zwol

@zwol@hackers.town

Nerd of all disciplines; presently computer science faculty at CMU. Online just barely long enough to remember USENET.

Analog interests include pottery, candlemaking, board games, science fiction and fantasy, photography, and cooking.

New follows welcome but I probably won't follow back because I can't keep up as is.

#fedi22 directory tags: #computerScience #softwareEngineering #infosec
#education #cseducation
#chemistry #candlemaking #pottery
#games #boardGames
#photography
#cooking

This profile is from a federated server and may be incomplete. Browse more on the original instance.

zwol, to random

once again failing to find anyone who will sell me small, metal hooks that mount to wall anchors. they need to project from the wall by only 1/4" (7mm) or so, but the moment "mounts to wall with screws" becomes a criterion, nobody's got anything less than coat-hook-sized.

zwol, to email

I have suddenly started getting a lotof obvious #email #spam making it through the filters on one particular account. Anyone else seeing this?

vertigo, to random

Really getting sick and tired of the "you can't share code if it's statically linked" mantra.

It's a lie that has been spoken so many times that it has taken on the status of truth.

But there exists one very significant falsification to that mantra -- AmigaOS.

This is my yearly reminder to people, it seems, to study exec.library and how OpenLibrary() works. Everything's there.

zwol,

@vertigo @WomanCorn @enkiv2 I had almost the same question: I was wondering how record types were handled.

I would also like to know how library versioning was handled. Usually when I see people talking about static linking as a desirable thing, a big reason is that the version of each library in use is frozen at link time, but with this, it seems like that wouldn't happen

zwol, (edited ) to Pittsburgh

Hey #Pittsburgh ers! I'm giving away the following items free to the first person who will come remove them from my office at CMU, which is Gates-Hillman Center 4124 (close to Forbes and Morewood, DM me if you need more specifics).

I'll be in my office until 3pm today. If you can't make it today, I'll also be here on Tuesdays from 11am-noon and 1-4pm, and Thursdays from 11-noon and 1-3pm, through August 8. If you really want one of these things and you can't make any of those times, DM me and we can work something out.

In descending order of weight:

  • HP LaserJet 4000 with 10baseT JetDirect card. Minor cosmetic damage, leaks toner on printouts, but otherwise in good working order. No packaging.
  • ReadySet brand solar charging system and accessories, in original packaging. Needs a new battery (sealed lead acid, F2 terminals, not sure exactly what form factor) but otherwise in good working order.
  • TaoTronics brand ultrasonic humidifier, the old white and blue model. Needs a new transducer; don't take this unless you have at least decent electronics-repair skills. No packaging.
  • No-name digital Toslink to analog audio converter. Unused, in original packaging; I ordered one and they sent me two. Includes a 1-meter Toslink cable and an USB-A-to-barrel-jack power cable (unknown length) but does not include a power brick (any 5VDC, 2-amp USB-A power supply will do).
  • Tenergy brand T333 battery tester. Works. No packaging.
  • No-name PCIE (1x, I think) to M.2 adapter card. Works. No packaging.
  • Brannan brand wall barometer, roughly 4" in diameter. Works. No packaging.
  • Tech21 brand iPhone 6s case. White. Unused, in original packaging. (Picture on packaging does not match actual case.)
zwol,

Photo of all the items described in the previous toot

endomain, to random

I've been trying to hold the line with firefox but I gotta be honest, it's so jacked out with ads that are hard to disable now that I just kinda feel like my experience would be equivalent with Chrome.

zwol,

@Macross @endomain You say this like finding websites that don't work properly in Firefox is an everyday experience for you, which I find very strange because it basically never happens to me. I can only think of two examples in the past ten years plus:

  • T-Mobile (US)'s customer portal, a recent break which all by itself is making me seriously consider switching phone providers;

  • The landing page for threads dot net, which appears to have fallen foul of a built-in privacy rule to the effect of "don't load Facebook's tracking pixels" and, yanno, fair.

zwol,

@Macross I was not trying to discredit or discount you, I genuinely wanted more detail. I'd still like to know specifically which websites you were having trouble with

zwol, to random

It has been a long time since I posted a data visualization with no explanation or context, hasn't it? Way too long.

