Wow, the v2 looks really good. You can write the same app for both desktop and mobile and just be responsive, that’s huge compared to so many frameworks. Also support for multiple webviews per window, I’m curious to see if it’s better than Electron (hard to be worse).
mark5891
Hi all, quick question. Does any of you know about cross-platform IO frameworks? (think of KIO and GVFS but alternatives to those) I'm making a comparing the different frmeworks, i can find only.
mark5891
Second question. Does anyone here know someone from dropbox that they could connect with me?
yep, there’s a “instant gratification w/o commitment” win still sitting on the table for someone who gets website stuffs working well w/ distributed compute + IPFS (or something like it).
I have a basic thingy that can republish to IPFS from a directory when something changes. It's too basic yet but I think it could make stuff a lot less painful. Could also hit pinning APIs, republish to IPNS, that kind of stuff. Should have some bandwidth to get back to it soon.
Second question. Does anyone here know someone from dropbox that they could connect with me?
😮 amazing! Well a connection from any cloud storage vendor dev would be much welcomed. I kinda need to know their opinion and requirements for native host of integration. Same for IPFS but it helps that I already mostly know those ^_^
For anyone else reading this is with the intent of writing a "internet storage filesystem library" aiming to take away the complexities of platform integration, the library would handle that (fuse on Linux, mac and windows would use their specific platform APIs). Other projects (like ipfs, Sia, dropbox, OneDrive, webdav, etc....) would just have to make a plugin for that library to gain cross platform native of integration support. I haven't worked out specifics yet (like choosing between read-only or read-write or how to handle filesystems where changing a file means a whole chain of up to the top level "mount point" would change, still have to figure that out)
Second question. Does anyone here know someone from dropbox that they could connect with me?
Reg. WNFS vs IPFS in this regard, not a concern. Ipfs (unixfs) would be a plugin, ipfs (wnfs) would be another plugin. Though a hypothetical wnfs plugin probably could handle unixfs too.
anyone know how secure those jumio type systems are seems like a goldmine for identity theives if you can phish a login basically take over the bank account
they sent me a jumio link via email but it was like 3-4 subdomains deep then when i verified the acc with my id and 'selfie' i got a reply back telling me that my account was being closed and i had to send them a bank statement (in my name) to transfer it out
no clue if my router has been poped again or what i didnt update it i know there are reverse proxy attacks that can happen in the browser
Just some pics i took in the past couple days here in the Azores. In fact, my view right now while typing this is the Atlantic ocean. Not bad for no job huh 😉
why does channel have all of the worse spam in this slack? WHO CAN WE BLAME?
maybe Rod is having a case of the Mondays
not related to spam, we should change the channel name. Matrix/Element UI doesn’t highlight the server name, so it just looks like an “anything browsers / platforms / standards”, not specifically decentralization or IPFS etc.
related to spam: i’m not sure there are full-time admins anymore for the Slack+Matrix+Discord combo… so the problem likely gets worse before it gets better. though the Matrix spam problem has always been bad 🤷🏽
why does channel have all of the worse spam in this slack? WHO CAN WE BLAME?
it’s always Monday somewhere
Dietrich Ayala (UTC plus 2)
So i had an issue last week where Glif wasn’t showing Ledger as a choice but was showing Metamask, in Brave nightly or Chromium canary. i tested release channels of both, and it did… the opposite 😵💫 i talked to Brave and to Glif, and it turns out it the problem is Chromium 122 enables iterator helpers by default which broke Endo (secure js sandboxes) which broke Metamask… https://issues.chromium.org/issues/41495590#comment11
building for the web is like coding in the dark inside a closet with thousands of ghosts in it
related to spam: i’m not sure there are full-time admins anymore for the Slack+Matrix+Discord combo… so the problem likely gets worse before it gets better. though the Matrix spam problem has always been bad 🤷🏽
hm, do images sent from Matrix show up on Slack and Discord...
dietrich1
yes
but images don't make it from {not Matrix} to Matrix
* but images don't make it from {not Matrix} to Matrix (to other not Matrix)
ribasushi
invisible man
mark5891
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Just going to take like a year or so of full time work.... not even kidding.
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
robin If one of you knows about any more potential candidate libraries/projects that make it possible to have cloud storage integration on your desktop (specifically desktop, as in Windows, Mac and Linux; not looking at mobile). Don't be shy, throw them my way 🙂
While at a glance this might look unrelated to IPFS... It's not 😉 Read the first part above linked document and you'll know my intent. "small projects" huh, lol.
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
It's an example that should also be possible, not one I plan on implementing.
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Remember guys, cloud/internet storage is the aim here. Not networks.
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
That being said, I'm doubtful about fuse. It really depends on the use case and what I'll go for in the end.
If I want to go for native integration AND having the cloud like features (say online, offline and pinned like Mac and windows have for these things) than fuse isn't that interesting as it won't allow that. But if I ALSO want to have command line support then fuse might be the only way to go. Still have to weigh the options there. Opinions are definitely welcome!
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
remoteStorage had / has a backend for some of this?
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Specifically— I’ve been told that they have some backend libraries that abstracts over cloud storage.
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
At least that's how i read it now at a glance, i might be wrong still.
