CID profiles, the white whale of 2025, has landed! 🐳 More soon but if you've ever been confused about why the same data yields different CIDs, this should help.
One million blessings to @lidel.org 🙏🏽
specs.ipfs.tech/ipips/ipip-0...
screenshot of the one-pager version of the servo readiness report
How do we get to more than just three web engines owned by three US companies?
It's a gargantuan question, with no easy or right answer.
I've put together a draft report, thinking about it through a very specific approach - please enjoy:
Servo Readiness Report
webtransitions.org/servo-readin...
Something like github.com/inkandswitch... for fast/small data
or storacha.network for blob storage.
In any case, dumb means that the server doesn’t know what it is storing for you.
Isn’t storage always simple? Not easy though.
simple = not complex
easy = without difficulty or effort (eg. crypto)
Hah, I had an inkling, appreciated some comments though.
So far we've solved private blob storage with encryption but I can't tell if that's right direction or not.
Key management is a huge pain, so having private storage with capability-based auth that's private by default might be a good intermediate?
I like it, I've been doing more storing by hash in general.
We're missing the private blob storage part of the equation. Dumb hash-based storage, but private. Storacha, but private data + better UX.
@expede.wtf is doing the other side github.com/inkandswitch... (dumb storage for small & fast data)
The latter for WebDAV for sure, it's a shit show, doesn't work well at all.
Dropbox, bit of both I guess, there's no real alternative with improved, or let alone the same, UX but with better tech.
For me it's gotta be tools and protocols that have been around for ages. Cloud storage (Dropbox, Google Drive, etc) and WebDAV.
I have no user analytics, so I actually have no idea if people use S3. It always works as intended, maybe that's why I don't hear about it 🤷♂️
I also might’ve been making this intentionally harder by providing the user with a choice of where to store their user data.
I have yet to find a way in which this makes sense to non-tech people.
This is what some people have been trying to do for at least 10 years.
I feel like the hard part is the UX?
Also, why is atproto supposedly better in this regard than remotestorage.io which has been around for ages. Even though the former is solely public data. Is it features, or timing?
I avoided that because it adds another layer of complexity, and I despise maintaining servers tbh 😅
Would love to try this out though. Let me know how it works out.
Not sure if such a server would need separate auth, or if restricting to certain domains is enough?
Ohhh interesting! First time hearing about this, definitely trying this out.
Does it clear out things from the cache at a certain point? Like if you don’t play something for a month?
I’m trying! 😝 That does indeed sound cursed.
I agree on Apple Music, it’s so bad 🙈 I’m making something where you can pick/build the music interface, hopefully that brings us back the good parts from back in the day.
Thanks for the input! If you’re down for it, curious to hear what exactly you liked about the old itunes.
PS. Thanks for the input, and all the rest of feedback you’ve provided over the years 🙏
Thanks for the input!
Regarding the clean metadata. Would you be down for a layer that sits on top of the files, ie. the metadata is extracted from the files and adjusted there? Or you prefer the files themselves to be changed?
Also, external audio scrobbler for tracking listens, or included?
Anything in particular that would make you hate it less?
Thanks for the input! Checking out Rekordbox 👀
If you’re down for it, happy to hear what exactly you want different 👀
Thanks Boris! Putting Ampache on the list to integrate with 👍
I wish I could add Bandcamp integration too, but it’s not looking viable; you have to contact them for api client registration.
What makes offline caching very good? As in, easy to cache something and works reliably?
Oh yeah for sure, that’s my whole jam.
The thing with OPFS is that you can’t access it from your OS, so the user has no access to it from outside the browser. The api is similar to the “file system access api” which is the one I referred to earlier.
I did implement the latter, but it’s not great.
That is quite the setup! Thanks for your input 🙏
Yes, on the former. Existing drive, not sure, always has to go through an extra layer I suppose. Like Dropbox's syncing client which then could talk to a local drive?
Native file system access on the web will probably never work well enough to be user friendly imo.
Alternatively, we opt for something like @rocksky.app and then we get the best of both. Diffuse doesn't need to store extra data and we get free data access thanks to atproto.
In any case, thanks for the input! Glad to hear music player feature ideas.
I've been thinking about the latter too. On the one hand, I feel like that's the job of a "scrobbler" like libre.fm or last.fm
But maybe Diffuse should track it too? Not sure, it would be part of the malleable stack then. Much easier than creating an API client to extract meaning full data 🫠
I feel you. But to be fair, Dropbox is one of the easier ones to implement, out of the "requires oauth client registration" ones that is. Unlike, Google Drive, which is a nightmare.
Just for you Jeff I'd add back Dropbox support; and @ruperts.world
Nice! Team gonic/opensubsonic here too 🙌 I have been on the fence for beets for as long as I can remember, I might finally do it…
For the very few people that do still download music and other audio files, where do you keep them?
Are you happy with that system; if not, what would the ideal audio file management system look like to you?
Also, do you wish you could do more with those files?
I was thinking that concept of paying Storacha as this separate service felt weird, but people pay for iCloud and Dropbox too.
Anyone got any thoughts on this? Am I missing something here?
Glad we got ATProto now for simple key-value user-data (although not private), but we need blob storage too.
We need a user friendly option where any software can outsource its (user) storage capabilities and have the user pay for their storage separately.
That's what Storacha does, but it lacks a couple things:
* The user needs a good overview of their data and must be able to delete it
* Private data
