@dominikg@mastodon.gamedev.place avatar

dominikg

@dominikg@mastodon.gamedev.place

Engine programmer at heart, currently working on a variety of projects.
Former CD Projekt RED Engine Programmer.
Former Wargaming Sydney Developer.
Many other Software Engineering jobs outside of games.

This profile is from a federated server and may be incomplete. Browse more on the original instance.

pervognsen, to random
@pervognsen@mastodon.social avatar

There's a bunch of C-like successor languages that say they want to eliminate undefined behavior. I've never been able to figure out how they intend to deal with reads and writes to memory since a lot of these languages take what I would call the "naive" machine-centric view of memory which is hard to reconcile with source-level semantics for variables, etc. You can't really rename all of this stuff as "implementation-defined" and get out of jail for free.

dominikg,
@dominikg@mastodon.gamedev.place avatar

@pervognsen
I thought one of the reasons for undefined behaviour was so compiler vendors could more easily adapt C to the target machine. Things like 1's complement vs 2's complement, Harvard vs Von Neumann architecture, etc. It was undefined in the standard so it could be made implementation specific.

Granted these are old reasons, but these days it would be things like strong vs weak memory ordering like you mention.

castano, to random
@castano@mastodon.gamedev.place avatar

Is there a way not to burn quithub's LFS quota on github actions?
I have a repo with GBs of LFS data, and it looks like the checkout action downloads the entire thing every time it's triggered.
This is not only is extremely wasteful, but also gets expensive very quickly!

dominikg,
@dominikg@mastodon.gamedev.place avatar

@castano @MouseByTheSea @wolfpld
I'm currently evaluating self-hosted Git servers at home, trying to decide between Gogs, (its fork) Gitea, and Gitlab.
So far they're all pretty similar for in storing stuff within Git, main evaluation point will be how they integrate CI/CD systems.
Gogs has the most basic one, Gitea has a runner but geared towards Linux, and Gitlab has the most extensive one but I haven't tried to automate Windows builds yet.
That being said I want to try SVN and Perforce too.

dougbinks, to random
@dougbinks@mastodon.gamedev.place avatar

A good reason to use colour rather than color in your code is that colour has the same number of letters as both albedo and normal so you can align your code better.

dominikg,
@dominikg@mastodon.gamedev.place avatar

@dougbinks Color for types, colour for names.

Donzanoid, to random
@Donzanoid@mastodon.social avatar

Are there any state of the art references for fast lexing/parsing on modern machines with SSDs and lots of RAM?

I'm interested in tight inner loops, good BTB use, no data stalls and minimal IO bottlenecks. Is the STB doc still the best reference?

@pervognsen any ideas?

dominikg,
@dominikg@mastodon.gamedev.place avatar

@Donzanoid @pervognsen I've found that looking at the Carbon language code to be quite useful, though since it leverages LLVM it ends up missing the last half of the process of optimisation and code generation. Also viewing @chandlerc data driven compiler talk was good inspiration.

ca1ne, to random
@ca1ne@mastodon.gamedev.place avatar

> With Recall, locating files in a large download pileup or revisiting your browser history is easy. You can give commands to Recall in natural language, eliminating the need to type precise commands.

"We suck at UX so we'll need to record everything and run on a 40+ TOPS CPU to be able to provide even the most basic functionality. Do you like this, is this good?"

https://www.windowslatest.com/2024/05/20/microsoft-confirms-windows-11-recall-ai-hardware-requirements/

I'm sorry, but AI is mostly just doing simple things, incredibly inefficiently, in the most creepy way.

dominikg,
@dominikg@mastodon.gamedev.place avatar

@dotstdy @ca1ne @pervognsen Is there a program like everything which you can just point to a directory and it will index the contents of the files in that directory structure? Not just the names but the contents.

pervognsen, (edited ) to random
@pervognsen@mastodon.social avatar

I was reminded of this design topic by a @vurtun post. The traditional UI paradigm for scrolling/pagination is still based on discrete indices or layout coordinates. For this kind of pagination there's a lot of advantages to instead using a semantically meaningful sorting key. It's inherently more stable with respect to changes in the underlying set, e.g. indices can shift around if new items are inserted, and layout coordinates are even more unstable if the size of layout elements can change.

