robryk @robryk@qoto.org

@deirdrebeth @CAnxiolytic

Ah, right, you can jacket steel in something softer. Do you have some rough estimate of how large a fraction of armor piercing bullets contain steel?

@deirdrebeth @CAnxiolytic

But bullets are usually not ferromagnetic: they are usually made from lead (steel is (a) less dense (b) more expensive (c) harder).

@CAnxiolytic

Is there a hypothesis on what exactly happened (did the gun fire by the normal mechanism of the pin striking the cartridge or somehow else (e.g. due to sparking?)?)?

@lauren

I think that part of the situation is that you have more time to be confused and notice the confusion: due to Russia being at war, the coup attempt involves traveling larger distances and thus larger delays.

@wilkie

The way fedi works right now becomes increasingly confusing and messy the more examples of intransitive federation there are, though.

@keir do you know what created the faint vertical lines?

@8petros

> Ale coś mi chodzi po głowie, że silniki się zasila ciekłym LPG, który odparowuje dopiero w silniku

Silniki z gaźnikiem zasila się gazowym LPG. W takich samochodach jest zamontowany wymiennik ciepła służący do ogrzewania LPG za pomocą chłodnicy, żeby go odparować (i dlatego te samochody trzeba rozgrzewać na benzynie i dopiero gdy chłodnica się nagrzeje przełączać na gaz).

Nie wiem jak jest z silnikami z wtryskiem.

Jeśli jest inaczej, to potencjalne rozwiązanie to użyć odparowywacza. BTW. Zauważ, że jeśli chcesz pobierać gazowe LPG z butli, to skądś musi przychodzić ciepło potrzebne do jego odparowania, więc tempo poboru jest ograniczone przez tempo wymiany ciepła między butlą a powietrzem. Podejrzewam, że samochód potrzebuje większego niż osiągalne w ten sposób tempa poboru.

@ilja @rysiek

It seems that microblog.pub does generate ldsigs for public activities when sending them out: https://git.sr.ht/~tsileo/microblog.pub/tree/v2/item/app/outgoing_activities.py#L222 (but apparently only when pushing, not when someone requests activities from it).

@ilja @rysiek

I thought that you could pass around activities signed by their actors (I recall a mechanism called Linked Data Signatures, but now when trying to look it up ended up in a maze of specs, all different: its spec was apparently subsumed by https://w3c.github.io/vc-data-integrity/, which uses some terminology that's confusing for me at first glance). Do I recall correctly that one could do that in principle in APub? Do you know if anyserver does that?

@rysiek

Ah, and for even more confusion:

if A defederates C, and someone from C responds to post from A (that they've gotten via e.g. third-party boosts) then this response:

a) will be visible on C,
b) will not be visible on A,
c) I _think_ will be visible on unrelated instance X if someone from X follows the response's author and the response was not unlisted or mentioned someone from X (I haven't verified this experimentally, but this follows from how I understand ActivityStreams).

@niconiconi Wouldn't the expressions be longer and thus take more RAM per cell?

@rysiek

Note that the standard doesn't say anything about blocks, so each piece of fedi software can do something different.

Mastodon implements suspension by refusing to accept any signature from the defederated domain as valid[*], refusing to actively talk to it, and iiuc dropping users from that domain from some lists (e.g. followers' lists). More importantly, it doesn't tell any other instance about this block (and i think this is actually a shared decision by most fedi software).

The latter property means that B can't even tell that A suspended C, so had no way of knowing that A would not want its messages forwarded to C.

The former property implies that A will ignore boosts of messages from C (because they point at an invalid message).

[*] I didn't verify experimentally, but it seems to, somewhat amusingly, have also applied to Delete messages, so i assume that if A defederates from C, it will stop honoring C's requests to delete messages it has previously received from C.

@Impossible_PhD

How much difference should we expect between placebo and the (nonexistent) arm that got nothing? (The reason I'm asking is that I would expect that most participants could tell whether they were getting the placebo pretty quickly.)

@timorl @lori i also enjoyed playing sail from bsdgames a long time ago

@delroth

I think a better argument than the data protection one is rather about this being akin to price dumping, where you do that to deprive others of users.

@grrrr_shark

> if they'd prefer to be somewhere that chooses differently, I will do everything I can to facilitate them.

I don't think this is a universally held standard, esp. that Mastodon's block semantics make that very hard to do after the fact (because the blocked instances might not see the "user moved" message, so followers won't notice; also follows from blocked instances get dropped on the floor immediately last I've checked). I think at least some people would be way less prescriptive if that was commonly understood as a baseline standard.

@rysiek I grow really weary of the broken "following" state machine on fedi; something sometime caused me not to follow you anymore without any notification (and this is not even the real brokenness, where if things go out of sync on both ends _user intervention_ is required to get them back in sync).