berkut

6 ms·

31.
by berkut 3y ago
I think they're overselling it, but Unreal is fairly heavily being used in the early parts of pipelines at various studios, certainly more than Unity engine itself... i.e. MPC had significant use of Unreal on set for mocap, Framestore
32.
by berkut 3y ago
https://pixelesque.net/blog/ (A newly-hosted - February 2023 - re-creation of two blogs, one archived I used to have on blogspot/wordpress).
33.
by berkut 3y ago
Which is interesting, because for MC simulations for rendering (path tracing for example) fast and high quality random numbers are very important... Although there are alternatives like using low discrepancy sequences (Halton, Sobol) which
34.
by berkut 3y ago
It also connects to many "Discovery" servers on the internet in its default configuration, which you don't need enabled if you're just syncing between machines on an internal network.
35.
by berkut 3y ago
If you actually look at the quote and not the flamebait title, he says the "experiment" of there being no loss to creativity with remote work is over, and that the tech isn't there yet for early-stage startups.
36.
by berkut 3y ago
Oooooh :) Thanks!
37.
by berkut 3y ago
Yeah, it's mildly annoying there's no (at least to my knowledge?) compiler hint like the '__builtin_expect' ones to tell compilers that the values are very unlikely going to be predictable with enough accuracy to allow t
38.
by berkut 3y ago
> "The reason you mainly want branchless code is for security reasons where a timing sidechannel could reveal information." Erm, that likely depends on the industry you're in? In HPC, branchless algorithms are often needed
39.
by berkut 3y ago
MySQL InnoDB does I believe...
40.
by berkut 3y ago
Seems better with abstract stuff than specifics, at least in the very limited tests I've done: https://imgur.com/a/tJ8VRa5 It seems to think a Seal is a Penguin, although it is front-on and arguably could be a chi
41.
by berkut 4y ago
Yes, but normally 4- or 8-wide is the norm: the wider you go the more sorting you have to do to traverse things in order or find the nearest hit which has an overhead (hardware may help with this, but it's still an overhead). Previous
42.
by berkut 4y ago
Those are generally being used on much smaller productions, or at least "simpler" fidelity things (i.e. non-photo CG animation like Blizzards's Overwatch). So for Pixar/Dreamworks style things (look great, but not phot
43.
by berkut 4y ago
If you only ever-so-slightly perturb paths, you generally don't get anywhere near as much of a benefit from monte carlo integration, especially for things like light transport at a global non-local scale (might plausibly be useful for
44.
by berkut 4y ago
Sorry, I mis-understood: I agree.
45.
by berkut 4y ago
In VFX at the high-end (the smaller studios do often take shortcuts) we do model full layered materials with thickness, at least for mid to hero-level assets, so we would do the glass with proper volumetric thickness, with fully-correct int
46.
by berkut 4y ago
SaveAs dialog appearing in the background in VSCode ( https://github.com/microsoft/vscode/issues/146422 ) is insanely annoying...
47.
by berkut 4y ago
I suspect that was really 1 GB? 3.5" HDs of over > 20 MB for IBM PCs were around in IBM PCs at the time.
48.
by berkut 4y ago
Their F-18s can't fly as high (without doing zoom climbs) as the F-22 (which has very large control surfaces which help with this).
49.
by berkut 4y ago
For unique_ptr, in a release build, no: but in a debug build there's overhead, and more to the point, stepping into the deference in gdb/lldb at least in my experience requires stepping through the internals, to the point some of
50.
by berkut 4y ago
IE 4 and 5 were innovative (IMO as someone who used both at the time - 1998-2000 - and actively converted family members to IE) compared to Netscape: it had a cache which worked consistently (important in 28.8 modem times) - Netscape would
51.
by berkut 4y ago
status.office.com had been down for 15 mins, but it's back up now...
52.
by berkut 4y ago
While I think that's somewhat true to a degree, I think that might be largely because where checkboxes are used (on desktop), there are normally OK or Apply buttons which do the actual action, whereas where toggles are used, things are
53.
by berkut 4y ago
It's "two" clicks on a very simple node network: if you look at some of the > 1000-10k node networks used in VFX for things like compositing/image editing and see artists re-arrange networks, it all adds up cumulat
54.
by berkut 4y ago
Seems a bit odd to have a 16-bit workflow (clearly not full float32 then?) if it supports EXR files, which natively support 32-bit float (in addition to half float)? Given they appear to be using GLSL for the pixel processing, surely 32-bit
55.
by berkut 4y ago
That's only available on the Pro phones though...
56.
by berkut 4y ago
I've got a ~9k loc Bevy "game" I'm writing that doesn't use dylib and takes > 20 secs every update, purely because of the linking (I'm not using mold yet, but am using lld)... i.e., I type 'cargo build&
57.
by berkut 4y ago
Yes, muscle memory :)
58.
by berkut 4y ago
Gold's pretty slow by today's standards! - it was fast in say 2009 compared to ld, but not really anything to write home about these days. lldb and mold (mentioned in article) are much faster.
59.
by berkut 4y ago
In my experience of building large C++ projects on machines with >= 64 threads (especially if using ccache or something), linking is often the bottleneck to incremental compilation as opposed to cpu-bound compilation. Some of our dev bui
60.
QBittorrent v4.5.0: The Hitchhiker's Guide to Legible Text(datalars.com)
224 points by berkut 4y ago|100 comments

More ›