zwol,

Incidentally, if anyone is clever enough to know how to take the SVG version of this and automate turning it into a modified SVG in which the middle trend line is continuous and the upper and lower trend lines are replaced by a ribbon with the same limits, in which each vertical stripe is a gradient from the upper color to the lower color (the colors will need to be tweaked so the middle trend line is still visible), please get in touch. I'm not sure this will be an improvement but I need to see it to know for sure, and ggplot can't do it.

zwol,

I figured out a better way to visualize it

thegibson, to random

I don't much care for this new local timeline not refreshing stuff, and trying to move me to live feeds for that.

zwol,

@vertigo @thegibson I thought the site was draining my laptop's battery faster than it used to

zwol, to random

Not only is this end table disturbingly top-heavy, the giant marble brick is held in place only by gravity. You can lift it right off.

I have ... concerns.

b0rk, (edited ) to random
@b0rk@jvns.ca avatar

what's a popular command line tool (other than git) that you wish had a clearer / more intuitive UI? I'm thinking of tools like dig which has this IMO pretty arcane output format

zwol,

@b0rk Nth'ing jq. With that one I think the confusing syntax is a reflection of a data model that doesn't fit the way I think about tree-structured data.

Also: imagemagick; matplotlib (notionally a library but you're supposed to be able to use it interactively); pulseaudio's control CLI (category: tools you only need if something is broken, so you have no mental model of how it works or what "normal" looks like, and the documentation is terrible).

zwol,

@b0rk My students struggle so much with gdb and I learned it so long ago that I don't grok what makes it confusing for them, or even to what extent it's C they're actually confused about.

zwol,

@b0rk @snorerot13 I have no evidence for this other than what it looks like, but I suspect a design goal for dig output — possibly the only design goal for dig output — was that you could copy and paste it into a BIND zone file.

edsu, to random
@edsu@social.coop avatar

I really like this positive statement about who @SocialCoop federates with, as a foundation for de-federating from Meta's threads.net when it comes online.

https://cryptpad.fr/pad/#/2/pad/edit/5xxQZvxvjfAP2eQTYenLuWOZ/embed/

Thanks @edumerco for getting this started! I thought it could use a bit of adjustment to also promote an open fediverse where new instances could come online easily.

zwol,
vertigo, to random

I find it both tragic and amusing that Intel tried to do the right thing in processor design thrice. And failed all three times.

The iAPX432 processors were late to market, so the 8086 and 8088 were commissioned to fill the gap. It turns out that the 8086 was measurably faster than the iAPX432 when it did finally deliver, so the 8086 won out. The i432 was basically DOA.

Intel tried again with the i960 family (see http://www.righto.com/2023/07/the-complex-history-of-intel-i960-risc.html), which is a RISCy version of the iAPX432. Unlike its predecessor, it was actually fast enough to be useful. But, the 80386 just destroyed it.

And the third time was the titanic Itanium architecture, which the Pentium family summarily humiliated.

Sometimes, I can't help but kind of feel sad for Intel.

zwol,

@vertigo Having had to write a tiny piece of GCC's back end for Itanium, I object to the suggestion that that design was in any way a good idea

zwol, (edited ) to random

Here's something I found in the aforementioned birdsite account archive and I think it would be fun to repost here (lightly edited). It's really long so I'll just put the intro here and the rest of it in an unlisted reply.

Imagine that you are an experimental relativistic physicist. You do stuff like Gravity Probe B <https://en.wikipedia.org/wiki/Gravity_Probe_B>. One fine morning, an acquaintance from the electrical engineering department knocks on your door holding two black boxes. "My grad student was futzing around with antenna design and built these", they say. "They seem to do something impossible."

zwol, (edited )

The outside of each black box is a standard CubeSat module, so 10cm x 10cm x 11.25cm, and they don't weigh much. They are completely featureless except for the CubeSat mounting rails and the following set of connectors:

  1. 2x banana jacks, red and black, marked +12V and GND, respectively. Nearby it also says MAX LOAD 5 AMPS.
  2. One BNC twist-lock connector for an RG58/U coax cable. On one of the cubes this is labeled SIGNAL IN MAX FREQ 10MHz. On the other one it is labeled SIGNAL OUT.
  3. One standard slide switch labeled ON/OFF.

