What app would you love to see on Solid?


#1

What app would you love to see on Solid?

(Please also add yours to the list yourselves.)


#2

Solid Office. Something like Libre Office (a free open source office suite) that uses Linked Data.

Ok, that’s more than one app so I cheated :wink:


#3

I would like to see a content enhancement app, that could take plain text and turn it into linked data. It should work in configurable stages, where in each stage the output from the previous stage would be further enhanced (made into a linked data graph) using one ontology, with the result after several stages being a graph of linked data using several ontologies. It should do all this in some kind of transparent or open way, without relying on external black box AI machines. It should be for built for Solid and run in the browser without requiring special permissions or the installation of anything. Thats what I want for Christmas, Santa.


Server side text indexing
Assistant. ( AI assistant talking like human )
#4

Pod View Apps. I envisage multiple apps here, but it takes just one to get this rolling.

We will I expect, have some standard locations for our data, such as folders named documents, music, pictures etc, but over time these will fill with lots of data, so we’ll also need to search this, and I think with the power of Linked Data this can be a lot better than what we are used to from search, and could well become the preferred way of interacting with our data.

For example: interacting through ‘views’ of all the data we have, that can leverage the semantics and other clues such as folder location (music v pictures v documents etc), access control (public v private v group v individual-x), links (related to x) and of course the resource name and content.

Essentially views that can be standard, or customised, or which are really just a nice UI for a search engine tailored for personal data stores, but which could of course reach out from pod to pod.

These can be fine grained or coarse as required, and the beauty is that there will be different routes to finding the data we want instead of just a fixed folder structure, or search by name. A route for each individual’s way of doing or imagining things, and a route for each task, each mood, instead of struggling to find the precise folder among thousands where that data ended up months or years ago.

I think this is a big topic, but we can start with one app and spawn a family of different Pod View apps. We can switch from one to another, and apps can build on features from other apps.

This is one of the beauties of the openness of the Solid ecosystem for me, where data and apps are no longer tied together, the potential for creating better apps, that make better use of the data, is multiplied many times.

Thanks to @JornWildt for the conversation which crystallised this suggestion. I imagine others have already expressed this idea, so I take no credit. Oh, well, if you insist :blush:


How data is partitioned?
#5

Two,
A local host server for windows
Anftp option to transfer code to my pod.
fyi
I develop my web site on S expression Web 4 CSS, html ,php
I want to transfer it to Solid
I want to test changes locally before ftp ing to my pod.

My web site is on a Raspberry PI. djb.mypi.co.
Now retired but always a big fan of |TBL.

djb


#6

A full-fledged database with ACID, CRUD, etc. Non-RDBMS such as graphDB would also be nice but my personal needs would best be met with relational.


#7

Privacy auditor — show me which data is shared how, for example in groups, but novel visualisation will be better; show me what access right a particular user has.

Automatic privacy settings/ACLs — sensible defaults, possibly time limited in some cases; interactive.


#8

Suggestion: Maybe Apache Stanbol can be used ?


#10

Hi Andreas,

I used Stanbol, thought it was great, and got these ideas from it. The thing about it is that it’s a pretty heavy Java OSGI thing and I don’t know how someone would easily run that on their pod. You could have a separate Stanbol server that someone else maintains but then you really don’t know if you can rely on it or configure it the way you want. I think the content enhancement can be done in JavaScript so there is not an unknown server involved.


#11

An email client! Would like to see how that works with PODs.

@MitziLaszlo Is it maybe possible to list all ideas in your first post? That would make it easier to quickly see what is already suggested.


#12

The first thing you need to do with Solid is prove the superiority of the concept in a capitalistic fashion. In other words, it needs to be demonstrated that Solid is built for commercial purposes. Most people aren’t going to be too excited by reading about “linked data” blah blah blah. Besides that, Solid should not be aiming to impress “most people”. Solid’s key demographic is INDIVIDUALS WHO RUN LARGE FINANCIAL INSTITUTIONS.

