DevCon talk Supercharging the SAFE Network with Project Solid

Developers

To developers I say, please don’t wait for me or anyone else on this. It took me a while to get here, and as just one slow old guy I can’t carry this by myself. So if you are interested I hope you will dig in, learn, play and I’ll be around to help anyone who interested in this stuff.

I’m going to be doing other non SAFE stuff that I put aside to get to this point, so won’t have as much time as I have been for a few months at least. So a good use of my time might be helping others do stuff with this, although I do expect to keep coding when I get the time.

So any questions, any idea you want to discuss, please get in touch.

Future Activity

Some of the things that could be being worked on (let me know if you are working on any so we and others can co-ordinate):

11 Likes

:blush:

https://twitter.com/timberners_lee/status/989245559277019136?s=20

Tim also retweeted this :slight_smile:

https://twitter.com/safepress/status/989198557633794048?s=20

37 Likes

wooo :clap::clap::clap::clap::clap::clap::clap::clap:

8 Likes

Woohoo that’s massive! :tada::heart_eyes:

8 Likes

That is great to see! I was hoping positive vibes would reach Tim!

I thought it was a great presentation and I didn’t realise you had taken it so far already! I’m both pleased and impressed! :slight_smile:

15 Likes

Mark, its no more than you deserve chap. All this work will pay off bigly :smiley: seriously though, this is one area of the front end that really matters a lot. Semantic plus security and privacy is massive. If we can throw in some homomorphic encryption (sorry for the horrible word) then we can have true medical research of mass data and do so privately. That would be immense actually. Lots to do but this along with truly scalable and shared neural nets and we have a hyperspeed jump for AI research where it matters. Spotting causalities and comparing genomic and proteomic data will save lives in my opinion.

Then it really begins, the real Internet :+1: one that works for us and us alone, not for corporates or governments, but for us the people, 100% of the people. Hats off to you chap.

49 Likes

Very happy to be proven wrong in regard Sir Tim never supporting Safe. Congrats to Mark for persevering with this.

17 Likes

This is what making the world a better place looks like. :earth_americas::earth_africa::earth_asia:

7 Likes

Timely article advocating personal data ownership using Project Solid on a decentralised cloud:

5 Likes

Admittedly I’ve had a hard time wrapping my head around the what and the why of SOLID.

I finally found a resource that gets at the heart of the matter when I simply searched semantic web: Semantic Web - Wikipedia

12 Likes

2 posts were split to a new topic: Redlands “Enterprise” smart city

SOLID is a recent development around the Semantic Web and I’m still getting up to speed on that one, but for the Semantic Web and Linked Data in general there are a couple good books. The ones I’ve read are maybe a bit too outdated now, but from the ToC this books looks like one of the better one that covers all the pre-SOLID basics quite well. Part III seems to be largely about server-side tech and might not be all that relevant, but I think the rest looks solid.

3 Likes

Mark (@happybeing) & Ian (@CommunityLink) - I’ve moved the posts into a new topic:

I hope this is okay with you both 'cause I agree with Mark that this isn’t the place for that discussion.

Thanks,
David.

4 Likes

Imagine some network like SAFE had it’s API as an ldp/solid API and forced (yes I know) the use of semantic data sharing and privacy in its apps. It would be a jump for devs right now but not a big one I feel. Then we could have the schemas for messaging, chat, docs, videos, research etc. all categorised and searchable. With all the privacy components of SAFE and the accessibility of SOLID, I think there could be a very compelling combination of two projects that have very related goals, but each strengthens the other, I love win-win situations and this one feels like it is heading in that way, well to me at least :wink:

Then we could have many apps using the same schemas, so you don’t like my messenger app, OK switch out the app, but keep your data, this seems sensible and also empowering for users, it takes our message of no need to worry about infrastructure to a new level and takes SOLIDs goals and makes them very accessible to devs and consequently, people.

Just a thought, maybe though one that has brewed for a long time :smiley:

25 Likes

(whispers from behind a corner) dooo iiiiit, dooo iiiiiiit!!!

12 Likes

Eggggssssssackly! :slight_smile:

With regard to the ‘forced’ bit, I don’t think that will be necessary. I believe that devs may end up with little choice for two reasons:

  • Firstly users will prefer apps which allow them this freedom, so devs who don’t do this will find it harder to acquire users on these new platforms.

  • Secondly, by following this path devs will be able to build apps that have much better features and are much more useful.

At the most basic level, an app that uses LinkedData can pull in different data types available in the user’s own storage (produced by other apps) and from public resources.

That can be for discovery/search, analysis, automatic citations/quotations/excepts, analysis/statistics or simple mashing data together from different apps. For example to create a social news feed.

Your social feed pulls in data from friends who share their data with you and vice versa. This can include any data they share, and the data you choose to include in your feed (posts, status, photos etc) in the order you choose with the presentation you prefer.

No more content forced on you for Facebook’s own purposes. No more crawling of your data, except by the individuals or organisations you trust with it, and only on your terms. No more “click here to agree or you won’t be able to use a social network with all the data you uploaded”.

Imagine if the EU were to force Facebook to offer users downloads of their content in LinkedData format. It would be trivial for people to migrate their data to Solid servers or Solid+SAFE and keep their social graph in tact. They can then select a social app they like - more likely many different apps, all capable of displaying timeliness and creating posts with the presentations you like best - and unable to spy or access your data without your permission.

It creates the world we want, and I don’t think there will be any need to force devs to take this route once we have some decent Solid+SAFE apps running which other devs can improve. Facebook on the other hand would be mad to let users export their data in such a portable format, but it might be an attractive way of solving the current problems, so I wouldn’t rule it out once enough people realise the potential. I explained that to a guy on twitter recently who said Facebook were consulting and were open to ideas like this. I’m skeptical though :wink:

22 Likes

If the network exposed an API to make data uniformly shareable, machine-readable, and searchable, what definitions would it agree on, how would would new definitions be composed, and how would the integrity of the definitions be maintained?

Specialised and tightly-controlled semantic schema may be more about addressing these questions and Uncertainty Reasoning for the World Wide Web.

Reading about Gene Ontology project, which is to

provide a uniform way to describe the functions of gene products from organisms across all kingdoms of life and thereby enable analysis of genomic data.

It makes sense that Gene Ontology is overseen by a trusted council which addresses the above questions and also

colloquialisms, community preferences, abbreviations, legacy names, the multiple ways of referring to chemical elements, capitalization, and all the possible variations that occur in natural language.

It seems to make more sense that each domain and associated community like gene research, astronomy, video media, audio media, etc, is more equipped in their specialised knowledge to oversee the structure and maintenance of their data semantics than the whole of the SAFE or other network.

One option is to align with http://schema.org/, porting the current structures and definitions over to the network, and being involved with the Schem.org Community Group

Although I read in this issue of Scientific American a piece of information that gave me the impression that the semantic web could benefit more from disparate schema, however comparable data, where something called inference engines could be used to find the keys that connect disparate schema:

Finding relations among different sources is an important step toward revealing the “meaning” of information.

What this might mean for the network is focusing on providing the essentials for building semantic schema but not committing to any in particular.

The last question I’m thinking about is how to scale the composition of RDFS.
For example, the simple act of a person uploading a video. How is the raw data annotated to conform with RDF? Is the user prompted to add standardised tags?
Seaching around Apache Jena and Stanford’s Protégé to find how they may address this problem.
When uploading videos to YoutTube, one manually enters tags to describe the video, but again, at what layer do we concern ourselves with this tagging? Is Schema.org concerning itself with video descriptions?

13 Likes

Very interesting questions Hunter. This is an area I haven’t looked into really, so my thoughts are raw and ill informed but I’ll be generous with them as usual anyway :slight_smile:

I’d say that it isn’t for the network to decide details like which ontology for which area (which I think is along the lines you conclude too). For one, how would that be enforced? Another, who are we to say - when there are others who understand those ontologies and the implications of such choices?

So I assume that our role is to provide implementations of protocols on top of SAFEnetwork that can handle the semantic web, such as Solid/LDP, and that it is for developers to decide how they use this.

There is no reason we shouldn’t publish guidance I guess, but I suspect that already exists and that maybe we should just find and reference that.

Similarly, I think there is a role for tools and other resources to help people answer these questions, and I see evidence of these - such as an index of ontologies which can be searched (I think I posted a link to one last year in one of the earlier threads about Solid). So we could help by searching for things like that and building resources to help people coming to this and asking the questions you asked.

Getting into ‘how would a video be tagged?’ etc is for me moving into an application domain, because for different applications and different media types etc, the answers may be different. Good questions, but maybe for another topic. I think this overlaps with JAMS so there’s an opportunity to discuss this with a real use case, and I know @Nigel is interested in whether and how RDF might be useful there.

5 Likes

If SAFE adopts RDF conformity and exposes an associated API, won’t the RDF only become useful once uniform resource definitions are adopted?
I don’t see yet how RDF conformity is decoupled from the resource definitions and how it doesn’t immediately concern the network with the application layer.

3 Likes

I warned you I was ill informed :slight_smile:

I think this is an interesting question. To recap, I’m thinking firstly that we can’t force devs to adopt a particular ontology, and secondly that’s not a bad thing because we can’t ensure that what we specify is optimum, or even adequate for all applications.

Where similar domains are covered by applications using different ontologies I think that’s a problem for the application layer. It can be tackled by the application, and it’s an opportunity for tools that assist in reconciling data encoded using different schemes.

I just read the Wikipedia link you posted, which is excellent BTW, and there it mentions that even within a large ontology there can be a problem with multiple terms for the same thing. It then says that machine learning might be applied to deal with this.

I can’t say much more so this would be a good question to ask on the LinkedData chat or the Solid chat if you fancy exploring it?

3 Likes