Mauve 👁💜 @mauve@mastodon.mauve.moe

@GeeksLoveDetail Yeah. Something similar happened before and disaing the battery optimization helped. Now even when I disable all the battery optimizarions it's still wonky. 🤷 My current hypothesis is it's blocking background execution more aggressively in general

😭 I made the mistake of updating Android and now my ebook TTS thing keeps randomly pausing if my phone is locked.

@quinn@octodon.social waaaant

I really appreciate that the goblins in Shadow over Loathing use they them. 🥰

@heapwolf Interesting. I'd love to dig into the source once it's out. 👀

@heapwolf Do you have libsodium working in socket? Been talking to folks about getting hyperswarm stuff working but native libs seem to be a major blocker.

@lumen Oh jeeze yeah, I can imagine the costs are even more unreasonable given the average income in Brazil 😱

Sharing the burden with other seems to be in the spirit of running federated infrastructure too.

@heapwolf *furiously destroys my body and mind to enhance shareholder value* aye aye!

@technobaboo Or maybe "Mime" isn't the right word here?

It's more like a "File" where it has a mime type describing what format it is, and a way to load it's data which might be from a URL or from raw data. E.g. the File API in web browsers: https://developer.mozilla.org/en-US/docs/Web/API/File

@technobaboo Hmm, I'm not sure. Personally, I'd have the mime type contain either the raw buffer for the data, a file URL for the data, or a IPFS/HTTPS URL for the data.

That way you can seamlessly attach either remote, local, or syntesized data to anything that takes a mime type object. 🤷 Not sure what sort of UI you were thinking of for this though.

@technobaboo For IPFS URLs, take the stuff in the PATH_GATEWAY spec, but replace `/ipfs/` with `ipfs://`.

https://github.com/ipfs/specs/blob/e06249febd4b819149039ad8c8719566a4bc113e/http-gateways/PATH_GATEWAY.md

I'm also working on a IPLD URL spec that's along the same lines.

But generally its `ipfs://CID_HERE/path/to/file.txt` and similarly `ipns://PUBLICKEY_OR_DNS_HERE/path/to/file.txt`.

Anyone know of a tool that takes those two column PDFs that papers typically get formatted in and turn them into a one column HTML page?

Or even better, have a way to get the original LaTeX for papers from their ID. :P

@technobaboo @gwil URLs generally don't contain mime types since browsers get them from the Content-Type header in the response (yeah, it kinda sucks). A lot of servers sniff the content and the file ending to guess the mime type. Either way you typically need to read from the resource to know what sort it is. 😅

@todrobbins I tried it in the past, but I didn't have time to actually practice since I always have so much on my plate. :/ Hopefully I'll have more time over my holiday break and then see how it goes from there.

I think it worked okay when I tried it last though. 🤷

I think I just wish there was a path to getting it on mobile, too.

Maybe my new years resolution will be to finally learn how to use Talon for talking to my computer. 🤔

https://talonvoice.com/docs/

@heapwolf Wake me up when the AI sequences a genome to print to grow a living machine that'll do whatever task is needed. 🤪

Preferably if it's a cool crab thing I can stick right into my spine.