KartikPrabhu, geoffo, [tantek], [KevinMarks], jeremycherfas, nickodd, [chrisaldrich], jeremych_, schmudde, [Emma_Humphries], deathrow1, stevestr- and [chrisaldrich]1 joined the channel; nickodd left the channel
#petermolnarskippy_: I host my DNS there, but that's not too interesting.
#petermolnarbut it has a decent API to alter DNS records, making it a good option for those self-hosting from a dynamic IP
[Ana_Rodrigues], willnorris, voxpelli and genehack joined the channel
#petermolnarI made sure the webfinger redirect is working and in place; I even checked what else mastodon might be fetching from me, but there was nothing.
#petermolnardo not ever apologise for providing a magical free service, [snarfed], please
[tw2113_Slack_], [Alan_Smith], [Emma_Humphries] and [Raphael_Luckom] joined the channel
#[Raphael_Luckom]Does anyone have a rationale for when to make a new bucket (aws, gcp etc) vs when to use a new prefix in a bucket you already have?
#skippy_if you think you'll do bucket-level operations on all object sin said bucket. Like, "move all of /2009 to cold storage".
#[Raphael_Luckom]I had that thought too...but are there really bucket-level operations that can't also be prefix-operations _within_ a bucket? Like `aws s3 cp <s3://examplebucket/foo> --recursive` .
#[Raphael_Luckom]Access-control seems to likewise work with bucket paths as well as whole buckets
#skippy_sure. i think it depends on how you do your tasks. buckets are organizational, not really functional.
[schmarty] joined the channel
#[Raphael_Luckom]I store lambda code in a bucket--I use the same bucket for all the lambdas I use. That's been working fine so far...but I have different log buckets for my test site vs my "prod" site. The idea of mixing test and prod infrastructure by using a single log bucket for both is kinda scary, but given that I can apply different access control within the bucket, I think that fear may be mostly superstition.
#[Raphael_Luckom]so far the only fact I've found that seems relevant is that there's a limit on # of buckets (100, extendable to 1K) within an account. But I'm not near that limit.
[Rose] and [jgmac1106] joined the channel
#aaronpkso gargron did a great job of writing up what is required for owncast to "join" the fediverse, essentially pushing "live now" updates out via activitypub
#Loqi[Gargron] What you need to become part of the fediverse:
- `/.well-known/webfinger` that responds to queries and points to
- `/actor`, a JSON response with properties such as `preferredUsername`, `inbox` and some others
- `/inbox` which accepts POST reque...
#aaronpkwhat would be a similar writeup to get that feature into an indieweb reader?
#aaronpki suppose the whole page would be an h-entry with the published date being the date the stream started, then that would be able to be subscribed to in a reader?
#aaronpkthen the addition of websub could make that realtime
#@axuan25↩️ I'm just being picky on the lib name :P If we're using terminology such as "User Info" or "Login" in the context of OAuth2, then it's mostly likely is OIDC (or IndieAuth). (twitter.com/_/status/1341454057383182337)
#LoqiIt looks like we don't have a page for "live now" yet. Would you like to create it? (Or just say "live now is ____", a sentence describing the term)
[Rose], schmudde, leg and [snarfed] joined the channel
#[Raphael_Luckom]I couldn't quite think through my question about how to use buckets, so I wrote a post about state storage generally to see if it helped organize my thoughts: https://raphaelluckom.com/posts/thinking_about_state.html
#[snarfed]any IG users missing Bridgy backfeed and want to try it out? alpha quality, obviously
[jeremycherfas], [KevinMarks] and [chrisaldrich] joined the channel; skalnik left the channel
#gRegorLove[snarfed], extension is showing enabled on brid.gy and instagram.com. I don't think I have any new comments that weren't backfed already, if that matters
#gRegorLoveonly blocked requests I'm seeing are intneral, graph.instagram.com