matja

3 ms·

1.
by matja 6d ago
Depends on what proportion of Samsung's output is current HBM4 and HBM4E DRAM. Everything in their statement can be true and it be a bad thing for consumers of non-HBM RAM. "HBM capacity to expand to 250,000 wafers a month" S
2.
by matja 6d ago
I'm surprised that in no reviews of the iPhone 18 camera there is mention of the bad quality bokeh, it has heavy ring shapes that looks like what you'd get from a cheap catadioptric lens. Edit: e.g. the tree in one of the images i
3.
by matja 7d ago
.len doesn't work on a list (because it might mislead you about the efficiency of the operation if it did exist), so it's better to use iterators - so then you can change the underlying type without changing your code. But then, s
4.
by matja 7d ago
> I'm not sure that nuking 2G of the underlying block device is a recoverable error on any filesystem that I'm aware of? ZFS and btrfs were designed from the start to handle this, by using checksums on every piece of (meta)data
5.
by matja 7d ago
Add: "call_back_day": { "criteria": { "none": "The user does not want a call back.", "today": "The user wants a call back today.", "tomorro
6.
by matja 14d ago
I wouldn't say so, because it was a co-processor: e.g., it depended on the 8086 to even generate addresses for memory operations. I think the IBM 7030 Stretch CPU (from 1964) would be a likely contender for being one of the first 64-bi
7.
by matja 14d ago
> Also, the arbitrary "Oh, and the registers are 80 bits wide" is also just one of those weird "Where did that number come from?". One of the features that was advertised (mentioned in the iAPX 86, 88, 186 Microproces
8.
by matja 14d ago
> I wish the stars would align for QEMU to gain support for sunlight/moonlight streaming directly in the hypervisor. I didn't know I wanted it until you said it. I have a few QEMU VMs where I use PCI-passthrough of a whole GPU
9.
by matja 16d ago
The levels tool maps a subset of the channel range to the full range, effectively changing the brightness and contrast in one step. The center point controls the gamma correction, so you can map a biased distribution (mostly bright or dark
10.
by matja 16d ago
You can do something similar in GIMP: 1) Colors > Components > Decompose > LAB 2) For the layers of the A/B (chroma) channels, Colors > Levels > "Auto Input Levels" (maximize the contrast), and move the mid poi
11.
by matja 16d ago
You're rejecting information solely based on the aesthetic quality of the delivery medium?
12.
by matja 19d ago
> every department is fighting for their lives against the others https://bonkersworld.net/img/2011-06-27_organizational_chart...
13.
by matja 25d ago
Looks like it messed-up and they're just for padding, remove them and it makes sense. I never noticed that before. As such, I propose a replacement :) : To accomplish this, append the following warnings to the software. It is
14.
by matja 26d ago
Obligatory mention, the GNU JUSTIFIED PUBLIC LICENSE: https://tom7.org/bovex/JCOPYING (for the rabbit hole: https://tom7.org/bovex/ )
15.
by matja 28d ago
The more popular your software is, the more likely you'll get bug reports from people with just bad hardware. I know a few applications now that perform a simple memory/CPU test when sending a bug report and reject those that are
16.
by matja 1mo ago
I meant that it's a difficult comparison between platforms without having all the details like memory latency, how the memory controller load increases latency (can increase by 4x on some platforms), cache latency, TLB entries, etc. Ex
17.
by matja 1mo ago
Thanks for this, really interesting parent project too, I went down a rabbit hole on that one :)
18.
by matja 1mo ago
It's higher bandwidth than any dual-channel DDR5 desktop machine, but Apple never quote the memory latency, so hard to compare otherwise.
19.
by matja 1mo ago
It installed and runs perfectly on my VM with an RTX 4080 passed-though using VFIO. It uses the nvidia module rather than nouveau (at least on my GPU).
20.
by matja 1mo ago
For those not on macOS, is there a video?
21.
by matja 1mo ago
Same, having worked on firmware that controlled relays (and NAND chips) and managed to burn them both out surprisingly quickly due to software bugs :p
22.
by matja 2mo ago
That's not very efficient
23.
by matja 2mo ago
Encode every 0 as 10 - then decode every 10 back to 0. Or more efficiently, the first 0 as 10, then decode the the first 10 as 0.
24.
by matja 3mo ago
OAuth is designed so that an end-user never needs to see an API key (OAuth refresh/access token) or even know what one is. When it is implemented to the spec, that happens well. I think that most of the "just give me an API key&q
25.
by matja 3mo ago
Amazon Route 53 is $0.40 per million DNS queries - which would terrify me if I used it, considering a typical 10Gbit server connection hosted at a unscrupulous ASN with no egress IP filtering is capable of sending a million DNS requests per
26.
by matja 3mo ago
> I feel like IPv6 is overengineered In what metrics? IPv6 is more simple to implement than IPv4. In Linux 7.1.1 IPv4 is 84kLOC, IPv6 is 59kLOC.
27.
by matja 3mo ago
I can think of a few reasons: There wasn't any such features for x86 when the patch was created, other than AES-NI. Many hardware platforms that have TPM, have it connected via a low-bandwidth LPC bus which would have nowhere near enou
28.
by matja 3mo ago
There was a patch called Tresor that did this, but I don't think it was updated for a long time. You have to store the encryption key in CPU registers and ensure it's not saved to RAM during task switching or power suspend operati
29.
by matja 4mo ago
Apt comic: https://www.workchronicles.com/p/comic-prevention-vs-cure
30.
by matja 4mo ago
> well that is fortunate I think that was one of the arguments of the Anthropic principle [1], that there doesn't appear to be any reason why there are 3 spatial dimensions and 1 time dimension, or why the fundamental constants are

More ›