"You put a signal in this one and it comes out the other one," your buddy says, "and here's the thing, we can't measure any propagation delay at all, not even when we put 1000 km in between them."

Being an experimental relativistic physicist, you don't have to do math in your head to know that that is impossible. A signal has to take at least 3.336 milliseconds to travel 1000 km.

... Or does it? Technically, relativity does not say it's impossible for any signal to propagate faster than the speed of light. It just says that, if a signal appears to propagate faster than c in one reference frame, then there is another reference frame in which the same signal appears to travel backward in time. Everyone assumes this means faster-than-light signaling is impossible, because everyone assumes backward time travel is impossible.

And we assume backward time travel is impossible because if you could do that you could set up time paradoxes.

Well, here's this pair of black boxes that sure seem to be communicating faster than light. You're an experimentalist. Time to get experimenting.

Can you, using these boxes plus any amount of stuff we already know how to build, perform an experiment that would cause a genuine time paradox (e.g. a signal enters the IN box if and only if it does not enter the IN box), if the boxes really do communicate faster than light, and relativity's prediction of what happens in this case is accurate? If so, how do you do it?

Failing that, can you perform an experiment in which a real observer in some reference frame does collect data that implies the signal from one box to the other went backward in time, and if so, how do you do that?

Get as fancy with "stuff we already know how to build" as you want. Money no object. Put the boxes on big complicated satellites in solar orbit if you need to. Just no more unobtainium besides the boxes themselves, and current-day constraints on measurement precision.

zwol,

@jamey Someone told me that they could, indeed, construct a time paradox under these constraints, but they didn't explain how.

zwol, to random

Poking around at the contents of my birdsite account archive, courtesy @darius 's make-the-archive-readable tool, and I am realizing that what I want is to go through the unredacted version of its contents, pull out everything that is still worth reading, and put those up on my website as static, unrolled threads.

I need the unredacted version for this, because a lot of the time something worth reading starts as a reply to someone else, and I get why the existing tool doesn't show anything that was a reply to someone else, but it does show self-replies to those replies and that means I can find the good threads but the first sentence of whatever I was saying is missing :-(

zwol,

So, um, anyone know of a tool that lets me browse my unredacted archive on my local machine?

@darius

zwol,

@darius It does? I see no replies at all in the default 'tweets' view and I don't see any way to make them visible either

fade, to random

When the smoke's in the sky
with a high AQI
that's a Thursday.

zwol,

@fade oof.

(also, hi! I think we talked a couple times long ago on LJ. It's nice to find old acquaintances in new places)

zwol,

@fade This is the same icon I used back then, and I think this is the first time anyone has recognized me by it. :ablobcool:

shauna, to writing
@shauna@social.coop avatar

In 2022 I wrote 154,966 words - that's one novel, and three short stories. I don't have much to share here, as only one of the short stories is published, but I'm proud of what I've accomplished anyway. :)

The published short story is called Sunlight and it's about two women who own a coffee shop struggling with a change in governance that threatens to destroy their business—and their marriage.

You can find it here: https://medium.com/after-the-storm/sunlight-cdb9bb0be8bc

Excerpt attached!

#writing

zwol,

@shauna this is a beautiful story and I love how much more it implies

  • All
  • Subscribed
  • Moderated
  • Favorites
  • JUstTest
  • kavyap
  • DreamBathrooms
  • thenastyranch
  • magazineikmin
  • tacticalgear
  • khanakhh
  • Youngstown
  • mdbf
  • slotface
  • rosin
  • everett
  • ngwrru68w68
  • Durango
  • megavids
  • InstantRegret
  • cubers
  • GTA5RPClips
  • cisconetworking
  • ethstaker
  • osvaldo12
  • modclub
  • normalnudes
  • provamag3
  • tester
  • anitta
  • Leos
  • lostlight
  • All magazines