That one guy known for the Reloaded modding suite.
```
cargo bench -p dxt-lossless-transform-bc1 --bench transform_standard --features bench
```
Linux E-Core: 50GiB/s
Windows (Any Core): 25GiB/s
Linux P-Core: 3.3GiB/s
WTF? This doesn't make any sense- and I'm not hallucinating, I doublechecked!!
github.com/Sewer56/dxt-...
```
cargo bench -p dxt-lossless-transform-bc1 --bench transform_standard --features bench
```
Linux E-Core: 50GiB/s
Windows (Any Core): 25GiB/s
Linux P-Core: 3.3GiB/s
WTF? This doesn't make any sense- and I'm not hallucinating, I doublechecked!!
github.com/Sewer56/dxt-...
Never forget `Nights: Journey of Dreams`, which wrote (presumably) environment data from stack memory into the padding of .ONE archives.
Which gives us a nice glimpse into the tooling they used.
Never forget `Nights: Journey of Dreams`, which wrote (presumably) environment data from stack memory into the padding of .ONE archives.
Which gives us a nice glimpse into the tooling they used.
I do work from here (remotely), and my non-work code stuff right after.
Maybe I'll game on this one day 😅, I'm always coding soooo much; ahahaha, so often that's a pipe dream!
Oh and I'll eventually need some wall decor that's not certificates.
I do work from here (remotely), and my non-work code stuff right after.
Maybe I'll game on this one day 😅, I'm always coding soooo much; ahahaha, so often that's a pipe dream!
Oh and I'll eventually need some wall decor that's not certificates.
Added support for G502X Plus (including LED) to libratbag.
Bought one, came in today, couldn't rebind it on Linux.
Well, it's now later in the day, and now I can.
Added support for G502X Plus (including LED) to libratbag.
Bought one, came in today, couldn't rebind it on Linux.
Well, it's now later in the day, and now I can.
I made a tool for creating and testing 'lossless transforms' that reorganize binary data for maximum compression efficiency.
Like the BC1-BC3 texture ones I made a while back:
sewer56.dev/blog/2025/01...
'Generic marketing chart' for memes included.
I made a tool for creating and testing 'lossless transforms' that reorganize binary data for maximum compression efficiency.
Like the BC1-BC3 texture ones I made a while back:
sewer56.dev/blog/2025/01...
'Generic marketing chart' for memes included.
It's great, really great.
Only caveat is I don't deep search as much as I used to; so it's a bit of a hard sell in my mind sometimes.
That said, I really, really respect what the folks over there are doing. They're legends.
It's great, really great.
Only caveat is I don't deep search as much as I used to; so it's a bit of a hard sell in my mind sometimes.
That said, I really, really respect what the folks over there are doing. They're legends.
Compiling the `bench` target gives you a nice, unrolled joy of SSE2/AVX2. Compiling the `lib` target, doesn't get you that.
That's where the performance difference comes from!! Same compiler settings, but different output on different target.
Compiling the `bench` target gives you a nice, unrolled joy of SSE2/AVX2. Compiling the `lib` target, doesn't get you that.
That's where the performance difference comes from!! Same compiler settings, but different output on different target.
They both generate the same assembly, but the Rust one is 4x faster when benchmarking; why?🧵
They both generate the same assembly, but the Rust one is 4x faster when benchmarking; why?🧵