Show newer

WTF ... Mozilla had always running JavaScript inside PDFs disabled by default.

But now with FF 88 this option is ENABLED by default. Which means, if a PDF file contains JS it will run without any user interaction. What can possibly go wrong?

To disable this:

about:config
pdfjs.enableScripting --> false

# FF 78.10 ESR doesn't include this option and still blocks JS in PDFs by default. Just tested.

@itnewsbot Finally! Calibri is an abomination; the new Bierstadt font has my vote. 👏

Hopefully they also replace that god-awful Segoe UI while they're at it. Kerning and x-height have suffered for way too long under the tyranny of insane design choices by .

@clovis Firefox Reader Mode usually gets past this but if not, you can use outline.com which always works.

@turtle21 I actually posted it on Twitter expecting nothing, but then a friend of mine replied laughing saying that he went through a "vanishing GRUB" moment a few weeks ago and could relate.

That was one in a million, but definitely worth sharing. 👍

@jumboshrimp Why not both? I have my regular personal aliases and domains that everyone knows about.

And then I have anonymous alts (or maybe not! 🙃) that no one will ever find out about.

It's trivial these days to be anonymous online.

@werwolf Definitely, it's not a major problem per se just a massive brick wall of a hassle when people run into it (mostly the dual-boot folks who for some reason are on MBR and not GPT).

Everybody gangsta until GRUB gets overwritten.

Using is simple, secure and easy. Just create an xmpp account, install an xmpp client, login and start chatting.

@stampirl "Best OS in the world"...

Linux: *looks quizzically at this Darwinian BSD sub-species* 😄

Isn't there a way to set it with the command line in macOS though?

Even on Windows you can run "w32tm /sync" and be done with it (I run that in a startup batch script that hits time.cloudflare.com).

@Ravenhelm Just make sure it's v0.9 so that you can release the 1.0 after you've DMT-tripped into the machine and back. 🤟

@jrballesteros05 Android 10 and 11 support custom DNS over TLS, but indeed it's hit or miss.

I'd recommend using NextDNS and installing their Android app which creates a local VPN and force-tunnels requests through their servers (with your blocking configuration enabled).

I use a personal WireGuard VPN with NextDNS as the tunnel's DNS, but the result is the same if you just use the NextDNS app by itself.

nextdns.io

@tchambers There's also an excellent plugin available for those who don't want to mess around with code, or use a child theme, etc.: wordpress.org/plugins/disable-

But I agree, if it can be mainlined and back-ported into the WP codebase that would be great, although I'm not entirely sure it'll play out that way.

Matt Mullenweg has said they're not considering any definitive action just yet, and there's a chance they might view setting Headers as something that should be left to users.

@rickwayne That's a very valid point indeed, and is probably applicable to many groups.

I don't know how enforceable the policy is, that's the puzzling thing. The only way to account for such bias is to use a weighted algorithm that tries to compensate for the mean (whatever that means; pun intended).

Adjusting the algorithm has to be done by humans, which will provide the desired (but technically false) data; the implications of which are unknown.

I agree with you, only time will tell...

@rickwayne At the risk of sounding pedantic, AI is not real. What many consider AI is really Machine Learning.

When the AI hype train stops it's quickly apparent that if an ML model is not trained on an expansive enough data set, it's useless (the exceptions are game engines which operate within finite conditions).

Only an air-gapped "AI" algorithm is without bias. But such an algorithm cannot become practical without external data, which must be correlated, leading to bias. Chicken, meet egg.

@rickwayne "[...] the FTC urges businesses to test their algorithms for bias, to base their systems on datasets that aren't missing key demographics [...]"

This is a very deep double-edged problem.

The people in charge will attempt to define bias, which can only result in bias (garbage in, garbage out).

Once a data set is anonymized, there is no bias. Great.

The bias comes in the minute the data is shared; whereby a combination of data points correlate with another known data set. Not great.

@binyamin I still haven't gotten used to Josh peeking out all of a sudden halfway down the article! :oh_no_blob:

His personal website is a poster child for his amazing skillset though, couldn't agree more.

@krinkle Thanks for the link! It's always fascinating to get insights into the infrastructure setup behind these large-scale projects.

Tangentially, archivebox.io was making the rounds on HN today.

Show older
Fosstodon

Fosstodon is an English speaking Mastodon instance that is open to anyone who is interested in technology; particularly free & open source software.