The Unclear Impact

Kristóf Marussy ferdi | @kristof@pleroma.marussy.com

I'm a PhD student working on the extra-functional requirements and formal verification of cyber-physical system architectures.
I also like free (as in liberty) software, privacy enhancing technologies, and cryptography.

blobfoxdrakedislike uni IdP system which silently throws you back to the login screen, missing important push notifications blobfoxdrakelike https://linx.marussy.com/cihivs0f.js

It seems that if you want to draw something like -webkit-mask-image on a <canvas>, you have to implement alpha blending by hand: https://git.marussy.com/ferdi-recipes/tree/recipes/pleroma/webview.js?h=pleroma-instance-logo

@trysdyn It sounds like a little car going veeEEEeeeEEEeeeEEEeeeeeee blobfoxhyper

@dantheclamman @dan Interesting question - what is "more" : Sci-Hub? OpenAccess publication in Nature Communications or Elsevier journals (6000 / 2000 USD - inaccessible to many)? Community-own publishing in traditional journals that charge to keep alive? OA in Journal of Open Research Software (100 EUR per publication to pay running costs)? Agree to your conclusion - "something deeply broken" - when I see which journal most people would/have to choose as dissemination channel.

re: alcohol poll

@bryn Still trying to figure out the difference between a shandy and a radler. Probably because I only ever had the latter, and found it delicious.

@jbauer mpv + youtube-dl works beautifully in conjunction with RSS feeds to follow youtube channels

I should teach my RSS reader to invoke mpv, though

@hazel pgtk-nativecomp?

I should learn more about Electron app security.

@TinfoilSubmarine That’s my fear, too. The question is whether e-mail is still an open platform (and you just need set up the right stuff in the DNS entries to show that you’re legit) , or is federation effectively closed now. Having heard some horror stories about sending e-mail to e.g., outlook.com mailboxes, it seems it’s more like the latter. blobcatsad

@TinfoilSubmarine On the one hand, mailbox.org is pretty nice, receives mail reliably with a wildcard alias on my domain, and lets me deliver mail everywhere after a little bit of SPF/DKIM/DMARC faff. On the other hard, they have effectively raised prices, and will likely be subject to ChatControl despite their best efforts. Plus I’d have an excuse to play around with SMTP servers. western_dragon

Do I want to run my own SMTP server?

@jk but a hat much smaller than you could be affixed to your hair by some sort of clip and look funny

Hacker News being riled up on OpenTelemetry in Github raises an interesting question: has there been research on using client-side performance data for fingerprinting users? My huch would say this is possible (the timing of different computations will be affected by the hardware and by other software you run as they interfere with the browser process), but it may be way too noisy to be anything useful (read: malicious).

dom.enablePerformance = false was put in place for use by e.g., the Tor project, so they seem to consider it a fingerprinting vector. Since a couple of days, the Github web interface claims my browser is unsupported and some of its javascript breaks if I use this setting.

Freenode

It's almost as if ownership matters 🤔

Community co-operative ownership of the means of communication is the way forward or these enclosures of the software commons will keep happening over and over and over.

Individuals cannot be trusted in the long run, capital can not trusted, the desire for constant growth cannot be trusted, even NGOs cannot be trusted if they don't have democratic control by their community.

How many times do we have to see examples of this in the wild before some tech folks will acknowledge this problem.

Libre software is part of a global commons. The organisations administering it should be too.

*mic drop*

@ionchy There has be some work [1,2] on formalising build systems using category theory as a bunch of multiary lenses [3], which might be related. But it’s quite weird (and it comes from model-driven engineering people, so it’s the sort of compiler that updates the source code when you modify the compiled target).

  1. https://homepages.inf.ed.ac.uk/perdita/MegamodelBuild/modelspaper.pdf
  2. https://link.springer.com/content/pdf/10.1007/s10270-019-00736-x.pdf
  3. http://ceur-ws.org/Vol-2355/paper3.pdf

andrew lee just seized over 700 channels on freenode because they mentioned libera.chat in their topic.

This includes projects like openbsd, wikimedia, FOSDEM, etc.

https://archive.is/uHw1g shows 720 channels that match what is being checked.

here's an example log: https://gist.github.com/pushcx/ab2a1d5b1d18e964c581ef18ccb3a79f

boost this if you care about foss in any way.

@meisam In particular, MacOS seems to gaslight users about unsigned builds and reports the app to be damaged instead of giving a proper certificate error. The latter would at least give a chance to testers to know where to opt in to running the unsigned nightly.

@meisam Luckily, I’m signing nothing, but an open-source project I’m contributing to is signing/notarizing their nightly MacOS builds (to avoid having to deal with Gatekeeper, I guess), and the setup – predictably – broke.

@aral It’s a bit ironic that a site about moving websites turns out be a hassle to move. blobcat3c

»