HNHacker News
TopNewBestAskShowJobs

lonjil

675 karma · joined May 2, 2017

submissionscomments
lonjil··on Moving the Linux Kernel to Modern C
No, no malicious code passed. All malicious code was rejected before making it into the kernel.
lonjil··on Valve’s Steam Deck wows reviewers: ‘The most innovative gaming PC in 20 years’
> The performance, poor as it is, requires the most expensive model at $630.

Why does it require the most expensive model?

lonjil··on Valve’s Steam Deck wows reviewers: ‘The most innovative gaming PC in 20 years’
near hour long discussions with game performance tests, thermal imaging, airflow testing, etc, don't seem to be the kind of stuff you're talking about
lonjil··on Raspberry Pi Direct: buy RP2040 in bulk from just $0.70
There's no privately known way of writing to it either. Mask ROMs are set in stone.
lonjil··on “Risc V greatly underperforms”
"thumb-2" isn't really a thing. It's just an informal name from when more instructions were being added to thumb. it's still just thumb.
lonjil··on “Risc V greatly underperforms”
Thumb has both 16-bit and 32-bit instructions.
lonjil··on “Risc V greatly underperforms”
> Genuinely asking, why? Do we think RISC-V should, or even could, try to compete against the AMD/Intel/ARM behemoths on their playing field?

Well, it's exactly what many RISC-V folks are trying to do. There's news about a new high performance RISC-V core on the HN front page right now!

> but it feels like the architectural decisions we are "stuck with" today are inextricably intertwined with their contemporary market conditions and historical happenstance. It feels like all the experimental architectures that lost to x86/ARM (including Intel's own) were simply too much too soon,

I just want to note that ARM64 was a mostly clean break from prior versions of ARM. Basically a clean slate design started in the late 2000s. It's a modern design built with the same hindsight and approximate market conditions available to the designers of RISC-V.

lonjil··on “Risc V greatly underperforms”
Most commonly used x86_64 instructions decode to only 1 or 2 µops, thus often also just as "complex" as the original instructions.
lonjil··on “Risc V greatly underperforms”
The ARMv8-M profile is Thumb-only, so on ARM microcontroller platforms there is no switching at all, and it does do everything, or at least everything you might want to do on a microcontroller, and has of course gotten a very large amount of use, considering how widely deployed those cores are.
lonjil··on Alpine Linux 3.15
The fix has now been pushed to the 3.15 branch.
lonjil··on Alpine Linux 3.15
> The main advantage is the absence of bloat, e.g. components like DBus

? Like most distros that start you off but nothing but a minimal base, you can choose whether to install DBus or not. Almost everyone using Alpine as a daily driver on their PC probably has DBus installed and enabled.

lonjil··on Alpine Linux 3.15
As long as you're OK with the limitations imposed by using musl, I think it's great on the desktop. I recently switched from Void Linux (glibc variant) and it's working very well. I run proprietary applications like Spotify, Steam, and Discord in Flatpak, and most of the rest from Alpine's repos.

One thing that surprised me is that KDE's software center gives package update notifications for Alpine packages OOTB, and the GUI Just Works(tm) for installing and upgrading.

lonjil··on Alpine Linux 3.15
It is designed for desktop use, actually. Many of Alpine's developers use it as their primary PC distro :)
lonjil··on Consider LibreSSL as default OpenSSL provider again
It seems like Alpine's attempt at upgrading to OpenSSL 3.0 caused them a fair bit of stress, only a bit of which was caused by applications still using ancient deprecated APIs.
lonjil··on Consider LibreSSL as default OpenSSL provider again
Ehh, it doesn't seem very hasty to me. The proposal is for Alpine 3.17 or 3.18, which are 1 and 1.5 years away respectively. That's plenty of time to consider the pros and cons, try it, and revert if it ends up being a poor choice.

> Seems like this stems from https://github.com/openssl/openssl/issues/16660 which contains the full backstory and some more information.

I wouldn't call that the "full backstory", since it is only about a single issue outlined in this proposal. Though she does explain that fact in that GitHub thread too, and also why they're looking into alternatives:

> Given that the OpenSSL 3 migration had an outcome where our contingency plan came into effect, I believe it to be the most prudent course of action to evaluate all possible options before committing to trying the OpenSSL 3 migration again, such an evaluation would be required by the TSC anyway.

