lewiscowles1986

lewiscowles1986, 7 days ago (edited 4 days ago) to random

Has anyone else experienced non-official package registry corruption / poisoning?

Recently at work, I've had to disable a poetry option to range-scan (introduced in poetry 1.8); and ensure that pypi is present in all our pyproject.toml

We are using AWS CodeArtifact. I'm unsure if it's the registry, or if using the range-scan headers (on by default if you accidentally upgrade) leads to missing dependencies; which get pushed to CodeArtifact.

All I seem to be able to do is solve, not replicate.

lewiscowles1986, 19 days ago to random

A really annoying example of how wasteful you manufacturers can be. Toys where it's a "surprise" overcoming choice... Such a bad idea!

Shameful to see two popular brands engaging in this

Image of a scam children's toy costing £3 with a "playful" question-mark, meaning there is no way to "know" what is inside.
Image of a scam children's toy costing £3 with a "playful" question-mark, meaning there is no way to "know" what is inside. You can find out all the variants of this one, but which variant you get, we've no idea about.

lewiscowles1986, 1 month ago to random

It takes a special kind of business to admit such a thing, but I'm not sure this was intended..

Since noreply@curve.com is their email, I'd love it if folks can share. Let them know they need to do better

lewiscowles1986, 1 month ago to random

I've been far too happy, and grateful since my surgery. So here I sit, in surgical assessment unit 2, because I've been in agony since 5am.

lewiscowles1986, 1 month ago to random

TIL via @Crell tht PHP is getting C-sharp style getter / setter "hooks"

nice.

IDK if the way I think about objects is the issue, but inherritance here feels weird.

If I have a struct and the only way to set values on it is via a header file where some macro modifies my input; then there is a layer of direct setting, that I'm not sure would translate to another struct with same address locations / names, without that header.

So if folks stack the hooks does perf get progressively worse?

lewiscowles1986, 1 month ago to random

Just found my wonderful blue logitech mouse.

My gosh it is so much better than the magic mouse; seconds to go from low-charge to fully charged (battery swap); works with all my OS's; actual scroll-wheel, so it does not accidentally trigger "for convenience"

lewiscowles1986, 1 month ago to random

Plausible analytics looks amazing

Thanks @jalcine for documenting your usage of them, and @plausible for making a fairly solid project, with documentation answering my almost every question.

This could be user failure, but do you have anything available for keeping US data within the US?

lewiscowles1986, 1 month ago to random

Is this real?

Says it is from the US Fish and Wildlife Service

lewiscowles1986, 1 month ago to random

Excalidraw is such a win for my note-taking.

As despite donating to my note-taking app, it being at-least temporarily discontinued; I set about hacking a single-page which I can use to load excalidraw files.

I have got rid of some UI I don't value, such as the burger menu, command pallete and library link; but library import / export, publish works; with dark mode and light mode via user browser preferences A.L.A matchMedia API

It's also #offlinefirst

lewiscowles1986, 1 month ago to StarTrek

First ever operation achieved.

Amazing job by the pre-op, surgical and post-op teams.

While laying, and pacing the corridoors in recovery, it was so easy to see how much worse outcomes others get.

I think I might be in less pain now, than I've been in for the last 2 years.

short of #StarTrek medicine I can't really work out how it could have gone better (although there was post-op discomfort obviously)

lewiscowles1986, 2 months ago to random

Okay, so if you had a test suite failing for a specific version of a scripting runtime, but commands pass, and produce the expected output. Would you still consider the code {runtime_version} incompatible, or would you investigate the test harness?

lewiscowles1986, 2 months ago to random

Thick people, overwhelmingly assert things with no intention of backing them up...

https://zirk.us/@noplasticshower/112184598111157869

OpenSource is not more vulnerable that Closed Source; that is a trick. IF you believe it, and can provide no evidence for your position, it means you are being a moron.

Just because idiots and thugs damage the commons, does not mean that a commons is a bad idea, or that worse things come from there. If you cannot handle that subtle nuance, you are at best intellectually stunted.

lewiscowles1986, 2 months ago to random

jiat0218@gmail.com absolute scummer.

I Dislike people like this, for the same reason I'd like to take a cricket bat to folks writing DRM, EFI.

You're eroding trust in the work we do ya doof.

Hoping this bad actor gets serious repercussions for such stupid attacks.

At the point you're using tests to hook build processes to inject code; it's not a mistake.

lewiscowles1986, 2 months ago to random

@simon
I was really surprised recently to be unable to find time-series data in datasette.

