Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
jabl
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
28 ms
·
571.
▲
by
jabl
3y ago
RH doesn't prevent you from redistributing the sources of a GPL'ed RHEL binary. That is indeed explicitly allowed by the GPL, nothing they can do about it. What they can do, however, is cancel your RHEL subscription if they catch
572.
▲
by
jabl
3y ago
For a somewhat simplistic explanation, RHEL became very successful as a tool to replace proprietary RISC Unix servers (often running Oracle DB) with x86(-64) and Linux, offering a traditional "enterprise" support model those propr
573.
▲
by
jabl
3y ago
Perhaps auditing tools should do something better than just comparing version numbers as released by upstream? Maybe Ubuntu LTS does less security backports than RHEL, but that doesn't make it better per se (saying this as an Ubuntu us
574.
▲
by
jabl
3y ago
SGI also did a lot of scalability work in that time frame, as part of their migration from MIPS+Irix to Itanium+Linux. In the end they got Linux working on IIRC up to 4096 processors.
575.
▲
by
jabl
3y ago
I thought they were using git.centos.org which used to be the RHEL sources? Centos stream sources are what will become the next RHEL major or minor release. So if(?) RH is adding further stuff on top of Centos stream after forking those cha
576.
▲
by
jabl
3y ago
Wonder if the timing of this is a coincidence, considering that RedHat just announced they're stopping releasing RHEL sources except to paying customers: https://www.redhat.com/en/blog/furthering-evolution-cen
577.
▲
by
jabl
3y ago
For an example, most high performance devices expose functionality as memory locations mapped into the CPU address space. In many cases it's necessary to allow userspace direct access to (part of) the memory the device exposes, such as
578.
▲
by
jabl
3y ago
Interesting that the regulations are done that way in terms of both an average and a maximum. I wonder how the average is determined, how many samples, and what is the size of a batch etc.? > 18% of gasoline’s weight is the BTEX aromatic
579.
▲
by
jabl
3y ago
I would guess the minimalism was a practical choice at the time, given it was a pretty small team making an entire OS. Had Plan 9 ever become some kind of commercial (or even open source) success, I'm sure a find command would eventual
580.
▲
by
jabl
3y ago
Not only does gasoline contain toxic components, it's also incredibly flammable. While using gasoline as a degreaser was a common practice, it's a pretty dangerous habit. Not to mention back when leaded gasoline was a thing, it l
581.
▲
by
jabl
3y ago
To nitpick, at least in the EU regulations limit the benzene content in gasoline to a maximum of 1%. I had a vague recollection the limit was actually slightly stricter in the USA at 0.6%, but maybe my memory fails me.
582.
▲
by
jabl
3y ago
Do we know they are fixed length? They could e.g. use a table with offsets to instruction boundaries in the cache?
583.
▲
by
jabl
3y ago
It's entirely possible for a green threads implementation to schedule the green threads over multiple OS threads, thus making use of multiple cores. The programming language "Go" being a popular implementation of this.
584.
▲
by
jabl
3y ago
OTOH PG suggesting that some young people with zero experience with multithreading pivot to searching for the grail is, just, daft. A topic which computer science PhD's have failed to find a solution to, despite decades of research. No
585.
▲
by
jabl
3y ago
Esperanto is doing (unless they have folded/pivoted since I last checked) RISC-V HW with an AI focus, heavily using the RISC-V vector extension.
586.
▲
by
jabl
3y ago
> My kids love Mr Beast. I'm uncomfortable with it and can't put my finger on it. Same. I'm going to sound like an old-man-yelling-at-cloud, but the content is, well, just mindless drivel. There's a near-endless amoun
587.
▲
by
jabl
3y ago
> You’re using a gish gallop style list of random claims. Am I? Looking back at my posts above it seems my main claims are: - Hydrogen storage isn't a solved problem, unless there's some recent breakthrough I'm not aware o
588.
▲
by
jabl
3y ago
> The "proposal" is simply pointing out that those claims are simply false. Then you can surely point out some peer-reviewed articles detailing this. Merely confidently stating something doesn't make it true. > Hydrogen
589.
▲
by
jabl
3y ago
So what are you proposing, exactly? Hydrogen storage remains an extremely tough problem, liquid, compressed, hydrides etc etc are all both expensive and offer poor energy density. Likewise fuel cells after decades of research remain expensi
590.
▲
by
jabl
3y ago
MTBE/ETBE is banned in many places due to groundwater contamination. But ethanol-less gasoline wouldn't be a disaster either, it would mean the refiners would have to do more processing (alkylation, reformation etc.) to increase t
591.
▲
by
jabl
3y ago
Yes. Another variation on this scheme (which AFAIK the Porsche synthetic fuel plant in Chile is doing) is producing methane from atmospheric CO2 and H2 from electrolysis, then converting methane to methanol (how most methanol is produced to
592.
▲
by
jabl
3y ago
Also when you convert forest into farmland and cut down the trees, the root systems of the trees will, over a period of time, decompose. Annual plants have much smaller root systems.
593.
▲
by
jabl
3y ago
Animals don't have the mental capacity reason about moral philosophy, so we can't really judge them for not behaving in another way. But we humans have, and yes, we can judge other humans for causing unnecessary suffering towards
594.
▲
by
jabl
3y ago
Hopper is the name of the GPU architecture, and Grace is the name of the CPU. Combining them in a device gets you a "Grace Hopper" superchip. (And yes, I'd guess the codenames where chosen back in the day with an eye towards
595.
▲
by
jabl
3y ago
I think it's more due to this device combines a "Grace" CPU and a "Hopper" GPU, thus creating a "Grace Hopper" superchip.
596.
▲
by
jabl
3y ago
This is about the "Grace Hopper Superchip", which has one of the Grace CPU's replaced with a GPU. Thus, 72 Neoverse V2 cores. https://resources.nvidia.com/en-us-grace-cpu/grace-hopper-su...
597.
▲
by
jabl
3y ago
No, classic Cray did have a fixed vector size (4096 bits, or 64 double precision floating point values) both in the HW and the ISA. What it did have, however, which most contemporary SIMD ISA's lack, was a vector length register which
598.
▲
by
jabl
3y ago
That's pretty much how every serious dense matrix (or more generally, multi-dimensional array) representation has been done since, well, forever. It's only C programmers writing "My First Matrix Program" that insist on a
599.
▲
by
jabl
3y ago
Sometime ago I stumbled upon Kingdom of the Wicked by Helen Dale. Haven't read it myself though.
600.
▲
by
jabl
3y ago
Nvlink provides cache coherent load-store access, so the point is actually that it's not DMA.
More ›