dominikg,
@dominikg@mastodon.gamedev.place avatar

@pervognsen Isn't that what implementing your own models for Qt views gives you? Though admittedly they are using row & column in their identifiers in the interface, but you are still free to implement your own translation layer on the model code.

grumpygamer, to random
@grumpygamer@mastodon.gamedev.place avatar

I added chests in addition to barrels... the feature creep begins.

dominikg,
@dominikg@mastodon.gamedev.place avatar

@grumpygamer are the classic crates coming next?

pervognsen, (edited ) to random
@pervognsen@mastodon.social avatar

Here's an idea. Does anyone disagree?

(Specifically for Rust but more or less the same in languages with similar reference/view-based slice semantics. I also haven't really convinced myself, which is why I'm asking what people think.)

Don't write s[i..i+n]. Do write s[i..][..n].

Edit: I don't think either of these is "the right thing" compared to calling a length-based subslice function.

dominikg,
@dominikg@mastodon.gamedev.place avatar

@pervognsen In an ArrayView container I implemented I added both SubView and Slice functions which would return another ArrayView. The SubView would take in offset & length, while the Slice would take begin & end.
Makes much more sense to have both as explicit functions as sometimes you have iterator-like limits and sometimes offset and length.

sinbad, to random
@sinbad@mastodon.gamedev.place avatar

I actually didn’t know how deliberately unprofitable Unity was, I assumed they had modest profits before the VC nonsense, but apparently not. Billions of debt. They did the thing of burning through money to wipe everyone else out and get into a position of dominance, before hitting the screw-tightening button on their captive customer base. You know, the standard late capitalism shitty monopolist manoeuvre. https://www.gamesindustry.biz/unitys-self-combustion-engine-this-week-in-business

People trying to make profitable engines never stood a chance

dominikg,
@dominikg@mastodon.gamedev.place avatar

@aras @sinbad profitable up until it got VC funding?

dominikg, to random
@dominikg@mastodon.gamedev.place avatar

For all this talk about people leaving Unity, it's a pretty sad state of affairs that the only alternative mentioned is Unreal. Seems like the industry is converging on a monopoly provider, then only time until they start to tighten the screws.

grumpygamer, to random
@grumpygamer@mastodon.gamedev.place avatar

No AAA publisher/studio wants to be the first to do layoffs, but once one does they all follow. The fallout always seeps its way into indie dev. Get ready.

dominikg,
@dominikg@mastodon.gamedev.place avatar

@grumpygamer Does that mean it's a good time to start a new studio?

grumpygamer, to random
@grumpygamer@mastodon.gamedev.place avatar

But... Panda is delicious.

dominikg,
@dominikg@mastodon.gamedev.place avatar

@grumpygamer It would also be confusing for those people that eat cars and planes. The ones you used to see on TV.

grumpygamer, to random
@grumpygamer@mastodon.gamedev.place avatar

I much prefer Sublime Text over VS Code, but it feels like Sublime Text dev has stalled. I've spent years writing ST plugins for Dinky. But honestly, I can't think of a VS Code feature I want that ST doesn't have, so it's probably moot. Maybe the AI coding, but that would be more to play with, not sure I'd really use it.

dominikg,
@dominikg@mastodon.gamedev.place avatar

@grumpygamer There's something to be said for "completed" software.
Though I understand the mindset when you see a library or project which hasn't had updates for years. Is it incomplete and abandoned, or just finished/completed?

  • All
  • Subscribed
  • Moderated
  • Favorites
  • •
  • JUstTest
  • mdbf
  • everett
  • osvaldo12
  • magazineikmin
  • thenastyranch
  • rosin
  • normalnudes
  • Youngstown
  • Durango
  • slotface
  • ngwrru68w68
  • kavyap
  • DreamBathrooms
  • tester
  • InstantRegret
  • ethstaker
  • GTA5RPClips
  • tacticalgear
  • Leos
  • anitta
  • modclub
  • khanakhh
  • cubers
  • cisconetworking
  • megavids
  • provamag3
  • lostlight
  • All magazines
    •