Has anyone contacted you about that, which you can remember?

lewiscowles1986, 2 months ago to random

I know there is the "slippery slope" argument. But this really frustrates me:

https://phpc.social/deck/@msw@mstdn.social/112130308373760101

Source Available, at-least means I can learn from a thing.

Most people do not, and maybe cannot compile the "OpenSource" software they use.

It's poorly documented, CI is hard to reproduce or is choc full of proprietary crap. Less than 10,000 people worldwide can use the thing fully, beyond the getting started docs.

So Source Available, bringing funding helps those businesses.

lewiscowles1986, 2 months ago to random

What is that thing called, where you give a steer, like "lets not do this right now" and someone responds "I'll close this work"

😬 It's not that I want to crush your geekery or stop this work. I'm just saying to focus on the pre-agreed stuff first.

Honestly, once you build a history of working on the prior agreed stuff, we're good, I'll try to find someone, even if it's not me to help.

lewiscowles1986, 2 months ago to random

Surely the problem with Redis behaviour

https://narrativ.es/@janl/112132188906467025

is that by insisting upon perpetual resource extraction for discoveries you had commercial options for, for over a decade, all you do is harm the mechanism of distribution (Open Source), and stagnate your little area of "specialism".

And while you were putting others out of work, taking food off their table, building your free, but also wants to be paid product, you've technically caused irepairable harm to users and specialism.

lewiscowles1986, 2 months ago to random

Hmm, is anyone else using mermaidJS to create diagrams, where state is temporal, but without using sequence diagrams?

Also is anyone using styling to highlight the active states, regardless of which type of diagram you are producing?

lewiscowles1986, 3 months ago to random

Why is everything the EFF does annoying

https://www.eff.org/deeplinks/2024/03/should-caddy-and-traefik-replace-certbot

If resources are a problem, maybe don't keep developing Traefik or Caddy.

Put them both in the bin where they belong, set fire to them, and salt the earth they came from.

It can't be "easier" to develop full server software, compared to a single-purpose utility.

Now ACME is proven it should be charged for anyway

lewiscowles1986, 3 months ago to random

14 minute video with someone howling about how they don't like sub-process calling

https://www.youtube.com/watch?v=3ysBj4cRU5c

me too bud, but I generally keep it short.

Things they omitted to mention, despite 14 minutes

sub-process calling can lead to unnecesarrily coupled or missing system dependencies.

dependent upon user input sub-process calling can lead to injection attacks, shell escapes, and other vectors which increase required efforts on your utility, and increase risk if choosing to use it.

lewiscowles1986, 3 months ago to random

IDK why, but I just reviewed some code, and it has my head spinning. I've never seen this syntax in PHP

https://3v4l.org/6uFuA#v8.1.27

It looks like because PHP arrays support non-numeric indexes (they are basically objects); you can send naked property names (without quotes even) and have it return an array with named keys.

Clever. Next do list comprehensions lol

I've not seen this in other languages either. Python kwargs was what I thought it might look like, but it's not.

lewiscowles1986, 3 months ago to random

I Love how bad actors just implode with nonsense then self-immolate.

https://m.fedica.com/@samir/111931172338498607

Totally absurd Gibberish

Then, pointing out that they seem to lack the capacity to think around this issue got a block...

https://m.fedica.com/@samir/111931212199469031

If you just want to ignore the terms and conditions of instances, and absurdly misunderstand, or misrepresent others viewpoints; then take exception to folks correcting you... You're gonna have a bad time!

lewiscowles1986, 4 months ago to random

I just got nodejs v20 running via nvm on a non-rooted 2020 iPad

ha.. Might need to try asdf with @ramsey php to see if I can run above php 7.4

Why? Because I can and I'm interested

lewiscowles1986, 4 months ago to random

Okay, folks who have completed a 48 or 72 hour fast (just food); how did you manage.

I Feel like I need all the tips. I've done 24 hours a few times, even completed Ramadan with some folks of that persuasion, as a sign of solidarity.

48 seems high and after reading about someone on a getaway to Switzerland who does >72 annually for a "tune up"; who says 3rd day is misery, I'd like to know from folks who have completed. How you did it without a straight jacket and locked room.

lewiscowles1986, 4 months ago to random

Actually heartbreaking story

https://www.newyorker.com/magazine/2024/02/12/a-teens-fatal-plunge-into-the-london-underworld