Trusting humans (to any extent) is quite vastly different than trusting corporations, to be honest. But if that’s your jingle, you can always do the work with a bulk tagger / gallery assembler. No AI, nor environment killing, needed.

Organic died shortly ago (they’d been enshittifying for a while). We go with CoMaps nowadays.

The Proton CEO praised Trump and his control and choices regarding IT and antitrust some months ago. People said it was a “one-off” but then the Proton board backed up the CEO. Writing’s been on the wall for a while.

But not Youtube. Which is Google.

Curious.

I’m done with Google! Watch me rant on Google’s Youtube! Earns me money!

So… yeah.

Why are we giving neonazis attention, again?

Fair point. Still, I have to question how many normal citizens can afford to do that. I’ve literally never heard a news heading like “Portugal citizen takes IBM to trial, execs in jail for 25 years”.

• Redundant (it’s not the only security mechanism)
• False / misleading / not-a-Linux issue (more of an internet culture issue).
• FUD. Wayland is not the only way to go; X11 is pretty mature.
• FUD with intent to sell Microsoft’s “Secure Boot”.

Nice try but doesn’t deserve wasting more time on. -4.4/10.

Have you tried hiring a couple humans? They’re much better than AI. And you help your local (physical or virtual) economy!

There are various ways to solve the first part and they’re pretty generic, the only “C++-specific” part is whether you’ll want to build the string step by step (adding characters) or build in one go then modify (build a prearranged string, for example of a specific size, then modify as needed).

To solve the second part we (you) also need to know the encryption algorithm. “Encrypt” is quite a generic word.

(or without internet access)

Having copyright means nothing if the people can’t enforce it. How mch does it cost for a normal citizen to take one (1) corporation to a jury trial, again?

I’m all fine with donating less to Gnome! I’ll start by donating $0.

I mean, I am not too concerned and it is your code. The if constexpr is just so much spiffier. You should do what’s best for you.

My own use case is that for pre-C++23 (and in particular, pre-C++20) workflows I’m looking to homogeneize the set of dependencies used to supplement the standard library and this looks like a great candidate lib. I already have my own expected, actually, but would rather someone else’s who knows what they’re doing. Since C++14 is the oldest standard I have to support directly for clients (big improvement: was C++03 before 2021) and it brought in the big changes that made constexpr actually usable, I treat C++14 as a a sorta “C++ LTS” in my codebase.

That if constexpr in the case you mention does make the constructor clean as heck. Thanks for the clarification and the commits, by the way!

, and we burned it at the stake.

Gotta love that about FOSS communities. Every time someone tries to do something interesting, we burn them at the stake. It’s a classic of the Mozilla fandom, for example.

Running WXP on the metal??? I thought I was mad!

Without a much deep look at the code: any particular reason to not target also C++14?

I feel that would be incredibly wasteful (and a browsing session can be several windows, too) for marginal zero or even negative net gain. Browsing would also need to set isolation profiles, because for some tabs, sites or windows you’d certainly want to have access to your localtime (plus it be precise enough). Ditto for each and every potential variable.

The truth is, not everything needs to be containerized.

Yes, funnily enough.

With the KDE / Kubuntu news, I might have to give it another try soon-ish, by the time of the next LTS most assuredly.

Well it’s easy to recommend Signal for this case. War plans and all!