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
wild, i had no idea Michiel was still working on remoteStorage
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
I'm puzzling a bit in figuring out how to combine all these libraries. It seems to me that there is so much already! The missing thing is just the glue between all of them and the platform-native way to present it.
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Yeah I don’t recommend the remoteStorage spec itself
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Just that the implementation apparently had Dropbox, GDrive etc in it as backends
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
bengo yeah, I think it’s no longer like that but CMS used to be a huuuuuuuuge thing. In general, one of the reasons tech folks are weak at business is because we don’t understand just how much money people will throw at what we see as simple problems
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
> I think it’s no longer like that but CMS used to be a huuuuuuuuge thing
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Oh, ok, I don’t know — I haven’t heard from anyone working on a CMS in a long time. I guess that it’s just that the market has settled around a few big ones.
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
I had a great experience collaborating with folks there; then they acquired Macromedia and overnight it was “ALL YOUR MARKET ARE BELOGNS TO US MUAHAHAR PAY UP SUKARS”
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
(I may be paraphrasing a bit, it’s been some years)
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Hi folks, a question for a rough schedule on how i plan on proceeding with this. I'd love to hear feedback/thoughts on this!
1. First up: finalize the design of this library i plan to make. I have a rough idea but nothing drawn out yet.
2. Gather feedback from at least a couple projects who i'd write a plugin for (i think Sia and IPFS)
3. Build a prototype demonstrating the plugin architecture with at the very least a "native" IPFS integration in windows and linux (macos gets the backseat as i don't have a mac)
At that point when i have a rough prototype i want to submit this for a grant. Potentially from the protocal labs/filecoin/wahtever_its_named_commons_fund, NLnet and perhaps from Sia too.
Is this a schedule that makes sense? Point 3 above is likely a couple months away at the soonest.
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Incase anyone was still wondering here, yeas i'm serious! This needs to exist and someone needs to make it. I want to be that "idiot" building it :D
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
😛 google... "Server Message Block" ahhh gotcha! Yes, i think i would go for that. Didn't know it had a name. But a client-server model like how KDE/KIO is doing it seems very appealing to me.
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Oh wait! Ehm, client-server: yet. SMB as you link to it, probably not. I think - not 100% sure yet - i want to use the platform specific cloud functionality. It allows for features like showing if a file i online/offline/pinned which you won't get with a full native filesystem implementation (like SMB i assume). On linux the story is difficult as there is no nice API.... KDE and Gnome both have their own thing... So SMB might in fact make sense on the linux side!
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
I think there are dragons in there, a lot of people have in the past been burnt working on SMB, but it’s at least widespread and designed for something kinda like that
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
no worries, I am happy to be the Guy Who Remembers Old Shit In Case It’s Helpful
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Is it a cloud storage plugin library that enables using many different storage accounts?
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Oke, so... Read that image from bottom to top. Need to work on it quite a bit to get it accurate but this should give you a general idea if the intent. Does that help you
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
To explain it a little:
• The dashed block right below the "Host File Manager" block is to be read as optional FUSE support. Meaning if you'd enable that than any of the plugins you enable (bottom row) would appear as-if you have local hard drives. No clue yet on how i'm going to make that technically possible but that's is the intent there.
• The "Custom Linux API" block below "Gnome" and "KDE/Plasma" is because the DE world in linux has nothing as general API so i have to make up something.
• The "Cross platform unified cloud API" would be akin to the windows and/or mac API.
• Each of the plugins should then be a breeze to implement. There's a big step missing here, that's how plugins are even handled as there will be a (local) socket there between the cross platform API and each plugin. The reasoning there is to have each plugin run in it's own process so that one badly behaving plugin is only going to affect that plugin and not other running instances. This is in the spirit of how KDE/KIO does things.
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Why...... is native cloud integration in a cross platform manner such a maze of different libraries!! Just look what i was able to find: https://hackmd.io/@markg85/BJ-zHg0hT There's like no (or not much) API sharing going on. Seems like i found myself a good project to improve things for everyone 😄
Desktop for sure 🙂 I was just pointed to the fact that rclone is like a near 100 match on that diagram. Still have to dive in much deeper to see where it stands and if it makes all of the above possible. I doubt it does the desktop integration bits but that's fine too. I much rather contribute to an existing project then to make all of this. Lots of testing still to be done on my end before i start building (or patching) anything.
mark5891
What is depin? Seriously, what is it? It's name implies - to me - unpinning something that was pinned. The description seems to imply decentralized pinning. It's anything but clear to me what this new out-of-the-blue term is...
What is depin? Seriously, what is it? It's name implies - to me - unpinning something that was pinned. The description seems to imply decentralized pinning. It's anything but clear to me what this new out-of-the-blue term is...
What is depin? Seriously, what is it? It's name implies - to me - unpinning something that was pinned. The description seems to imply decentralized pinning. It's anything but clear to me what this new out-of-the-blue term is...
What is depin? Seriously, what is it? It's name implies - to me - unpinning something that was pinned. The description seems to imply decentralized pinning. It's anything but clear to me what this new out-of-the-blue term is...
What is depin? Seriously, what is it? It's name implies - to me - unpinning something that was pinned. The description seems to imply decentralized pinning. It's anything but clear to me what this new out-of-the-blue term is...
