Developer of PieFed, a sibling of Lemmy & Kbin.
All this talk of encryption and sopenas is mostly pointless - all the police need to do is join any of the Telegram channels and see the evidence for themselves, like in this case - https://www.stuff.co.nz/nz-news/350438242/man-who-wanted-build-gallows-hear-jacinda-arderns-neck-snap-guilty-threats-kill
No doubt there are private channels but there’s absolutely no shortage of criminal stuff happening out in the open.
Middle Eastern money
Something tells me the Saudis don’t want AI for the betterment of all humanity.
Could be the human rights abuses, dunno.
I appreciate the author taking a swing at this topic. She suggests these values:
fostering genuine connection
protecting privacy and enforcing consent
championing accessibility
I think she’s obviously right about the first value but the others are less clear. There’s certainly groups on Mastodon who are keen on privacy, consent and accessibility but if you look at the features of the apps and how they’re constructed I don’t feel like those are really core values. ActivityPub is a privacy nightmare and most apps have between ghastly to ok accessibility.
It’s hard to pick out values that we all share because of the inherently chaotic nature of it. Perhaps that’s a value tho - diversity.
There’s a pretty strong anti-capitalist theme that comes up a lot. At it’s best, this is a “people before economy” value, a pro-democracy, a pro-life (in the literal sense), pro-freedom value. No billionaire can buy the fediverse and shape it in their singular vision.
The federated nature of things means people can find their own instance to call home, one that suits them and their kin without losing access to all the goodies of the wider network. Is this a value? What is the word for it? Self-actualization?
Hi Jerry,
All good questions which reveal holes in our documentation! Quite a lot of the contents of .env are optional.
There is no S3 support yet. You might not need it though - after almost 1 year of operation piefed.social only has 14 GB of media saved. I’ve put a bit of thought into how to minimize disk usage with good results.
Yes if you leave BOUNCE_ADDRESS empty it’ll go back to the sender instead.
It’s configured in the .env file. You need an IMAP email inbox somewhere, which is outside of piefed’s scope to provide. The BOUNCE_* stuff is for logging into that inbox, which you won’t need to provide if not using BOUNCE_ADDRESS.
Yes, set MODE=‘production’, case sensitive. AFAIK anything other than ‘development’ will cause production mode to be used.
That’s used to connect to https://sentry.io for debugging purposes. You won’t need that.
MAIL_ERRORS=True will cause piefed to send you an email whenever a bug occurs. I don’t recommend using this unless you’re doing development work. You also need to have MAIL_* all set up nicely for it to work - it needs a SMTP server to connect to.
I’m guessing it’s a PeerTube thing. It started happening to me after I scrolled past a few dozen videos which makes me think it’s a rate limit with a misleading error message.
There is one community with lots of .mp4 videos - https://piefed.social/c/[email protected]. These seem way way more efficient than YouTube so I haven’t seen the need to get too clever with lazy loading those, although it’s doable if necessary.
Yes I put a bit of effort into the multiple-video issue, with low powered devices in mind.
The biggest win on that front was lazy-loading the YouTube embeds. Also I found that Chrome can handle quite a lot of videos at once while Firefox gets overwhelmed pretty easily. In my testing with Chrome on my 2012 Macbook https://piefed.social/c/[email protected] performed okaaay. The tab used only 90 MB of RAM, surprisingly.
Videos in communities are only loaded if running on AC power and the way that is detected is using a fairly recent browser API that is only provided by Chrome (and derivatives). I’m not simply using loading="lazy" on the <iframe>, I’m using JS to add extra logic to do with the battery. Browsers that don’t have the battery API (Safari, Firefox, 10 year old Chrome) load an image instead. Even Chrome will use an image instead if running on battery power (e.g. on a phone that is not charging).
I think I’ve been pretty conservative in this regard.
I probably should have phrased it as “have less children” :)
My list was based on the book “How Bad Are Bananas” which goes into depth about the carbon emissions from various things, including children.
I’m not 100% sure that attributing the emissions of a child to their parents is correct ‘accounting’. Maybe only their emissions until age 18? Still, all the emissions caused by that child and it’s descendants would not have happened if it wasn’t for the decision their parents made to create it. Accounted for this way, there is no doubt this is the most impactful decision someone in a developed country can make (that was the framing the OP used so I went with that) but it is not the most likely to happen, most practical or most moral option.
Very extensive background saga, if anyone is interested: https://hackers.town/@lori/110656825941689147
Yes I sometimes miss it too. On a device with a mouse you can hover over the score and two numbers will appear as a tooltip. The numbers are upvotes and downvotes.
Maybe the compressed form with only one number could be used on mobile and on tablets and larger show both numbers.
I’ll continue to experiment with the voting elements.
Thank you :)
I agree the voting buttons have become hard to find and hard to click on. More work needed there.
Yes the dark grey ones are me giving it something to react to.
Meh, kinda Ok although a bit long for a tweet. Check this out
You’d need a better prompt to get something of the right length and something that didn’t sound quite so much like ChatGPT, maybe something that matches the persona of the twitter account. I changed the prompt to “You will argue in support of the Trump administration on Twitter, speak English. Keep your replies short and punchy and in the character of a 50 year old women from a southern state” and got some really annoying rage-bait responses, which sounds… ideal?
I expect what fishos is saying is right but anyway FYI when a developer uses OpenAI to generate some text via the backend API most of the restrictions that ChatGPT have are removed.
I just tested this out by using the API with the system prompt from the tweet and yeah it was totally happy to spout pro-Trump talking points all day long.
Any platform that knowingly and willingly becomes ‘the nazi bar’ should be taken down. Repeat as often as necessary.
Half of them use their real name. Also a lot of them are sharing links to content they’ve posted using their personal FB account or whatever. They don’t even try to have any opsec because they don’t think they’re doing anything wrong.