad and devs wish they could drop it" starts circulating (which would possibly require rebuilding framework/tooling from scratch again).

As someone who browses the vast majority of the web (outside my regular pool of "web apps") without JavaScript enabled, this sounds sorely needed (your custom j/s loader means I will never see the images in your article).

If an action game is all about a specific vision of difficulty, I want to hear all about how it ensures that on a gaming monitor with 4ms of display lag vs a TV with 90ms. Otherwise it's multiple reaction time difficulty tiers, they're just not based on the wants or needs of each user.
What about adaptive difficulty? Recording detailed telemetry so the game knows how each user's reaction times clump vs different in-game demands for action. Feedback/reward improved times & timing consistency?

Either Nintendo are channelling actual magic or I'm going to be extremely surprised if Zelda BotW VR isn't extremely poor in several ways.
(Is that mock-up footage implying you'll right-stick rotate your head around Link even in VR? Portable mode means it's a 300MHz small Maxwell GPU that has to hit 60fps locked, not the 20-30fps of the base game; 720p LCD was not picked for low persistence; etc)

UK market: 80% of console/PC software revenue is digital (and grew 20% YoY); also corresponds to a -30% YoY drop in pre-owned game revenue (now an absolutely tiny fraction of total economic activity). More details:

GameDev depression, anxiety; Kotaku 

Catching up on things I missed last year: Fugue in Void. Definitely some interesting moments in the hour or so; made me wonder how short you could go (for commercial "experience" titles). Currently on sale on Steam:
Critical LP:

I assume someone is already writing documentation including "top 10 things you're focusing shader resources on that a user won't be able to see clearly the other side of lossy compression".
Visible but subtle optimisations like doing post-processing on low res chroma & full res luma become almost invisible (so definitely more heavily recommended) when feeding it through 4:2:0 chrome sub-sampling. Probably a lot of (more subtle) things to learn if you've not studied video encoding before.

me: *Thinking about all the interesting things we'll do with so much rendering power & the mess some detailed non-photorealistic techniques will make when encoded to YouTube for sharing clips*
Google: "And here's at least 10.7TF of GPU for every player... that we're locking the other side of a low latency video encoder!"

As with the ATVI layoffs, thoughts with the 4% of EA staff (350 people) who got the news today.
Seems like the big publishers are slashing non-art/code jobs despite continuing to make everything a "living game" that requires significant support / community engagement to thrive and return on development investment.

EU Copyright Directive, politics 

I do a lot of builder pattern stuff and while I'm not planning to stop, this is certainly a post worth thinking over to make sure you're not doing so without considering alternatives.

Masto meta 

Mitsoda is writing another VtM: Bloodlines & that's also what Cara Ellison is up to?
I'd not heard of Hardsuit Labs (Seattle-based AAA support studio up to now?) but this seems exciting. Paradox are finally making use of WoD after buying it from CCP.

Masto meta 

I guess I'm going to download the current DXR samples & have a play next month so goal achieved there?
When MS abandoned their fallback layer/made it clear it was never intended for end users, I had somewhat placed DXR on a "check again in 2021" shelf.

I had been planning to write this blog post for a while but having read Angelo Pesce's post just yesterday, there was bound to be a bit of contextualisation/reaction to that which slipped in.

Show more
Gamedev Mastodon

Game development! Discussions about game development and related fields, and/or by game developers and related professions.