Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
  • Aug 11 2021 20:52
    @RubenVerborgh banned @mikeadams1
  • Jan 04 2021 20:23
    @RubenVerborgh banned @WebCivics_twitter
  • Jan 04 2021 20:18
    @RubenVerborgh banned @SailingDigital_twitter
  • May 27 2019 06:08
    User @Mitzi-Laszlo unbanned @in1t3r
  • May 23 2019 06:49
    @Mitzi-Laszlo banned @in1t3r
  • May 16 2019 09:49
    @Mitzi-Laszlo banned @mediaprophet
  • Feb 01 2019 22:04
    User @melvincarvalho unbanned @namedgraph_twitter
  • Feb 01 2019 21:49
    @melvincarvalho banned @namedgraph_twitter
Arne Hassel
@megoth_twitter
Hehe, not in our mind, of course ;)
Jeff Zucker
@jeff-zucker
But also by extending the UI vocab you could have consistent classnames etc. which could be used ine either turtelCSS or straight CSS
Arne Hassel
@megoth_twitter
But might be a difficult technology choice that's difficult to enforce, is my point
Jeff Zucker
@jeff-zucker
If there were an existing UI vocabulary that could be used to easily create semantic-based CSS, rather than enforce you could just say - as long as you use these classnames, you can write a theme for the databrowser, not dissimilar to CMSs people are used to.
Arne Hassel
@megoth_twitter
You mean incentivize developers to express the UI of their extensions to the data browsers as data models?
(again, certainly a possibility, but afraid it's asking to much...)
Jeff Zucker
@jeff-zucker
That's an interesting idea but I was thinking of something much simpler for the devs - you all (I say because I'm not the one doing the work and its easy for me to say) create a databrowser that includes classnames standaradized on the extended UI vocab, and you publish a list of how the classnames related to the different parts of the browser, and the theme writer just writes a theme in standard CSS but using the voacb for classnames you created.
Devs like to be handed lists of class names, saves on thought :-)
Arne Hassel
@megoth_twitter
Ah, I was thinking of using the Solid Style Guide by inrupt (which are documented at design.inrupt.com) could be that interface that allows theme designers to write their custom CSS rules
But not married to that idea though, just what I think is reasonable at the current time
Jeff Zucker
@jeff-zucker
Sure, you have to pick your priorities. But even with the style guide you need some way to say tie it to the parts of the databrowser.
Arne Hassel
@megoth_twitter
True - by using design.inrupt.com as the common ground we tightly couple extensions to data browsers to the efforts we at inrupt are doing... So an extended UI vocab might be a better common ground to go about this
But the extended UI vocab might also become just an extension of the inrupt effort either way =\ It is a very difficult thing to solve...
Glad to have this discussion, even though it is a difficult one... Hopefully we'll arrive at a solution that is for the betterment of the Solid project either way
Jeff Zucker
@jeff-zucker
I'm sure you will, good luck juggling all of the good things you can do. I'm sure whichever ones you focus on will bring improvements.
Arne Hassel
@megoth_twitter
Thank you ^_^
Ruben Verborgh
@RubenVerborgh
We are starting an effort to write the Solid technical workings in a clear, implementable spec. The current draft structure is available at https://solid.github.io/specification/, and your feedback is most welcome at solid/specification#5
Jeanne-Kamikaze
@jeannekamikaze
Folks, I think Larry Sanger is looking for you: https://larrysanger.org/2019/06/declaration-of-digital-independence/
Mitzi László
@Mitzi-Laszlo
@jeannekamikaze thanks for sharing, on point
Tim Berners-Lee
@timbl
A good read .. it is mostly .. with one bit of architecture "Accordingly, social media applications should aggregate posts from multiple, independent data sources as determined by the user, and in an order determined by the user’s preferences."
Graham Klyne
@gklyne
See also: https://inessential.com/2019/07/02/no_algorithms, https://inessential.com/2019/07/03/no_algorithms_follow_up. (Gotta chuckle at the final sentence of the follow-up.)
Mark Hughes (happybeing/theWebalyst)
@happybeing
@timbl I had a short conversation with Larry Sanger and interested him in Solid: https://twitter.com/lsanger/status/1147547907962413056?s=20
rat10
@rat10_gitlab
The DeSemWeb workshop at ISWC 2019 was cancelled :-( Was it for lack of contributions or other reasons? And is there any other place where Solid developers meet f2f?