> and instead the author chose to write on Twitter and in AlpineLinux GitLab about the frustrations.

Eh, I wouldn't read too much into short twitter vents. As for the Alpine Linux GitLab, well, she's the head of Alpine's security team and a member of the TSC, so coming up with proposals as for what should be investigated w.r.t. security critical packages is one of the things she's supposed to be doing.

lonjil··on ISO3103 Tea Preparation [pdf]
It's not about comparing quality, it's about comparing consistency. Brands want their blends to be consistent from year to year. So every year they might create hundreds or thousands of different combinations of the specific teas they've acquired, until they find one that tastes as similar as possible to the blend they had last year.

Also, green tea is just fine at boiling, if it isn't of poor quality :)

lonjil··on ISO3103 Tea Preparation [pdf]
For most kinds of tea, boiling is fine or even preferred if the tea is of good quality. I would never steep any of the white teas I own below 90°C, and most days I prefer boiling. Likewise with green (and yellow) tea, unless Japanese. Oolongs and blacks I always steep at boiling.
lonjil··on Apple Joins Blender Development Fund
For Cycles, Vulkan is simply not an option. Metal compute is capable of everything OpenCL is, but is more modern / nicer / whatever, while Vulkan Compute is pretty crap and definitely can't be used as a replacement for OpenCL.

The only cross platform alternative that actually works is OpenCL, but Blender is moving away from OpenCL, mostly because literally every GPU vendor is doing so as well.

lonjil··on Stop Using Encrypted Email (2020)
> doesn't understand the operation perfectly IS going to be black-bagged by their local black-ops group

More likely, they simply aren't noteworthy targets. If you rely on PGP to secure your emails, those emails will probably end up not actually secured, but you won't notice because your emails aren't valuable anyways and are probably secured by other stuff, like TLS, and Google's server security.

lonjil··on Witchery: tools for building distroless images with alpine
Because then you'd need to make literally everything you use support the unikernel. Containers, and VMs running Linux, are nice because everything already works on Linux. Also, I don't see what you would actually gain. Like what layer is actually being removed here?
lonjil··on The past, present and future of image formats
JPEG XL is really new, so it would be a lot more surprising if it did have support anywhere yet. Hell, AVIF is pretty new too.
lonjil··on New UUID Formats – IETF Draft
Just put an index on whatever nice sortable unique field you happen to have, then?
lonjil··on Valve Steam Deck
Steam Link and Steam Controller were dropped because they weren't selling very well, but the steam link application continues to be updated, and the Steam Controller continued to get firmware updates until 2018.

If this Steam Deck thing sells well, they'll probably make another one. If it sells poorly they'll probably keep the software updated and use put discounts on it to clear out stock.

lonjil··on Valve Steam Deck
I think it's about as fast as a regular hard drive. So I guess use that as a point of reference.
lonjil··on Valve Steam Deck
SD class UHS-I is about as fast as a SATA HDD, so it will be quite a bit slower than your SATA SSD.
lonjil··on SQLite: Vulnerabilities
Note also that the SQLite devs, chasing 1% perf improvements every year, continually rewrite huge chunks of the entire library, sometimes replacing entire components. Their comprehensive testing ensures that there's no API or ABI breakage, and of course the fuzz testing does catch many bugs, it also ensures that SQLite always has a constant flow of new bugs.
lonjil··on Understanding thread stack sizes and how Alpine is different
If you need a bigger stack, you can get one. This stuff is merely the default.
lonjil··on Understanding thread stack sizes and how Alpine is different
> What, then, is the guaranteed thread stack size?

If you need a big stack, you can just ask pthread to give you one.

lonjil··on Understanding thread stack sizes and how Alpine is different
So in your view, having even a single GNU tool installed, or even available for installation, means you're using "GNU/Linux"? Alpine uses musl and busybox rather than the more common GNU equivalents.

Kernel overcommit has nothing to do with GNU, but default stack size has a lot to do with which libc you use. Musl has a different default than glibc. Overcommit is mentioned because it is the justification for glibc having a large stack size by default. Musl has defaults that make fewer assumptions about how the system is configured.

lonjil··on Why it took us thousands of years to see the colour violet
I'm pretty sure the Med can actually be a bit reddish dark like wine sometimes.
← PreviousPage 6 of 7Next →