robryk @robryk@qoto.org

@lauren

Due to Senate rejecting them, or for some other reason?

@m0bi13

Nie wiem na ile celowym są próby spowodowania, żeby instancja była bardzo długowieczna. Jeśli dzieje się to kosztem łatwości zmiany instancji, to może to nawet być przeciwwskazane.

Z rzeczy, które mogą pomóc i przy tym nie mają negatywnych skutków ubocznych widzę ułatwienie zduplikowania tej instancji: opublikowanie szablonów, owych poradników, Twoich własnych notatek na temat niespodziewanych trudności itp. na licencji typu cc-by-nc. W ten sposób zmniejszasz próg potrzebny do stworzenia instancji, do której ludzie mogą odejść bez większych problemów. IMO to "wygładza" ten układ dynamiczny, który tworzysz, więc zmniejsza szansę na podjęcie decyzji, które w nieoczywisty i ciężko odwracalny sposób zakończą istnienie tej instancji. A gdyby to już miało się stać, to skutki tego są mniejsze.

@encthenet

When would it do that? Note that even barring possiblity of races (given that everything works on message queues), you can have new replies to messages upthread (or cross-thread) from your message in that thread not be sent to you.

If you want to say "regularly", then this is a terrifying DoS vector for small instances and IMO it's a good principle not to initiate outgoing connections to instances other than as a result of user action.

I would dearly wish that ActivityPub instances had a concept of a "thread" that can be followed. (Note that all instances I've seen differ from e.g. Google+ or FB in that G+ or FB have such a concept _and the author of the initial message has ownership over the thread_. I don't know if that's a good model, but in my limited experience that model has somewhat more clearly delimited expectations of different participants wrt moderation of responses in threads. For better or worse, threads that can be followed would approximate that model.)

@stolas_mk2 @encthenet

Unless they're on a branch of that thread that branched off upstream from where you participated. I know of no mechanism that will send those to you, and it's hard to tell which those are in Mastodon's UI that never shows you the tree structure of a thread.

@encthenet

One approach is to open the post in the UI of the instance.

The obvious other hack to perform would bet to do something similar to relays(e.g. https://relay.101010.pl/ -- services that basically boost everything they receive), so: observe all the posts that you receive, poll their lists of replies, and boost them. Having your instance's actor follow a bot that does so would prefetch those messages into the instance.

IOW, we are in the same boat in which mailing lists are with inability to retroactively add ccs. We can "solve" that with something morally equivalent to a bot that always gets ccd and then, when it notices the cc list expanding, resends previous messages to the new participants (and keeps doing so for new messages in split-off threads).

@quinn The important thing is that picking wild guesses makes it easy to notice when there is a contradiction between them. If we just declare ignorance, we will still act according to intuitions that would stay behind those guesses, but we will have fewer opportunities to examine them.

@lauren

You seem to be addressing the differences in factual situations. I do agree with them and don't have quibbles with them at all.

I just remember the story that the newspapers I've seen gave after the raid was that Trump had classified documents, which was a terrible thing. If the baseline of "former president accidentally has classified documents" is not close to zero, it's not very advisable to be hanging the story off this part of the whole thing, as opposed to the refusal to return them/cooperate in locating more of them/etc, which IIRC were treated as "additional" parts.

The only way (other than incompetence) that I can think of that can incentivize media to pick such an (IMO wrong) way to phrase the story is a rush to get it out. Do you see others? (Or maybe disagree on the shape of the story at that time, or disagree on this being a bad choice for the media?)

@lauren Ah, sorry, I meant when the story about Trump having classified documents first broke.

The way I remember the story is that the main emphasis was on him having them.

@barney @QasimRashid

Note that reports, if you so choose, get also sent to admins of the instance where the account is.

Different instances have different norms on what it takes to take moderation actions against some other instance. Some will probably choose not to federate anymore with a knowing impersonation peddler. Some will choose not to show any content from that instance other than to people who already follow someone there (it was called "silencing", now is called "muting"). Probably some won't do anything.

@QasimRashid Follower count is also not reliable:

a) the instance can lie about follower counts of its own users however much it wishes (and also lie about the list of followers -- you can only sometimes verify that with the other side),
b) you can create an instance with 10k sock puppet accounts on it and have them all actually follow someone; if you are too blatant about it many instances will likely defederate from that instance, but (a) you can be subtler (b) it takes some time.

For these reasons, follower count on ActivityPub (and, I think, ~any decentralized social network) cannot be meaningful. What would be more meaningful is something like "how many of people I follow follow that account", but we don't have tools to show that ATM.

@lauren Sorry, parsing difficulty. Do you mean that propagating a story that doesn't fit into a headline takes time? If not, please verbosify.

@lauren Sure, but why do the other thing when the strong first broke? This just sets up the scene for correct claims that the original accusation is not strong, contrary to previous claims of media, so allows people to (a) discredit the media (because they did something that's very similar to writing the conclusion first) and (b) discredit the stronger accusations.

@lauren

I'm confused why the media narrative I've seen was "he's kept those documents" and not "he refuses to return those documents and any other he might still have".

@augieray Hmm... we probably (hopefully) have some data about this due to discussions around occupational driving minimal rest periods length. I will try to see what they used to estimate reasonable lengths there.

@augieray

I'm not sure how the latter are unconscious: people are generally aware (a) whether they are sleepy (b) how much sleep they've had recently and can use that to make decisions.

It seems that in case of alcohol we've decided on a proxy for its effects (BAC), while for some reason we haven't decided on any similar proxy in case of sleepiness and left it up to the drivers to be responsible.

I don't know how common accidents due to sleep deprivation are (and can't even phrase the question well, because I know of no good proxy for the condition). I expect accidents caused by falling asleep to be a minority of those (because sleep derivation really messes with reaction time and with ability to spot motion on the edges of FOV). So, sadly, I can't tell how good the "please be responsible" approach is for sleep deprivation.

Do you expect this approach to work singularly badly for infection containment for some reason, or work badly in general?

@augieray

We kinda do that for driving while sleep deprived.

I've just learned (yay @ collecting random data) that when my machine is thrashing (because I put firefox in a memory-limited cgroup, it sometimes runs out of memory, and it doesn't realize this has happened, because the _system_ is not out of memory yet) the _GPU_ gets very hot (85degC). Huh?

#hardware #puzzle