Lightnews — Scholar-powered news

bryan newbold @bnewbold.net · 6h

yeah!

2

bryan newbold @bnewbold.net · 6h

you can always do "Other" and just write in a comment

2 1

bryan newbold @bnewbold.net · 6h

I think anybody being able to report mis-labeled content is good.

if it is the person who was labeled, that is an "appeal". if it is another person (perhaps an advocate for that person), I guess it is debatable if it is a report, appeal, dispute, etc

1 2

bryan newbold @bnewbold.net · 7h

we have this system in place to give labelers control over what types of reports are appropriate for them (linked).

my stance on this is around whether it is good to encourage labelers which have no standard mechanism at all to receive private feedback from impacted accounts/humans

proposals/0009-mod-report-granularity at main · bluesky-social/proposals

Bluesky proposal discussions. Contribute to bluesky-social/proposals development by creating an account on GitHub.

github.com

1 4

bryan newbold @bnewbold.net · 7h

conceptually splitting out appeals as a distinct type of action makes sense.

from a process and architecture perspective, appeals and reports are very similar: they both require a database somewhere to track, auth, review, etc

2 1

bryan newbold @bnewbold.net · 7h

totally agree! services should be able to radically filter down which reports are in-scope and almost never end up on that list if they end up there.

in terms of current UX i'm on-board not showing up there. I'm mostly speaking to the semantics of "what is a labeler conceptually"

5

bryan newbold @bnewbold.net · 8h

hrm, maybe if stream.place is running app code and you can easily add the .well-known

at a larger org like 'did:web:microsoft.com' that could be a whole lot more complex

1

bryan newbold @bnewbold.net · 8h

it's also to enable more complex hierarchies of namespaces. eg:

'uk.ac.school.dept.project.someRecord'
'dev.lexhost.username.project.someAPI'

wanted to avoid having to attempt a bunch of different resolutions, and have different outcomes based on which succeeded/failed

1

bryan newbold @bnewbold.net · 8h

tooling can help with this a bunch: an index website of all the popular lexicons makes discovery easy. then pull/sync tools can enumerate all the "groups" that are hosted by an account

bryan newbold @bnewbold.net · 8h

I think having prevalent linting / CI could help with this. and I think it is segmented to big projects under active development.

a worrisome bit is changing DNS provider/control and needing to copy all the records. but I think zone transfers should automate that?

1

bryan newbold @bnewbold.net · 9h

I also think there is a huge gap between "easily identified" and "completely free from all errors and misunderstandings"

2

bryan newbold @bnewbold.net · 9h

a specific important case of this is appeals.

on the other hand, I definitely think that labelers need to be able to back-pressure reports: they should not need to accept an unlimited number of reports of any kind from any account

1 5

bryan newbold @bnewbold.net · 9h

I get the desire for it, but I think not accepting reports is basically running "open loop" with no human feedback, and isn't a great pattern. it feels pure automation and not augmentation; I think automation almost always still needs humans in the loop

5 9

bryan newbold @bnewbold.net · 9h

4

bryan newbold @bnewbold.net · 11h

yes

1 4

Reposted by bryan newbold

Skywatch Blue / Anti-Alf Aktion @skywatch.blue · 15h

Do you use Skywatch? Have you previously used Skywatch? Have you heard of Skywatch?

If you answered yes to any of these questions, we'd love to hear from you and have put together a short survey. Please fill it out to help us figure out how to better serve our users!

Skywatch Survey

Skywatch thrives on community input and exists to make Bluesky a better place for you. Please complete the short user survey below to help us improve the Skywatch User Experience

forms.gle

37 71

bryan newbold @bnewbold.net · 1d

better link
youtu.be/X2fGzmphx4U?...

Boston bomb scare suspects press conference about Hair.

YouTube video by hyperrreal

youtu.be

1

bryan newbold @bnewbold.net · 1d

reminds me of the mooninite press conference.

I think you can only pull off the shitposting thing once or twice a year at most though

2007 Boston Mooninite panic - Wikipedia

en.wikipedia.org

1 1

bryan newbold @bnewbold.net · 1d

Data Validation - AT Protocol

Expectations around recommended data limits and validation.

atproto.com

2 1

bryan newbold @bnewbold.net · 1d

it does burn disk for mirrors.

the abuse rate was higher in the spring. we tightened things up but it is still ongoing.

we think keeping the system open and simple is important enough to let this slide for now. aka, didn't want to add authentication

1 10

bryan newbold @bnewbold.net · 1d

anybody can point to any domain in that field.

there is no such server

1 4

bryan newbold @bnewbold.net · 1d

we are pretty committed to lexicon neutrality, and PDS hosting is a tiny fraction of infra costs for us today

2 1 13

bryan newbold @bnewbold.net · 1d

storing records is cheap, and we have rate-limits and norms to keep that from getting out of control.

customer support (eg, account recovery) is more expensive.

and blob storage could end up expensive for some apps (eg, long video)

1 9

bryan newbold @bnewbold.net · 1d

it is just spam/trolling, they are not functional accounts.

pretty sure it is nostr people grumpy that PLC is not based on a cryptocurrency.

the indie PLC org can make a call about how/when to purge spam like this

1 1 5

bryan newbold @bnewbold.net · 1d

it is just spam/trolling, they are not functional accounts.

pretty sure it is nostr people grumpy that PLC is not based on a cryptocurrency.

the indie PLC org can make a call about how/when to purge spam like this

10