HNHacker News
TopNewBestAskShowJobs

PreInternet01

2,490 karma · joined December 13, 2022

submissionscomments
PreInternet01··on Extinction of the Influenza B Yamagata line during the Covid pandemic (2022)
Well, the strains that I listed are in the Sanofi Quadrivalent vaccine that is most popular in Europe (and being administered right now) as far as I know...
PreInternet01··on Extinction of the Influenza B Yamagata line during the Covid pandemic (2022)
Yeah, influenza type B 'Yamagata' has been mostly gone since 2020, but that's not necessarily because of COVID, but just because the virus is so variable. In fact, it's highly unlikely for an influenza sub-type to hang around for more than 18 months.

In Europe, for 2024/2025, types A 'Victoria' and 'Thailand' (with type-B 'Victoria' and 'Yamagata' following closely, so you can see it's still there-ish) are expected to be the most virulent. but, as always with influenza, this might be wrong, which means vaccination might be slightly less effective, although still beneficial.

PreInternet01··on Developer pockets $2M in savings from going cloud-free
Yeah, I don't get these articles (with this one being pretty similar in tone to the 'Hetzner is 4x cheaper than AWS' one I saw on here earlier today).

If all you need is simple compute and bandwidth, and you can spare the resources to admin a server, of course you'll be able to find colocated, dedicated-server or even VPS solutions that fit your requirements and are way cheaper than AWS/Azure/GCS/etc.

The whole selling point of 'cloud' is that you can scale infinitely at a moment's notice and have a whole array of batteries-included services at your disposal. Don't need that? Don't do it, even if it's fashionable...

PreInternet01··on Gang of Raccoons Mobs Woman in Casper Hotel Parking Lot
There are definitely signs: https://www.naturepl.com/stock-photo-baboon-warning-sign-cap...

But, as with most situations, "let's put up a sign" somehow didn't resolve the issue...

PreInternet01··on Gang of Raccoons Mobs Woman in Casper Hotel Parking Lot
Thanks for your detailed explanation of where I went wrong. Correcting non-native speakers on their mis-use of fringe vocabulary sure feels good, doesn't it?
PreInternet01··on Gang of Raccoons Mobs Woman in Casper Hotel Parking Lot
"Leaned behavior" to me means that they previously saw other, non-baboon, groups rob tourists in the same way, which seems unlikely?

This feels more like a situation where the baboons realized that "these humans have yummy stuff" followed by "hey, and this is how we can distract them in order to grab it"

But, well, semantics...

PreInternet01··on Gang of Raccoons Mobs Woman in Casper Hotel Parking Lot
Worst gang-like behavior of animals I every saw was, like, a decade ago at the Cape Point parking lot in South Africa.

This a relatively large lot (being at the round-turn point of a one-way route), with both personal vehicles and tour busses coming and going. There are also a lot of baboons. On the narrow-ish mountain road up and down, these already pester the vehicles every way they can, but once parked, it's when things get truly special, especially for the tour busses.

First, the baboon females parade past a bus, with infants on their back. The tourists are delighted by this, and rush out with their cameras. Then, the females retreat into the bushes, which further distracts the humans.

Then, the males, which have been hiding in the nearby trees, descend upon the bus, and absolutely ransack it. Everything that can be taken (and that's mostly bags and coats) gets dragged into the trees. Then, they triumphantly go through the loot, redistribute anything that's edible or otherwise interesting, and toss the rest.

It's scary how well these animals instinctively understand the way we behave...

PreInternet01··on OpenBSD Is Hard to Show Off
Counterpoint: possibly OpenBSD doesn't really have any relevant 'kernel processes' (in line with the model: the kernel is there to serve user-mode, and not a goal on its own). (And tangentially: why do I need multiple `psimon` kernel processes? Tire pressure is good to keep in mind, I guess, but, at the kernel level? Repeatedly?).

I truly don't know, by the way; this was merely an example to illustrate my perceived difference-in-complexity between Linux and OpenBSD, which might be entirely misguided, but definitely part of the issue the article linked in the submission hints at?