To me, the exciting part about solid is a unique ID, constantly encrypted, with a personal encryption key that can be used on transactions across the internet. The big deal here is not having everybody’s credit card #s stored in one place for hackers to concentrate on attacking. Data migration and emails are all very nice but that is not what is going to compel actual companies to put money into developing for Solid.

If you just develop an app for Pizza Hut or Dominoes or someone and prove that you can increase the security of actual financial transactions, the battle is won. Tim Berners Lee has discussed how the internet was not designed for privacy; it was designed for sharing and that led to huge security weaknesses when the web became a commercialized tool. Well, DEMONSTRATE HOW THIS SYSTEM SOLVES FOR THAT!!

Solid could spend years as just another ecosystem fighting for its niche. However, if you demonstrate that it is in fact far less vulnerable to hacking–and even invulnerable to a large-scale hack, you will have companies pouring billions into its development.

So give us a pizza ordering app or partner with a bank and adapt their software to work with Solid. Get even a midsize bank to allow you to migrate their user-interface over to Solid and then invite people to attack it.

If you can do that and it works, nobody will have to come on this forum and ask for any apps in Solid because EVERYONE will already be migrating.


#13

Made it into a wiki so anyone can edit


#14

Naive question maybe, but why can’t the financial institutions just do it themselves?


#15

They can. However, that question presupposes that they want to. Solid is a new technology. It’s unproven. It is perfectly reasonable for a company to be reticent to spend their money implementing a technology that may or may not be supported ten years from now.

There are any number of technologies constantly competing for preeminence on the web. Solid needs to show why it is the technology that deserves to win. If someone can demonstrate that Solid is significantly better vis a vis web security versus the status quo, there won’t be an ecommerce company anywhere that doesn’t use it. Once that happens, support for the platform is guaranteed and all of the other apps that make the world a better place will come to be.

Trying to emulate things that already work fine with current technology and show how Solid allows for a slightly better user experience is frankly a waste of time. Number one, a company like Facebook isn’t going to want to implement a technology that leads to information portability and therefore allows users to pick up and leave; so an existing social networking app being an early adapter is out. Number two, email works fine already and people don’t really care what platform it is running on so long as it works. The same goes for data migration. Data can be migrated just fine right now. So those are out. Number three, data portability is a great thing, but making the average person realize that is something else. People get comfortable using a product and they tend to stick with it–hence the first-mover advantage.

When you are in a fight (in this case a fight for web supremacy) you go for your opponent’s weakness. The current web has very weak data protection. Facebook, Yahoo, Google, Amazon–you name them, they’ve been hacked. Billions are being spent to try to protect data. Solid is designed around encryption and decentralization. Show everybody from banks to your local auto parts store that they can solve this problem by moving to Solid and all you have to do is get out of the way. But don’t expect them to do it for themselves. Having a great product is only half the battle. You have to make people realize how great it is.


#16

I think another weakness of the current web is the lack of collaboration between apps/systems due to the data silos. If we can show what can be done through app collaboration or just through data sharing, Solid will have also a chance.


#17

An idea from @tuelsch and @pheyvaer on https://gitter.im/solid/app-development – A hub for ontologies, where ontologies could be registered and that could notify subscribers of new ontologies or of ontology changes.


How to build and publish my own schema?
#18

An idea for a guided browser, due to @JornWildt here, where a browser or browser component could be guided by a guide or data shape kept on a pod, and remind the user with visual cues when the browser strays from the guide, and/or restricts the user to browsing according to the guide.


#19

@tuelsch How difficult would it be to implement that in the web app you already have?


#20

@pheyvaer, @tag42git I don’t think it would be very difficult, allthough the feature to submit your own ontology is missing at the moment (besides pull-requests, but thats for devs only). Users with a pod could get their notifications delivered to their pod or, alternatively, plain old emails.

Another feature I’m looking into, is visualizing the relationships between properties, classes and ontologies (like http://visjs.org/network_examples.html but nicer), to enable visual exploration of whatever “thing” you’re currently viewing.

Also, I might have to step up the indexing game server-side, because the current implementation is not really scalable to, say, 10’000 ontologies (like swoogle claims to have indexed).


#21

Sounds great, but don’t let the best be the enemy of the good :slight_smile: