211–220 of 263 posts
Maybe I've been programming too long, and I saw Solid POD and I read "a solid pile of data." Perhaps that was the intended play on words. I never saw the english word "pod".
It's like two peas in a pod. Or a pod of whales. https://www.merriam-webster.com/dictionary/pod
Earlier quoted context omitted.
Would the content-addressable distributed web really "give ... users the same friendly experience"? I'm a bit outside my wheelhouse, and I don't fully understand what Solid means when they say "You own your data", but with content-addressable strategies, often users _don't_ really have ownership once something is published in the sense that you can't delete or change something if anyone else is serving a copy.
Very much depends on the concept of ownership, doesn't it? ;) I mean, you own it, but so does everyone else. I think the traditional(?) concept of ownership you are referring to, that allows people to "delete" information is very unnatural and feels akin to trying to reverse entropy. You can't unring a bell, and you can't delete data from the world.
I guessing it would be less likely an issue if your friends / associates have a downloaded copy of your messages for the past year if each week a different key pair was made to decipher a section of the data store or something.
Paging Kenton Varda and the company/product/system he and his team created that does something very similar called Sandstorm: https://sandstorm.io/ . (Kenton also led the protocol buffers team at Google which is their main format for data interchange). I'm curious how this is different.
It looks like Solid has progressed since the last time I really looked into it. But, from what I recall, Solid focuses on standardizing data formats and storage... but not compute. That is, all of your personal data is stored in some central location chosen by you, and then web apps can access that data (if you give them permissions). The web apps themselves still run on their own servers, controlled by the respective developers.
I'm skeptical of this model because:
1) If the code still runs on the developers' servers, there is no way to place technical restrictions on what they can do with your data. They can make a complete copy of whatever data you give them access to, they can store additional telemetry on the side, etc.
2) I think developers will resist standardized data formats because it makes it hard to develop new features. If you want to build any feature that requires storing additional data, it needs to be supported by the format. Perhaps the formats are extensible, but if multiple vendors do not agree on the extensions, then your data is no longer portable, defeating the purpose. I find it hard to imagine that any developer will voluntarily restrict themselves in this way without a huge incentive, and I don't see what that incentive is. (Certainly, not enough people care about data sovereignty for that to be an incentive.)
My opinion is that data sovereignty efforts must focus not just on storage, but on compute. The servers on which code actually runs should report to the user, not to the developer. Developers should build apps, which run on the user's servers. This way, developers are still free to create whatever data formats they want, but the user ultimately controls the storage. Other developers can attempt to develop "compatible" software which can read the same data, but this doesn't hold anyone back from adding new features to their own software.
But my own attempt to create such a platform didn't work out. So, you know, you shouldn't necessarily listen to me...
applications :( Does anyone else feel nostalgia for the pre-"web app" days of the internet? I'm talking about personal sites on Geocities and web rings built on communities of shared interest. The browser was an application for navigating hyperlinked information. Other applications include email clients, news readers, FTP clients, and IRC clients. You never had to download a megabyte of minified JavaScript just to re…
http://wiby.me is a search engine that indexes pages like the ones you miss. Full disclosure though, I made it.
Earlier quoted context omitted.
AFAICT there are literally no apps to try this on. Why did they decide to announce this and publicize it before a basic, well-polished, example was ready? The pod signup process is using vanilla bootstrap, there wasn't even a visual design pass.
Well, I guess we just asked the opposite question: Is there any reason to keep this sub-surface any longer, and we didn't find any such reason... :-) So, it is just "release early, release often". So, the code and the spec has been out there for a long time. I mean, some of this stuff is really old. Us geeks have been thinking about decentralized social networking since the dawn of ages, and it seems like the rest of…
I read the website and I don't really get it. If I give an app read access to my data, they've got a copy. Maybe they don't have the canonical copy, the original, but they've certainly got A copy. How do I have control?
As far as I can tell you've reinvented Facebook with a better API. Why wouldn't someone just embrace and extend to take control? "Oh yes, we made some changes to our PODs so you won't get full functionality with other people's PODs. You can just move your data to our company with a click of a button, CLICK HERE, and everything will work great! Also, we no longer support exports of data, sorry. But we'll take really good care of your data, we promise!"
Maybe I'm dumb and I just don't get it, that's certainly possible. But maybe you're not doing a good job of explaining it.
Earlier quoted context omitted.
I don't recall anything on the internet being instantaneous 20 years ago. Surely you're misremembering? I also don't find that new smartphones have much touch-based interface latency; but even so, that kind of latency doesn't have anything to do with the internet, nor the halcyon days of late 1998. To my recollection browsing the web in the late 90s was far more exciting but also far less useful than today. The moder…
My computer from 20 years ago was a lot snappier than anything I've used today. And this is not just nostalgia, there is data to back it up: https://danluu.com/input-lag/
The w3.org site describing solid is much better (even if some links don't work): https://www.w3.org/community/rww/wiki/SoLiD https://www.w3.org/DesignIssues/LinkedData.html inrupt.com site is just marketing hype and crap experience.
After reading what little documentation is available and installing a single-user solid server myself, I wonder if this is how people felt when the World Wide Web was first invented. I find myself asking: Okay, now what? Except I think even the World Wide Web, when it was first invented, at least had a demo that worked. Solid doesn't even seem to work. Following instructions for "learn[ing] how to install and run you…
This may be hint regarding how to authenticate: https://github.com/solid/node-solid-server#testing-solid-loc... > In order to really get a feel for the Solid platform, and to test out solid, you will need the following: > 1. A WebID profile and browser certificate from one of the Solid-compliant identity providers, such as solid.community. EDIT: Seems if you run `solid init` and accept the defaults, it will configure…
Hmmm, a centralized WebID eh? No thanks.
Even the "How it works" page [0] indicates how your solid pod becomes your id.
"In order to prove ownership of your data, you need a way to identify yourself. Rather than relying on a third party, you can use your Solid POD to say who you are. So no more “Log in with X” or “Log in with Y” on the Web — just “Log in with your own Solid POD”."
It'd be nice if it didn't replace one type of SSO with another, but proposed a way that SSO wasn't required at all.
After reading what little documentation is available and installing a single-user solid server myself, I wonder if this is how people felt when the World Wide Web was first invented. I find myself asking: Okay, now what? Except I think even the World Wide Web, when it was first invented, at least had a demo that worked. Solid doesn't even seem to work. Following instructions for "learn[ing] how to install and run you…
AFAICT there are literally no apps to try this on. Why did they decide to announce this and publicize it before a basic, well-polished, example was ready? The pod signup process is using vanilla bootstrap, there wasn't even a visual design pass.