robryk @robryk@qoto.org

@lcamtuf Note that instances that have people that follow you will generally fetch media too. "Remote" media is treated as a cache, and various instances have various approaches to how long they cache it for etc. I would hope that they drop it when they receive a deletion event for the associated post, but (a) it's hope (b) that notice might never arrive if something broke their mutual federation in the meantime (e.g. one blocked the other; iirc in either direction).

@niconiconi Crash it due to not upholding expectations around rise/fall rates (and something in hardware starting to have problems in the direction of metastability) or because of software that deals poorly with that (by e.g. unmasking the interrupt too early or sth. like that)?

@delroth Though that's a counterpoint to those claims made in 2016, not necessarily to them made now.

@varnvlog@toot.community Unions as in worker unions, or do you mean something else? Also am curious which country you're thinking of (I would expect unions to significantly differ between countries, but I don't have enough experience to be in any way certain.)

@Foone I wonder if there's a way to get some client to prominently show new posts with introductions hashtag from followees that have empty notes (or to have a view that shows only such).

@BugbearPancakes Those websites also run a moderation/... model that's inferior to distributed things. In distributed setups, you ~always have at least a passable business relationship (if not a relationship based on trust) with whoever has effective moderation power over you (users with admins, admins with other admins/hosting providers/...). This IMO keeps things more stable over long periods of time by allowing/requiring feedback on such decisions to actually matter.

This is why I'm also concerned about cloudflare being the obvious place to go for ddos protection to.

@b0rk When you have a time-travelling debugger available (like Mozilla's rr), stepping backward is very powerful. If you don't, and the code you're executing is deterministic enough across multiple runs, you can step backward to last execution of line N before point of interest by:
- executing once with a breakpoint that counts on line N,
- noting the count at the point of interest,
- executing again and actually breaking on that breakpoint when count reaches that value.

@matthew_d_green I'm not sure I'd call that humility: conversely, you can be pretty sure that you've proven something and (usually) you can be very sure that you've disproven something[*]. There's some amount of asymmetry, and some amount of reliance on some "ultimate" forms of argument, while downweighting intuition, analogies, and empirical observations.

[*] very informally; I mean "somethings" like statements that start with a universal quantifier and have an easyish to evaluate expression under it

@matthew_d_green I started thinking recently that anyone in any sort of position of authority in infosec should take an academic introduction to cryptography course, as an easy way of at least realizing that the approach of "things are true or false, we might just not know", "modus ponens works", "vacuously satisfied implications are true, not some weird third state", ... exists (and, I'd hope, adopting it at least somewhat).

@newtsoda There's this annoying contributing problem where people shuffle meanings across {sympathy, empathy, ...}, do it differently, try to use the words to talk to each other and fail terribly.

This is a worse version of the problem around the meaning of apology.

@matthew_d_green At the same time you learn that precision matters, and that a proof or counterexample cannot be ignored other than by changing the assumptions we operate under. One also learns that structuring things so that being precise is easy very often makes problems and possible approaches easier to notice. I would dearly wish that all infosec-adjacent people shifted a bit in this direction.

@grrrr_shark

Hm~ I have an impression that various weirdnesses are more socially acceptable here than in Poland (can't point at any specific things off the top of my head though). I wonder if it's me being more oblivious (not speaking the language natively), or Poland being worse on this axis than Germany, or my impressions are affected by change in my age.

@aredridel Did they also have such problems with "purely transactional" emails (asking a company/institution to sell them something/do something/...)?

I'm surprised that older people have this problem: I would expect that they had much more exposure to business mail in their life, which has a very similar concept (at least the DIN norm for letters does treat it as a very obvious part of letter, on a similar level of obviousness as your/mine reference number). Thus I wonder if that's purely a problem in personal emails, where there is not necessarily a single thing they're trying to achieve via the message, or if my assumptions are wrong.

@wizofe @SamTheGeek @matthew_d_green

Ah, sorry, I was trying to point out that there's a way in which this is objectionable independently of one's definition of freedom (or rather, independent of one's view of the decisions taken, but only looking at the process that caused them to be taken).

@wizofe
I think many people find laws that don't prohibit individuals from doing something, but make it somewhat more onerous in a way that has no direct benefits, objectionable, regardless of the law's intent. Similarly, I expect even more people find the concept of government exerting pressure onto anyone in a non-public way objectionable, regardless of the intent of that pressure.

@SamTheGeek @matthew_d_green

@kuba I wonder whether doing something similar with fireworks would be considered better/worse.

I also wonder when we will get some anarchist group trying to spoil such advertisements by adding light sources in their midst.

@griffinkate Wow, one set of the other party's attorneys have resigned because they weren't getting paid _and_ (reading between the lines) were asked to participate in things they must not support: https://storage.courtlistener.com/recap/gov.uscourts.nysd.517471/gov.uscourts.nysd.517471.65.0.pdf