What make it register to you as decentralized physiscal infrastructure,
aschmahmann? I'm now curious if it's just my understanding of the wording that confuses me or if it's just really a bad description? Probably just me...
What is depin? Seriously, what is it? It's name implies - to me - unpinning something that was pinned. The description seems to imply decentralized pinning. It's anything but clear to me what this new out-of-the-blue term is...
> What make it register to you as decentralized physiscal infrastructure
I knew what it meant from seeing people talk about it recently 🙃
What is depin? Seriously, what is it? It's name implies - to me - unpinning something that was pinned. The description seems to imply decentralized pinning. It's anything but clear to me what this new out-of-the-blue term is...
What is depin? Seriously, what is it? It's name implies - to me - unpinning something that was pinned. The description seems to imply decentralized pinning. It's anything but clear to me what this new out-of-the-blue term is...
What is depin? Seriously, what is it? It's name implies - to me - unpinning something that was pinned. The description seems to imply decentralized pinning. It's anything but clear to me what this new out-of-the-blue term is...
(edited) ... idea if the ... => ... idea of the ...
bit-
B
i want to use spotify, so i thought about using it as a web app. however i have to use widewine? what is is and why should i / should i not allow it?
D.W.e
G
any ff devs?
I have an extension which requires you to set permissions to connect to a proxy and have to load it manually every time. Is it possible to automate this or set it so it loads automaticallly and sets the permissons?
Have you heard about DKIM? It's specific to email but seems to share some of the same goals: https://www.mimecast.com/content/dkim/ perhaps an extension of that for the web is (good?) step forward? I don't think a DNS-based solution - which DKIM is - would be usable for dynamic sites but it might be an option for static content.
I'm having a couple questions:
1. Once you have data provenance, that only means you have content and a hash/public key/something to identify that content to be valid. But you have to validate it against something. How do you plan on that something to exist? To me that could resemble public keyservers (openpgp) but that's just me guessing based on what i know.
2. What about content that has shared ownership? Like a research paper with multiple authors? Or collaborative work in general. You'd have multiple entities to check before you can claim the content to be valid. Should the current proposal reflect this?
(edited) ... is (good?) ... => ... is a (good?) ...
plebotomisty
P
Hi, I use opera browser infrequently, I noticed that in the current version the free vpn is blocked by the firewall in my country, some of the previous versions allowed you to choose from 3 vpns America, Europe and Asia, current version gives you no choice just a constant failed connection
Anyone know which builds of opera had the variety of 3 free vpns to choose from?
Looking through the (many) open issues, feels a bit like a city council meeting - a mix of practical and crazy. I guess we’ll see where we land…
At one of the city council meetings I went to there was a lady who proposed that we paint on the (currently blank) backsides of the "Welcome to Grovetown" signs.
Very doable & practical, also kinda out-there.
Does the web need this?
What i see happen increasingly more is efforts to make websites behave more and more like apps but using web technologies. PWA is already the "web page disguised as app", IWA adds a layer of security.
It's like rebuilding the classic desktop app development (meaning real development in real frameworks, not electron). I'm not so sure these developments are a net-positive for either the web or the desktop...
boris What do you call a project which is only sort of developed in the open. For example, Signal server source is available but there is the issue tracker is turned off, and they have a sort of weak discourse forum for reporting issues. There isn't even docs on how to run the server. https://github.com/signalapp/Signal-Server
boris What do you call a project which is only sort of developed in the open. For example, Signal server source is available but there is the issue tracker is turned off, and they have a sort of weak discourse forum for reporting issues. There isn't even docs on how to run the server. https://github.com/signalapp/Signal-Server
I feel like if anyone would have coined a phrase for this it would be one of you two
boris What do you call a project which is only sort of developed in the open. For example, Signal server source is available but there is the issue tracker is turned off, and they have a sort of weak discourse forum for reporting issues. There isn't even docs on how to run the server. https://github.com/signalapp/Signal-Server
So, their license is open source, including their code base
boris What do you call a project which is only sort of developed in the open. For example, Signal server source is available but there is the issue tracker is turned off, and they have a sort of weak discourse forum for reporting issues. There isn't even docs on how to run the server. https://github.com/signalapp/Signal-Server
boris What do you call a project which is only sort of developed in the open. For example, Signal server source is available but there is the issue tracker is turned off, and they have a sort of weak discourse forum for reporting issues. There isn't even docs on how to run the server. https://github.com/signalapp/Signal-Server
But they run their network as a closed, permissioned instance. Where clients and servers are all run by the Open Whisper Foundation
boris What do you call a project which is only sort of developed in the open. For example, Signal server source is available but there is the issue tracker is turned off, and they have a sort of weak discourse forum for reporting issues. There isn't even docs on how to run the server. https://github.com/signalapp/Signal-Server
They don’t “practice” open source for their codebase. The name for the open way of working is “commons based peer production” — which they don’t do.
boris What do you call a project which is only sort of developed in the open. For example, Signal server source is available but there is the issue tracker is turned off, and they have a sort of weak discourse forum for reporting issues. There isn't even docs on how to run the server. https://github.com/signalapp/Signal-Server
boris What do you call a project which is only sort of developed in the open. For example, Signal server source is available but there is the issue tracker is turned off, and they have a sort of weak discourse forum for reporting issues. There isn't even docs on how to run the server. https://github.com/signalapp/Signal-Server
boris What do you call a project which is only sort of developed in the open. For example, Signal server source is available but there is the issue tracker is turned off, and they have a sort of weak discourse forum for reporting issues. There isn't even docs on how to run the server. https://github.com/signalapp/Signal-Server
boris What do you call a project which is only sort of developed in the open. For example, Signal server source is available but there is the issue tracker is turned off, and they have a sort of weak discourse forum for reporting issues. There isn't even docs on how to run the server. https://github.com/signalapp/Signal-Server
boris Thanks for sharing, it's nice to have a specific definition to delineate the differences here. "Commons Based Peer Production" feels appropriate to me 🙂
boris What do you call a project which is only sort of developed in the open. For example, Signal server source is available but there is the issue tracker is turned off, and they have a sort of weak discourse forum for reporting issues. There isn't even docs on how to run the server. https://github.com/signalapp/Signal-Server
boris What do you call a project which is only sort of developed in the open. For example, Signal server source is available but there is the issue tracker is turned off, and they have a sort of weak discourse forum for reporting issues. There isn't even docs on how to run the server. https://github.com/signalapp/Signal-Server
the source is open licensed, but everything else is opaque:
• governance
• contribution model
• operations
boris What do you call a project which is only sort of developed in the open. For example, Signal server source is available but there is the issue tracker is turned off, and they have a sort of weak discourse forum for reporting issues. There isn't even docs on how to run the server. https://github.com/signalapp/Signal-Server
community has little/no entry point, let alone “citizen” level engagement.
boris What do you call a project which is only sort of developed in the open. For example, Signal server source is available but there is the issue tracker is turned off, and they have a sort of weak discourse forum for reporting issues. There isn't even docs on how to run the server. https://github.com/signalapp/Signal-Server
boris What do you call a project which is only sort of developed in the open. For example, Signal server source is available but there is the issue tracker is turned off, and they have a sort of weak discourse forum for reporting issues. There isn't even docs on how to run the server. https://github.com/signalapp/Signal-Server
source-available software is also allowed for software you aren't allowed to modify, re-use etc. and this is how i mostly saw it being used
boris What do you call a project which is only sort of developed in the open. For example, Signal server source is available but there is the issue tracker is turned off, and they have a sort of weak discourse forum for reporting issues. There isn't even docs on how to run the server. https://github.com/signalapp/Signal-Server
Source available usually means some form of restricted license, and nothing about the community processes.
bit-
B
in about:debugging I discovered this [attached image].
What does this mean, what are Service workers? Why do I have over 20 service workers from google drive, google chat and more - and do I have any reason to get rid of it from a privacy perspective?
* in about:debugging on arkenfox firefox I discovered this [attached image].
What does this mean, what are Service workers? Why do I have over 20 service workers from google drive, google chat and more - and do I have any reason to get rid of it from a privacy perspective?
* in about:debugging (on arkenfox firefox) I discovered this [attached image].
What does this mean, what are Service workers? Why do I have over 20 service workers from google drive, google chat and more - and do I have any reason to get rid of it from a privacy perspective?
I was playing around with loading images from ipfs if failed from initial source for a side project, in practice would use @helia/http to get the verification also, I think this is a pattern worth exploring. Also the messaging, "We couldn't find your image, but we're checking the Interplanetary File System" || "We are verifying the image with the provided cid" https://codesandbox.io/p/sandbox/friendly-khayyam-5c89k6?file=%2Fsrc%2FApp.js
I was playing around with loading images from ipfs if failed from initial source for a side project, in practice would use @helia/http to get the verification also, I think this is a pattern worth exploring. Also the messaging, "We couldn't find your image, but we're checking the Interplanetary File System" || "We are verifying the image with the provided cid" https://codesandbox.io/p/sandbox/friendly-khayyam-5c89k6?file=%2Fsrc%2FApp.js
Just building out the wanted functionality of Subresource integrity, even if clunky, to tell the story of what is wanted here
dietrich
i wonder if i renamed this matrix room if the bridging would break...
dietrich named the room "ipfs-browsers-and-standards"
i wonder if i renamed this matrix room if the bridging would break...
there’s a difference in display/publish name - i couldn’t change the latter. but i’m assuming any sophisticated Matrix client will show the display name not the raw published name.
I was playing around with loading images from ipfs if failed from initial source for a side project, in practice would use @helia/http to get the verification also, I think this is a pattern worth exploring. Also the messaging, "We couldn't find your image, but we're checking the Interplanetary File System" || "We are verifying the image with the provided cid" https://codesandbox.io/p/sandbox/friendly-khayyam-5c89k6?file=%2Fsrc%2FApp.js
What is depin? Seriously, what is it? It's name implies - to me - unpinning something that was pinned. The description seems to imply decentralized pinning. It's anything but clear to me what this new out-of-the-blue term is...
it was all the rage at EthDenver. neologisms spark wildfires when a few VCs publicly announce enthusiasm for them
mark5891
IPFS Camp in Brussels! Nice! The tickets... ouch, pricey! Is there an option for me to attend with a "sponsored ticket"? cc
mosh
As some of you know, i'm in the middle of an open grant proposal which would make me work on IPFS for about a year or so. Attending certainly makes sense.
Super quick summary for that grant. A cross platform plugin architecture library that allows to "mount" internet based storage as native local filesystem. IPFS would be a plugin.
mosh
As some of you know, i'm in the middle of an open grant proposal which would make me work on IPFS for about a year or so. Attending certainly makes sense.
Super quick summary for that grant. A cross platform plugin architecture library that allows to "mount" internet based storage as native local filesystem. IPFS would be a plugin.
mosh
As some of you know, i'm in the middle of an open grant proposal which would make me work on IPFS for about a year or so. Attending certainly makes sense.
Super quick summary for that grant. A cross platform plugin architecture library that allows to "mount" internet based storage as native local filesystem. IPFS would be a plugin.
Ha, will do! Thank you for the pointer!
ashtrix
A
Best practices for efficient packet analysis: Tips and techniques for optimizing workflow and maximizing productivity.
lidel is there a nice issue somewhere summarizing IPFS CIDs for SRI?
^ Note that the debate in VCWG basically boils down to whether the multihash-like prefixed raw-byte hashes used in SRI could be deterministically mapped to multibased multihashes and whether or not an IETF version of multihash is normative-enough and stable-enough to be relied on for all of the VCWG's "hash-linking" needs
lidel is there a nice issue somewhere summarizing IPFS CIDs for SRI?
so lidel's option A (translate SRI multihashes to CID multihashes in browsers that swing that way) and VCWG both seem to be asking the same question: do we need anyone's permission to use multiformats for this?
lidel’s option A (translate SRI multihashes to CID multihashes in browsers that swing that way) and VCWG both seem to be asking the same question: do we need anyone’s permission to use multiformats for this?
bumping this question up to the room, and sharing w/ #multiformats:ipfs.io for visibility
lidel is there a nice issue somewhere summarizing IPFS CIDs for SRI?
I just want to bring up this sha 256 vs blake3 thing. Yes, a CID is hash-transparent and therefore implementing anything in the browser in SRI should carry the same transparency.
But
There's some massive advantage here when using blake3 instead of sha 256.
Say you have a video file as SRI hash and say that hash is sha 256. You can't verify it's content without getting all it's content. You can verify each block as it flows in but you don't know the full-file hash ever till the moment you've downloaded all the data. Even if you have the local file, you can't say sha256sum file == SRI which is a big shortcoming and a difficult thing to explain for web development.
Blake3 on the other hand (look at this awesome documentation from Iroh) does have this feature. The root hash is the whole file hash, you can rely on b3sum file == SRI You can add metadata at the cost of a little more data but gaining the ability to incrementally verify the data. At this point it's like IPFS with a merkle tree and incrementally verifying the incoming data. Another huge feature that i quote from the linked iroh docs is: The chunk size constant can be modified & recalculated without affecting the root hash. This opens the door to experiment with different chunk size constants, even at runtime. This should be a big winning feature for browsers specifically as you want to get your data ASAP.
Moral of this message? I'm not so sure anymore if going the transparent CID route is the best for the web. While going the transparent CID route is idealistic and nice, it also seems to be limited where tech simply caught up since CID was invented. Perhaps it needs to be tied to a hashing function that delivers some substantial benefits over just simply hashing. Or CID might need a rethink to incorporate the fancy new features we all want so badly (incremental verifiable data).
Just my 5 cents.
lidel is there a nice issue somewhere summarizing IPFS CIDs for SRI?
obligatory: be very careful with making a standard continent exclusively on Blake3
it's design and rationalization almost guarantee it can not be FIPS certified
lidel is there a nice issue somewhere summarizing IPFS CIDs for SRI?
Last i checked blake3 outperforms sha 256 - even hardware accelerated ones - mostly everywhere... so not so sure what hardware validation (that's what FIPS-140 is about) is worth in that case. But you can make the argument i try to make more generic. You want to have something standardized that allows incremental verifiable data with preferably common tools. Using the IPFS tools as a way to verify a CID isn't common nor easy (as a reminder, the default CIDs generated wouldn't even match the CIDs that web3/storage was generating due to (i think) raw-leaves and there was no version change there for cid). I know the rational and reasoning behind it but it - imho needlessly - complicates things. You want a simple way to generate your SRI and to verify it.
lidel is there a nice issue somewhere summarizing IPFS CIDs for SRI?
point remains: there is nothing wrong with the general "bao" construction that blake3 popularized
it's the hash function itself that can not be taken "off the internet" in a lot of contexts
lidel is there a nice issue somewhere summarizing IPFS CIDs for SRI?
I specifically am flagging blake3 the hash, since this is a standards-thread
having bao with sha256 as the "compressor function" would remove the concern (incidentally this is quite literally the design of the tree-hash the filecoin snarks operate over)
lidel is there a nice issue somewhere summarizing IPFS CIDs for SRI?
That would complicate things massively! You'd need to:
• Get that "fancy sha 256" recognized through some standards body (NIST?) before browsers would even consider using it
• Build - and distribute - tooling to use it. Like a sha256sum
lidel is there a nice issue somewhere summarizing IPFS CIDs for SRI?
there's nothing facy about it - it is sha256
I don't understgand why we are arguing - you are literally making my argument
the problem I am highlighting is that there is likely NO path for Blake3 to be recognized by NIST
lidel is there a nice issue somewhere summarizing IPFS CIDs for SRI?
Just so that i get this right, sorry if it feels like asking things twice. Do you mean a merkle tree sha256 chunk structure, like we have right now in IPFS? I thought you meant something more fancy then that with the added advantage of knowing the full file hash in advance. So more "like blake 3" but just with sha256 instead.
lidel is there a nice issue somewhere summarizing IPFS CIDs for SRI?
I am not proposing a solution, just using existing tech as pointers to broaden the common understanding of what's out there
the only statement I am making is that for the context of this thread blake3, specifically as defined in https://github.com/BLAKE3-team/BLAKE3-specs, is very problematic
lidel is there a nice issue somewhere summarizing IPFS CIDs for SRI?
> So more "like blake 3" but just with sha256 instead.
that said ☝️ this is one avenue
moreover that exists: it is almost exactly what fil uses for the data proofs (the difference is that the standard sha256 output is shortened to 254 bits for annoying snark reasons)
it's also the thing w3s is using for their inclusion proofs: https://web3.storage/docs/concepts/podsi/
lidel is there a nice issue somewhere summarizing IPFS CIDs for SRI?
blake3 is annoying a (binary) merkle tree construction algorithm AND a cryptographic compresion function
the first part is awesome, and what we should embrace more
the second - much less so
note that the organizers are, crudely speaking, the polkadot equivalent of the "ethereum foundation":
https://web3.foundation/projects/
that said I might attend if they want to talk XCMv3<>CASA 😎
ProtocolBerg upped the game by being the least maxi-deranged conference in recent memory, now all the historically one-blockchain conferences are adding a chain-agnostic track 😏
ianconsolata
Know folks who have done good work with libp2p? Have you done good work with libp2p? $125k in retroactive funding is now available for rewarding some of the best contributions to the ecosystem. If that sounds like a project you know of, this might be your jam: https://blog.libp2p.io/2024-04-08-libp2p-rpgf/
(I’d put it in a philosophy channel if we had one, but we probably shouldn’t 🙂)
@thompson_plug:matrix.org
T
This message is being deleted…
Phish Bot
This message is being deleted…
+ | Spam Police
This message is being deleted…
dietrich
i get regular questions about peer ids and how they're used in ipfs, their generation and stability, and how/whether developers should use them in their applications as identities, but there's zero documentation that i could find on this, so filed an issue. i know ppl here have ✨ thoughts ✨ so please share on this issue if you have a moment: https://github.com/ipfs/ipfs-docs/issues/1849
"to the emergent properties of interactions between ‘things’ in ecosystems … a move from linear to systems thinking." Interestingly, I always take this approach too, but come at it from a physics perspective where we rigorously study emergent phenomena from individual interactions.
Ian Preston it’s absolutely possible to start from physics as well, but there’s a lot that becomes untractable fast if you want to deal with societal levels of complexity. Ecology and biology have developed a deep and formal understanding of applied complexity in which cows aren’t too spherical, it’s a good toolbox.
If something has become intractable then you simply have the wrong level of abstraction. Of course I'm not going to model societal emergent behaviour by modelling the groups of atoms in each person.
Ian Preston what does modern deployment of java apps on consumer desktop OSes look like? eg say you made an ipfs-desktop-alike that bundled Nabu instead of Kubo.
Ian Preston what does modern deployment of java apps on consumer desktop OSes look like? eg say you made an ipfs-desktop-alike that bundled Nabu instead of Kubo.
Depending on your audience your options are in ~decreasing difficulty for user: 1) give them a jar, and expect them to install java separately, 2) bundle the jvm with your app using jlink https://www.baeldung.com/jlink 3) build a static native image using graalvm native-image AOT 4) create an OS specific installer using jpackage https://www.baeldung.com/java14-jpackage
Ian Preston what does modern deployment of java apps on consumer desktop OSes look like? eg say you made an ipfs-desktop-alike that bundled Nabu instead of Kubo.
Ian Preston what does modern deployment of java apps on consumer desktop OSes look like? eg say you made an ipfs-desktop-alike that bundled Nabu instead of Kubo.
We use 3 on Peergos to build native binaries for mac/windows/linux. It's more fiddly than 2 or 4 for the developer because of the restrictions imposed by AOT, but workable.
Ian Preston what does modern deployment of java apps on consumer desktop OSes look like? eg say you made an ipfs-desktop-alike that bundled Nabu instead of Kubo.
In AOT mode it's basically like Go binaries. No JIT, just a GC.
Ian Preston what does modern deployment of java apps on consumer desktop OSes look like? eg say you made an ipfs-desktop-alike that bundled Nabu instead of Kubo.
Ian Preston what does modern deployment of java apps on consumer desktop OSes look like? eg say you made an ipfs-desktop-alike that bundled Nabu instead of Kubo.
well, maybe better to think about it as an adoption trade-off
Ian Preston what does modern deployment of java apps on consumer desktop OSes look like? eg say you made an ipfs-desktop-alike that bundled Nabu instead of Kubo.
shoulder as much of the user pain as possible to increase the probability they actually use the thing
Ian Preston what does modern deployment of java apps on consumer desktop OSes look like? eg say you made an ipfs-desktop-alike that bundled Nabu instead of Kubo.
Surely 4 is the most end-user-friendly as it has a native installer and OS integration?
Ian Preston what does modern deployment of java apps on consumer desktop OSes look like? eg say you made an ipfs-desktop-alike that bundled Nabu instead of Kubo.
i was thinking the ipfs desktop scenario, where there’s already app packaging etc, and it just manages the external process
Ian Preston what does modern deployment of java apps on consumer desktop OSes look like? eg say you made an ipfs-desktop-alike that bundled Nabu instead of Kubo.
Ian Preston what does modern deployment of java apps on consumer desktop OSes look like? eg say you made an ipfs-desktop-alike that bundled Nabu instead of Kubo.
> it just manages the external process
eg you swap out kubo for nabu
Ian Preston what does modern deployment of java apps on consumer desktop OSes look like? eg say you made an ipfs-desktop-alike that bundled Nabu instead of Kubo.
Ian Preston what does modern deployment of java apps on consumer desktop OSes look like? eg say you made an ipfs-desktop-alike that bundled Nabu instead of Kubo.
Ian Preston what does modern deployment of java apps on consumer desktop OSes look like? eg say you made an ipfs-desktop-alike that bundled Nabu instead of Kubo.
Ian Preston what does modern deployment of java apps on consumer desktop OSes look like? eg say you made an ipfs-desktop-alike that bundled Nabu instead of Kubo.
"when individuals are well informed about the problem they face and about who else is involved, and can build settings where trust and reciprocity can emerge" - I assume you've seen the evolution of trust,
"excessive concentration of power is a threat … it’s not just about prices or output but it’s about freedom, liberty and opportunity." If they really believe this then they should tackle the biggest concentration of the most universal power - money, and fix the tax systems to make billionaires impossible.
Argh, this sucks -- and DNSLink might go away too. I wonder if there's some compromise we can strike, like keeping the more privileged status of the IPFS Companion extension.
As far as I’m aware, there’s no DNSLink implementation in Brave. It’s just the Kubo node that does the DoH resolution and brave just forwards ipns:// requests to the gateway endpoint.
FWIW, the Service Worker Gateway may be a good way alternative and something that we can integrate with IPFS Companion so that folks are at least verifying the bytes. CC
Daniel Norman yeah, there probably is. When you go to a website that has the dnslink thingy, something must detect that. Brave detects that and therefore must have at least some technical knowledge of dnslink. I'd think (assumption) that the code path here splits in brave. If there is a dnslink record, use ipfs to get the data and present it to the user, if there's not then follow the normal dns mechanics.
Well, who invented this type of access to files over internet? I kinda miss some README there, I'm used to know what the thing I am handling is, from different aspects than folder name.
Well, who invented this type of access to files over internet? I kinda miss some README there, I'm used to know what the thing I am handling is, from different aspects than folder name.
?
0831brn
This message is being deleted…
@jackerik:matrix.org
J
This message is being deleted…
+ | Spam Police
This message is being deleted (The message that this message was replying to was deleted.)…
We miss most good things in life just from sheer volume. If you wanna get data onto IPFS I have a decentralized database I've been working on if you wanna collab
a bit of a tangent but still:
does anyone know if there is any effort/traction to get Ledger (the USB crypto-HSM thing) to work on Firefox again?
webmidi, once considered a major victory and potential turning point when it got the go-ahead, is stuck for years and there's basically one engineer doing any of this work.
a bit of a tangent but still:
does anyone know if there is any effort/traction to get Ledger (the USB crypto-HSM thing) to work on Firefox again?
the only interesting potential approach could be to talk to Floorp or some project like that, about out-of-tree patches, as a way to grow interest/pressure.
a bit of a tangent but still:
does anyone know if there is any effort/traction to get Ledger (the USB crypto-HSM thing) to work on Firefox again?
sorry if I sound as if I am dismissing your issue links: I am trying to understand the options from a wider-angle view
if the answer is "only way is got to convince FF" - I know what this means
but because I don't know the entire chain I assume there's more than that
a bit of a tangent but still:
does anyone know if there is any effort/traction to get Ledger (the USB crypto-HSM thing) to work on Firefox again?
the standards position repo is super instructive in understanding how the mozilla folks are thinking about all aspects of these decisions. a bunch of longreads, but helps curb "magical thinking" as developers with opinions on what Mozilla should do or might do.
a bit of a tangent but still:
does anyone know if there is any effort/traction to get Ledger (the USB crypto-HSM thing) to work on Firefox again?
but i'm a broken record there. i left Mozilla 6yrs ago and still feels like a few times a week ppl be like "yeah but maybe Mozilla will..." or "i think Mozilla should" or "why doesn't Mozilla...". it's a great organization in so many ways, but sells a vision that sets developers up for disappointment. and since there's no other avenues for hope, that magical thinking just... sticks.
a bit of a tangent but still:
does anyone know if there is any effort/traction to get Ledger (the USB crypto-HSM thing) to work on Firefox again?
but I didn't say anything like this...
I know how "standards bodies" are
I was on the IPLD team ffs 🤣
I am trying to determine what sets aside a yubikey from a ledger, politics aside
a bit of a tangent but still:
does anyone know if there is any effort/traction to get Ledger (the USB crypto-HSM thing) to work on Firefox again?
a yubikey isn't a yubikey from the browser perspective. it's a class of device implementing WebAuthn, a spec that multiple browser vendors spent years working on together.
a bit of a tangent but still:
does anyone know if there is any effort/traction to get Ledger (the USB crypto-HSM thing) to work on Firefox again?
Mozilla doesn’t get that pass from devs, because the vision they sell makes it seem like they’d obviously be pushing hard to grow the capabilities of the web.
a bit of a tangent but still:
does anyone know if there is any effort/traction to get Ledger (the USB crypto-HSM thing) to work on Firefox again?
> Frame has the keys to the kingdom
what do you mean? They don't have your keys, they just have a coupled desktop app along with their extension which gives them more power and harder adoption
a bit of a tangent but still:
does anyone know if there is any effort/traction to get Ledger (the USB crypto-HSM thing) to work on Firefox again?
Is webauthn usable for crypto-y things (e.g. where the signature happens on the device and you can see what it's signing), or mostly just a thing to support yubikeys?
a bit of a tangent but still:
does anyone know if there is any effort/traction to get Ledger (the USB crypto-HSM thing) to work on Firefox again?
but i tried again recently and doesn’t seem much different from back then
Ian Preston
Big breakthrough in the peergos app sandbox. We managed to convince the browser to let us run wasm in it. So now we can run full fledged apps like the VLC media player, or microsoft office doc converters/viewers in our exfiltration proof sandbox.
Big breakthrough in the peergos app sandbox. We managed to convince the browser to let us run wasm in it. So now we can run full fledged apps like the VLC media player, or microsoft office doc converters/viewers in our exfiltration proof sandbox.
what's the best writeup (even if technical) on the general parameters/scope of this sandbox functionality?
Big breakthrough in the peergos app sandbox. We managed to convince the browser to let us run wasm in it. So now we can run full fledged apps like the VLC media player, or microsoft office doc converters/viewers in our exfiltration proof sandbox.
Big breakthrough in the peergos app sandbox. We managed to convince the browser to let us run wasm in it. So now we can run full fledged apps like the VLC media player, or microsoft office doc converters/viewers in our exfiltration proof sandbox.
Big breakthrough in the peergos app sandbox. We managed to convince the browser to let us run wasm in it. So now we can run full fledged apps like the VLC media player, or microsoft office doc converters/viewers in our exfiltration proof sandbox.
this is the kind of thing where a draft spec or even a deep dive into the pattern could be powerful
Big breakthrough in the peergos app sandbox. We managed to convince the browser to let us run wasm in it. So now we can run full fledged apps like the VLC media player, or microsoft office doc converters/viewers in our exfiltration proof sandbox.
exfiltration-proof web apps w/ wasm is the kind of thing HN eats for breakfast
Big breakthrough in the peergos app sandbox. We managed to convince the browser to let us run wasm in it. So now we can run full fledged apps like the VLC media player, or microsoft office doc converters/viewers in our exfiltration proof sandbox.
yep, I'm curious about what was special about wasm specifically
a bit of a tangent but still:
does anyone know if there is any effort/traction to get Ledger (the USB crypto-HSM thing) to work on Firefox again?
about support for things like webHID in gecko/firefox, Mozilla changed their stance on WebSerial recently (it's now "patches welcome", see https://mozilla.github.io/standards-positions/#webserial) and already implemented webMIDI which are both in the same bucket as webHID in terms of capabilities
a bit of a tangent but still:
does anyone know if there is any effort/traction to get Ledger (the USB crypto-HSM thing) to work on Firefox again?
Gabriele said webMidi will ship when hot plug is added which could be late this year at earliest. serial is "if someone else does it", which isn't a realistic maintenance commitment. doesn't look anything like real change. yet.
Big breakthrough in the peergos app sandbox. We managed to convince the browser to let us run wasm in it. So now we can run full fledged apps like the VLC media player, or microsoft office doc converters/viewers in our exfiltration proof sandbox.
wasm needs sharedarraybuffer, which needs self.crossOriginIsolated. This was true for our outer universal context on the random subdomain, but not for the iframe that code loads on its own domain to load the app code in.
Big breakthrough in the peergos app sandbox. We managed to convince the browser to let us run wasm in it. So now we can run full fledged apps like the VLC media player, or microsoft office doc converters/viewers in our exfiltration proof sandbox.
Big breakthrough in the peergos app sandbox. We managed to convince the browser to let us run wasm in it. So now we can run full fledged apps like the VLC media player, or microsoft office doc converters/viewers in our exfiltration proof sandbox.
Big breakthrough in the peergos app sandbox. We managed to convince the browser to let us run wasm in it. So now we can run full fledged apps like the VLC media player, or microsoft office doc converters/viewers in our exfiltration proof sandbox.
the iframe domain is a 'random' subdomain which in our case is a hash of the path to the app.
