Notices by Scott Feeney (graue@social.coop)

PSA: @signalapp has been modified sometime in the past few months (I think) to send filenames of attached images.

It's not obvious it's sending filenames because they're only visible if the recipient saves the image on desktop. This can reveal the order in which photos were taken, how many photos were taken in between, if a photo was taken on a different phone, etc. It's a surprising choice for a privacy-focused app, and partially negates benefits of stripping EXIF metadata.

Every line is pitch-perfect. McSweeney's, you glorious bastards, you've done it again. https://www.mcsweeneys.net/articles/a-company-reminder-for-everyone-to-talk-nicely-about-the-giant-plagiarism-machine

BikeHopper now shows you at a glance where the steep hills along your route are!

This required me to learn way too much about the obscure SDF image format, used by MapLibre.

@lmorchard @mekkaokereke This would also reduce the incentive for websites to block Mastodon, so we might get to actually see more preview cards!

Wikipedia is one of the few big websites that still make it possible to argue the internet was a good thing for humanity and not just a tool for amoral oligarchs to surveil us, lure us into surrendering community connections and knowledge into their control, and erect digital tollbooths.

@michaelcoyote You can ask Eugen Rochko why he refuses to merge the already-written code, which has been sitting around for 2 years, 6 months and 28 days, to make blocking someone hide their reply from under your post.

https://github.com/mastodon/mastodon/pull/18468

@evacide

Although the fediverse is in no way scraping-proof, if someone did this so blatantly here they would be widely shunned, defederated, and blocked. The Bluesky spokesperson only offers a milquetoast statement about maybe letting you opt out in the future, suggesting they think this is OK. Point in fedi's favor.

https://www.404media.co/someone-made-a-dataset-of-one-million-bluesky-posts-for-machine-learning-research/

@mcc Twitter was also a “proprietary service operated by a single company, but with an unusually good ‘API’ for plug-in 3rdparty tools” until they dramatically locked it down in May 2012 and banned third-party clients.

I remember that moment well. It led to a wave of migration to a federated, open-source alternative: Tent.

For a moment I really thought Tent was gonna change the world. Fun time capsule now: https://scott.mn/2012/10/20/tent/

One of the worst misfeatures of #Mastodon from a safety angle: If you get an abusive reply to your post, you can't remove it even by blocking the person. Even after blocking, your server still shows the abusive reply to everyone else.

There has been a patch ready to fix this for 2 years, 6 months, and 1 day. @Gargron has not approved it, and has provided *zero* explanation for why he will not allow it to be merged.

https://github.com/mastodon/mastodon/pull/18468

@Andres4NY @mekkaokereke cc @cheeaun: would you be open to adding support into Phanpy for new features like starter packs, limiting replies, and composable moderation, if pioneered by a server app other than Mastodon?

@jenniferplusplus @inthehands Phanpy has done this, removed the reply button from the home feed! You have to either click into the thread or click the ...

idea: a functional programming meetup called “Coffee is for Closures”

This is brilliant, and makes me want to read more Ted Chiang.

"[AI] is a fundamentally dehumanizing technology because it treats us as less than what we are: creators and apprehenders of meaning. It reduces the amount of intention in the world."

Hard to select just one pull quote; whole piece is gold.

https://mas.to/@gleick/113058537194470078

Lazyweb, is there a way, built into HTTP, to resume a partially completed upload?

What I'm finding searching the web seems to imply that there's not, and you would need to implement ad-hoc endpoints to query upload status, resume, and then combine the file chunks on the server. This seems surprising and frustrating to me; the feature is a must for reliable mobile uploads of media files, and this means it's impossible when I don't control the API... is it really so?

@corbet My first thought was that should be no problem for the Internet Archive to back up.

...then I thought, why does a scrappy nonprofit have to do it instead of the $400 billion company keeping them up in the first place?

@hazelweakly this cartoon is not an exaggeration. I just came across a tech company offering exactly this.

@dansup Thanks for clarifying. I definitely do remember you creating a list on FediDb.org of all the servers that had signed the FediPact, and pixelfed.social being one of the biggest servers on that list. The pact says, "i hereby agree to block any instances owned by meta should they pop up on the fediverse." https://fedipact.online/

So this is a change of position, and one I personally don't agree with, but that is your prerogative. @nathanu

@dansup Wait, I thought pixelfed.social was part of the FediPact to block Threads? What changed?

@nathanu I'm not sure I'm gonna go that far, but I am disappointed, and would appreciate some transparency as to what changed @dansup's mind, and if there's a possibility of reconsidering based on Facebook's lax moderation practices toward bigoted content.

@rysiek I wonder why this paints such a rosier picture than the graphs on https://fedidb.org. Not just >double the peak, but a totally different trend. Cc: @dansup