robryk @robryk@qoto.org

@kuba

Na które sposoby to jest inne od pryszniców na basenie? (Strzelam, że owe nie powodują wrażenia, że trzeba mopa, i ciekaw jestem które różnice są istotne.)

@freemo

Sure, but if there's an open vent in the tank, and there is noticeable sound coming from the vent, then I'd assume that it's at atmospheric pressure and thus can be punctured without much risk (assuming nonsparking tooling and/or long duration venting upside down so that it's unlikely to flash). I wonder whether the warning is trying to warn me of some remainign risk that I don't see or whether it's (overly) simplified.

@freemo

The thing that caused me to wonder now was a shaving cream can, where I could test the pressure inside by deforming it with my fingers first.

I also saw such a warning on camping gas bottles (some mixture of propane and butane), which did have a way of ensuring they're empty provided: there's a plastic shim you can push into the valve to permanently open it (so that you can vent the bottle after you used up the gas). In such cases I don't really see a reason not to amend it to say "don't puncture unless the shim has been inserted for a few minutes".

@shauna

I thought about this once and my conclusion was that large part of unacceptability of murder (compared to e.g. causing on average one death by increasing pollution) is due to the potential use of death threats (implied or explicit) as an extortion mechanism.

Why do pressurized cans warn against puncturing them "even if empty"?

@psn So it's ~uniformly larger among roads that don't appear to be local, or is it that some roads that you'd expect to be in the local road category are not in it? (IOW would you expect these roads to just be slightly smaller or to be those narrow roads?)

@psn Do you mean that the distribution of size of roads in the middle of nowhere is shifted towards larger, or that it's bimodal, or something else?

@th I wonder whether this involves complications around tooth shape compatibility (so that they don't need to slip against each other and thus don't wear themselves out by friction).

@niconiconi in 1d you still have the problem of whether the clopen status matches

@riley @alda

How do you square this with the benefit of getting in the way of a mistake? After all, checklists, physical markings to annotate the correct surgery site, reidentification of the patient etc. also take time and get in the way, but are considered beneficial.

@badlogic If copying a Skeleton destroys the original, it seems to me that it's not intended to ever be copied, so it would make sense to remove its copy conductor.

That said, this remains a footgun for mutating accesses.

@badlogic @Ronflaix

That would create a complementary footgun, where you are preserving a reference to e.g. an entry in a vector (acquired via e.g. `front()`) across operations that invalidate it and it's not obviously recognizable due to auto.

Tak to wygląda

MPK w Krakowie używa openstreetmap do pokazywania pozycji pojazdu w środku. Wszystko byłoby dobrze, tylko używają wersji sprzed budowy trasy łagiewnickiej, więc tramwaje radośnie jeżdżą poza torami.

@b0rk

There's one pitfall (that I totally fell into when setting up my own Nix system) that I think people will end up in if they follow these steps: what does `<nixpkgs>` refer to? What version of nixpkgs and how is it updated?

It's important, because that's what nix-shell and the imo terrible nix-env use. (Apart from standard concerns about old state, if you use either of the two for programs that need to link libGL, you'd better it be the sameish version as the one your system was built with.)

What will happen is that nixpkgs will refer to a nix-profile-managed version (and different users might have their own, including root having their own).

I don't know what's a good solution here that's not a hack of some sort and doesn't require people to keep track of this other pointer to a version of nixpkgs. My hacky way of dealing with that is to ensure that `<nixpkgs>` points at a dump of nixpkgs used to build the system. Sadly that gets hacky as soon as you have overlays or other customizations of nixpkgs in your system configuration.

A similar issue is the meaning of nixpkgs in flake repository (used by `nix shell` and the like), but that will probably only become important as people adopt flake-based tools more.

@rq @koakuma @chozu @emmy

Re AGPL: it's also perfectly kosher to put a proxy in front of an AGPL piece of software that strips the offer: after all AGPL's terms only affect modification and distribution, but not execution. (h/t @delroth for pointing that out to me once)

@dunkelstern kein Problem, nimm dein Zeit: ich werde zurück zuhause nur am Ende der Woche sein und bis dann wird nichts unternehmen können.

@niconiconi is it a barrier where "enter" and "wait for all to enter" are separate calls?

@dunkelstern hast du vielleicht Fotos von diese Vorrichtungen für Platinen? Ich habe nichts besseres als eine Klemme an einem dritten Hand, so bin ich interessiert für andere Möglichkeiten.