Matthias Evering
@ewingson
Michael Thornburgh
@zenomt
i've been thinking about an issue for a couple days. say WAC allows me (an acl:AuthenticatedAgent) to create a resource somewhere, and i do. is my (strongly authenticated) identity recorded with that resource? i haven't found anything in the specs nor in @timbl 's https://www.w3.org/DesignIssues/ specifically about this. i feel like the creator's (at least) webid should be a maintained metadatum (like its content-type and last-modified time), and especially for this item i shouldn't be able to lie about it. OTOH that information is generally sensitive and probably shouldn't be publicly disclosed by default, though it should be available in some use cases (like "unforgeable author of a post in our group online chat thingy").
do we throw this (strongly authenticated) information away and use some other method to establish authorship? perhaps Shapes can be used to adapt metadata on arbitrary resources (like images) as well as on linked data. is sensitive resource metadata gated by WAC? Shapes?
Michael Thornburgh
@zenomt
is this solved by the resource 1) claiming to be by me and 2) being in my sp:storage? i don't think that's sufficient if any part of my sp:storage is writable by anyone else.
Michael Thornburgh
@zenomt
maybe my profile specifies a container that i assert i'm the sole author of its contents, and i (and only i) put stuff that i assert i'm the author of in there. that might address many of the "distributed social" use cases. stuff that says it's "by me" but isn't in that container: "maybe, who knows?".
@timbl @RubenVerborgh i feel like this must have been considered already. if so, pointers/links would be appreciated.
Jeff Zucker
@jeff-zucker
@zenomt interesting ideas, somewhat relevant is Ruben's remarks about there being no such concept as "pod owner" - https://github.com/solid/solid-spec/issues/190#issuecomment-500343236 ... and it is quite possible to for-example lose ownership of resources and containers in "your pod" (e.g. by messing up a .acl file) ... "owner" and "creator" are both worth tracking :-)
Ruben Verborgh
@RubenVerborgh
Currently, no standardized audit logs
But indeed, there are very good cases for them
Michael Thornburgh
@zenomt
so in today's Solid, is it just the Honor System™? that is, is there currently any strong way to know "who POSTed this thing to my ldp:inbox?" or "who wrote this comment on my vacation photos?" or "who wrote message #5 on our closed group discussion?"
Jeff Zucker
@jeff-zucker
and with which app?
Michael Thornburgh
@zenomt
i was saving that for later. :)
@jeff-zucker i'm pretty sure the answer today for "with which app?" is "not yet". however, i'm curious if there are in fact methods or conventions that i haven't discovered that address any of the cases i enumerated above.
Michael Thornburgh
@zenomt
which would most likely have to be along the lines of "location-based provenance"
Jeff Zucker
@jeff-zucker
There are some interesting ideas in this issue about where things should go, which is somewhat related to where they come from: solid/vocab#26
Justin Bingham
@justinwb
@zenomt auditing is one important element. another to consider is data signing like https://w3c-dvcg.github.io/ld-signatures/
ultimately i think you want a combination of things
Michael Thornburgh
@zenomt
re ld-signatures: i don't understand why they're reinventing the wheel there. we already have JOSE/JWS which can totally work today for any RDF serialization format. but that's a whole nother conversation that doesn't really belong here. :)
however, if there was a reasonable (that is, works easily in the browser) way to cryptographically tie my webid to a signed linked data representation, or in fact to a signed anything, then the webid auth problem would also be solved with the same mechanism (vs webid-tls or webid-oidc).
Michael Thornburgh
@zenomt
@justinwb @RubenVerborgh i think focusing on the term "audit" downplays an important aspect, which is that the creator/maker/author is immediately relevant for use cases like "author of this comment on my photo or message in this conversation" and probably most non-anonymous social activities.
Michael Thornburgh
@zenomt
cryptographic non-repudiation isn't necessarily required for cases like "i trust my own pod / my group's pod / github to tell me what it knows about resource X".
Tim Berners-Lee
@timbl
Talking about netnewsire , it has been on my agenda to generate a RSS/Atom etc feeds from the DesignIssues … Hmmm .. Brent also works on https://www.omnigroup.com/omnifocus/ which would also be a classic solid app
Ruben Verborgh
@RubenVerborgh
Author of a chat message is explicitly captured in the message description in RDF
Sarven Capadisli
@csarven
Politicians Don't Trust Facebook—Unless They're Campaigning https://www.wired.com/story/facebook-privacy-candidates-pixel-campaigning/