PreInternet01··on OpenBSD Is Hard to Show Off
Sure, why not: https://openbsd-ps-vs-coreos.tiiny.site/

In fact, OpenBSD has gotten a lot more verbose over the years, due to sub-processes being listed explicitly. Good example here is `smtpd`, but even then, the difference with Linux remains... stark.

PreInternet01··on OpenBSD Is Hard to Show Off
So, the article seems to be mostly about desktop usage, but for me, the best way to show off OpenBSD is to just run `ps ax` after a clean install, and compare that to, say, Fedora CoreOS (just because that happens to be the other common OS I use for my server VMs).

On OpenBSD, the list of processes is mercifully short, and I can tell you what each of them is supposed to do. CoreOS? Not so much -- and, yeah, I know, I could of course switch to non-systemd-infested distro, but even there, there would be at least a screen-full of processes of which I could maybe identify half?

This pattern extends to configuration: OpenBSD has a handful of very-well-documented files in /etc, whereas under Linux, things are literally all over the place, and you might even end up editing irrelevant configs, because they happen to be regenerated by yet-another-layer of tooling, or your distro is simply not using the tool you think you're configuring...

Still, I have no clear preference for one OS over the other! With workloads being mostly containerized these days, it's so easy to roll out a new VM, that the level of understanding required to, say, nurse a broken OS on physical hardware back to health, is now simply irrelevant. Just blow away the old and deploy the new, and forget about it.

On OpenBSD, the deployment tooling uses tried-and-true commands like `fdisk` and `ifconfig`, whereas for CoreOS it's YAML (pronounced: 'hell'), but the truth is... it doesn't matter that much. Getting the base OS installed, adding some volumes, creating sub-interfaces for the relevant VLANs, installing Docker(-compose), and getting the containers to (re)start is all very, very similar.

So, yeah, I sort-of get what the article is trying to say, but the ground-truth I think is that in most cases the OS isn't that much of a differentiator anymore, as they're pretty much all OK-ish...

PreInternet01··on Bankrupt Fisker says it can't migrate its EVs to a new owner's server
Sure... "bankrupt-adjacent company with vaguely hard-coded cloud dependencies inconveniences their customers" is a story we've seen many times in the past already, and will continue to see uncountable times in the future, mostly because most people reading stories like these are employed by startups doing the exact same thing.

So, the issue becomes "how to fix this"? This very same site tells me that EU-style hardcore regulation "makes it impossible to run a business", so, I'm open to suggestions, other than "Open Source Everything and Rewrite it in Rust"?

PreInternet01··on Why the ISO format has to die
Yeah, but I'd call that 'remnants of El Torito'? You basically have tooling to burp out the MBR, boot catalog and UEFI FAT32 shim, all pointing at the native code that can read your actual (very-much-non-ISO9660) file system into RAM and bootstrap it, and you're good to go on USB, ISO, or whatever.

The first 3 parts aren't that complicated (literally a few sectors with fixed content, maybe a few pointer fixups), and no reason IMHO to want to kill off ISO. Generating that actual file system is the hard work, but also required for live-USB distros, which circles back to the point being made in the article eluding me...

PreInternet01··on Why the ISO format has to die
What a strange article. What's called an 'ISO' is, these days, pretty much a thin wrapper for an UDF file system -- you can definitely use it differently as well, but I'm pretty sure nobody has created, say, an ISO9660 plus Joliet plus El Torito image in the past decade or so.

And, no, UDF isn't great either, but I wouldn't say it "has to die": it's a pretty convenient distribution format due to being widely supported, as it's really simple to implement.

So, this mostly seems to be a rant against live-CD-style Linux distros, since those are hard on maintainers (plus: toxic community multiplier)? On the one hand that might be true, on the other hand, the 'hey, here is how you get an ext4 or whatever filesystem in RAM' tooling around that is so mature and convenient that it's hard to see why, and I can't distill any real arguments from this...

PreInternet01··on Show HN: A pretty decent PDF to CSV converter
So, for a test PDF I just created in Microsoft Word (Insert/Table, 3 header columns, 3 data rows, then apply formatting to reflect this), it worked pretty decently indeed.

Not sure about the business model or anything else, but from a technical point of view, this sort-of made me smile.

PreInternet01··on Mkfs.fat on Linux vs. OS/2 2.1
No, I'm pretty sure this check was only done when the (root) directory needed to be accessed, which was on the same track as the boot sector, so reading it (which was required for synchronization anyway) was free.

Glancing at https://github.com/microsoft/MS-DOS/blob/main/v4.0/src/BIOS/... seems to support this...

PreInternet01··on Mkfs.fat on Linux vs. OS/2 2.1
> The rest of the "Extended BIOS Parameter Block" looks like it isn't relevant data, like the volume label or the volume's serial number

No, that's the exact reason why OS/2 checks the OEM ID. The volume label was initially used to detect disk swaps (so that if an application has open files on disk 1, it does not destroy data on disk 2 by writes after the user replaces the floppy in the drive, which could not be detected in hardware until much later, and even then not unless you were using a Mac, basically).

Of course, the volume label is user-set (so duplicates are possible), and more often not set than meaningful in any case. So, DOS 3.3 added the (unique-at-format-time) volume serial number. Which was always all-zeroes (or garbage) on pre-3.3 disks, hence the need for the OEM ID check.

PreInternet01··on Turning Everyday Gadgets into Bombs Is a Bad Idea
Very well-written analysis with factual support for three main points about this kind of attack: it's entirely feasible, it's cheap, and it's virtually undetectable.

And the devastating effectiveness has just been live-demonstrated to an audience very interested in that kind of stuff.

So, yeah, air travel is going to become, eh, a bit different, soon. Get those cellphones and laptops with removable batteries while they're still available (and ensure the packs are models you can easily buy at your destination), as anything battery-powered is going to be off-limits in both the cabin and the cargo hold as soon as the now-extremely-motivated international terrorist community gets around to downing the first, oh, three airliners or so?

PreInternet01··on Ask HN: How does my school block DoH?
At least Firefox supports a way for LANs to disallow DoH usage via regular-DNS responses: https://kb.isc.org/docs/using-response-policy-zones-to-disab...

Possibly, Chrome/Edge support the same or similiar mechanisms, but I haven't checked recently.

Alternatively, your school uses a HTTPS middlebox that knows how to distinguish DoH requests from regular SSL traffic to 1.1.1.1.

In any case, this is most likely not an 'attack' on you, and characterizing it as such is unlikely to lead to fruitful discussions around the policy.

PreInternet01··on Super Micro: Fresh Evidence of Accounting Manipulation, etc.
So, yeah, I absolutely love some parts of Supermicro (one word, no InitCaps...), especially the AM5 bits that power most of my servers, but...

Going public in the US was absolutely not the correct choice for them. Other than, quite obviously, not being able to keep up with the accounting requirements (which are, like, 80% of the job for any public company, the other 20% being split 80/20 between 'keeping shareholders happy' and 'doing the right thing'), they just lost sight of their core market, which used to be "cutting edge reference design boards for people who know what they're doing."

Instead, we got "packaged solutions" that included overpriced RAM and SSDs/HDDs. Well, sure, if I'd wanted that, I'd rather go to Dell/EMC! Because they at least have the delivery capabilities, which Supermicro has also been under-performing on for the past years.

So, yeah...

PreInternet01··on Galois Theory
[removed by author]
PreInternet01··on Some thoughts on OpenSSH 9.8's PerSourcePenalties feature
> What is the advantage of bothering to do this?

For SMTP: log optimization.

Like: yeah, I don't care that you're self-DDOSing if you're an actual spammer. But, just in case you're a legitimate sender lacking SPF records, and you just need some help, I still want some kind of signal.

For SSH, I guess you might want something similar, just to catch authorized-yet-misconfigured remotes...

PreInternet01··on Some thoughts on OpenSSH 9.8's PerSourcePenalties feature
PerSourcePenalties-like abuse mitigations are not very subtle, but often very effective.

My own experience with this is not as much with SSH as with SMTP. If your particular IP emits a single non-deliverable message every 30 minutes or so, that's fine.

But: more than a handful of SPF failures for the same sender/recipient combo within 10 minutes or so? Yeah, you're now on the general-deny-list. And persisting even then? Automagically tar-pitted on the firewall, and have fun...

PreInternet01··on Texas sues GM for unlaw­ful­ly collecting and selling dri­vers' pri­vate data [pdf]
Thanks for sharing the exact workflow, and sorry for potentially derailing the discussion on this! (For the record, I upvoted all substantial comments, and truly enjoyed my comment going from +4 to -2 and back again...).

I've never had iOS (either the Cisco or the Apple variant) randomly apply CamelCaps, though. One more thing to look out for in the future :)

PreInternet01··on Texas sues GM for unlaw­ful­ly collecting and selling dri­vers' pri­vate data [pdf]
Well, I'm not a native English speaker, but I'm pretty sure that in "UnlawFulLy" the capitals don't represent pronunciation stress points? Those would look more like "uNlawfullY", right?
PreInternet01··on Texas sues GM for unlaw­ful­ly collecting and selling dri­vers' pri­vate data [pdf]
wHaT's uP With the cApiTaLizAtioN iN this TitLe?

I've spent, like, several minutes of my life trying to figure this out, and... came up blank?

The title is not copied from the PDF, where such issues would be rare but not unimaginable. I also couldn't find any articles linking to this PDF with this particular title.

So... a misfiring accessibility solution? Solar rays directly going for the Shift key?

(P.S. Yeah, I know "Please don't complain about tangential annoyances—e.g. article or website formats, name collisions, or back-button breakage. They're too common to be interesting." -- but this seems sort-of uncommon?)

PreInternet01··on Psssst! Your date of birth can be a random number!
Yes, well, unless the poor CSR cannot actually see your answer, but has to enter it for a 'yes/no' response. Had a fun time with Rackspace support around the turn of the millennium, where I had used the output of 'openssl rand -base64 16' or somesuch, and that had to be entered verbatim, with the person on the other side apparently not understanding what a 'forward slash' was, and, yes, it was urgent...

Anyway, I don't find these questions intrusive as much as 'not fit for purpose', with a bit of (cultural) insensitivity mixed in. I once had to console a very-upset very-serious person from an African country with regards to the the 'name of pet' question they just had to answer as part of a mandatory signup for a US-based third-party service, and which they just couldn't find an answer to.

And whenever I'm asked about "name of high school" or (even worse) "high school mascot", I'm tempted to flip out with something like "yeah, which of the seven-or-so I attended, for no-fun-at-all reasons, and most of which I don't even remember, but I can assure you none of them had a mascot", but then I just take a chill pill and mash my keyboard...

PreInternet01··on Psssst! Your date of birth can be a random number!
Oh, yeah, my password manager is full of answers to 'security questions' like 'place of birth', 'name of first pet' and 'high school name'. Just about the only time that gets awkward, is when the same answers turn out to be required during certain phone conversations. Yes, I do in fact, happen to be born in Qjhdfoudvprx -- quaint little, very traditional village...
PreInternet01··on CrowdStrike releases root cause analysis of the global Microsoft breakdown
Yeah, sure. So we just accept "your holiday will be disrupted" since "IT is just a very competitive sector"?

I won't, but you be you...

PreInternet01··on CrowdStrike releases root cause analysis of the global Microsoft breakdown
> legally mandating multiple vendors for private companies is unnecessary

The days-long airport chaos that most of the world just went through recently would suggest otherwise?

PreInternet01··on Sentry is now Fair Source
It is how Sentry used to work, though, and if you want to forward me the emails proving that, just supply me with your work email.

Good to hear that things changed, though: for me, it's a bit late, but, hey, I'm not your target audience anyway.

← PreviousPage 2 of 14Next →