Just bumping this to remind people of Solid World tomorrow
I have a message for all Solid people :
!! STOP RE-INVENTING THE WEB !!
This are the words I had in mind this morning when I woke up…
Yes, please, stop trying to reproduce all the tools we all already use and know, let’s go now further more deep in what we can imagine…
I understand that for the first phase of Solid developement that I would call a ‘Proof of Working Technology’ @timbl with the Inrupt’s & Solid Community devs had to re produce what already exists on Solid and say
- 'Hey, you see, my friend, that LongChat works on Solid too!! ’
But developing Shighl tool for Solidarity and now Chat & Inbox modules on Popock show me that we should’nt stay on the highway and take the next ‘aire de picnic’ for having a good discuss on a ‘What a chat could be on Solid’ .
I think that is the way @MitziLaszlo wanted to drive us and we should reuse all the threads she started to rethink what the tools should be…
Just a basic UI consideration :
Imagine a Solid chat mobile app. A basic action could be to mention someone in a chat message…
Webid’s are long with https://spoggy.solid.community/profile/card#me and I use more time to type a webId that to type my message . That could not be used.
@spoggy is not Solid compatible as there could be spoggy.solid.community, spoggy.inrupt.net, spoggy.gouvernement.fr or many spoggy that do no be the same person…
So we could for this think of a good identifier for all apps, not only Chat… So the solid.community.net could be a good identifier but that’s to long to type too and no one want to have on his mobile device app lot of lines of spoggy.development-division.supply-department.the-company-you-work has send you a private message… or a mail.
And playing with Solid for two years show me that trying to reproduce is not a good way, there are lot of micro-issues that would be a mountain to resolve… The way we reproduce a chat flow allow anyone user as @aveltens did to usurpate any other user webid, just using a foa:maker, he said he was @timbl… And that should not be possible in a Solid chat. Sure that a Solid server should use the .meta or something like that to authenticate the author of a message but we should think of what Solid can brink us for a Conversation Tool.
In a conversation, or in a post, we should reinvent the interactions, not only trying to reproduce ‘send’ & but , a user could want to explore semantic data linked to each word if a post, and should find an infinity of exploration possibility. A user could extract a part of the conversation as a new idea and link to it in a article, something like dokieli do with documents, or the _underscoreProtocol propose http://www.uprtcl.io
Please do not reproduce the solid databrowser with the same archaïque tools even with React or friends
We are now sure that all the basic tools (notepod, bookmarks, chat, inbox…) we know can now be basically reproduced, with some difficult micro-issues, and some complicated rdf data flow when we want to implement like, dislike , groupsfunctionalities ( where store that data, what acl, )
I have to go to work for now, but please don’t waste time to reproduce all functionalities of the old web, take 5 mins to rethink the interactions, the UI that can be think now as components ( and that’s another discuss ) and
!!! Stop reinvent the web!!!
Take a picnic pause, with a good food a rethink interactions between actors and in the UI perspective
(I’ll ignore your opening paragraphs which I can’t make sense of). I will address the one UI issue which you raise, because I think there is an important point there that, even though we represent everything with a URI, we should not show them to people, or ask them to type them.
Here is what Imagine happening when a user in a solid chat does an “@mention”. They type an “@” or press a toolbar button which has an “@” on it (As this is the current language people use right now on the web. The icon on the button could more logically be a picture of a person, but either way.)… They are then presented with a list of names or nicknames of the people in the chat, or in a chat related to the group. They chose one of the names, and then the hypertext version of the message gets a an anchor inserted, where the href is the webid of the person mentioned, and the text is the name or nickname they chose from the list. So the mention target is the full unambiguous URL of the person, but the user doesn’t have to see it. They could use a people picker to mention anyone in the solid universe. They could go in another tab to another completely different chat, and drag the person’s avatar from a chat message there, so they could mention anyone anywhere, unlike before solid.
(((This does mean we have to move to a hypertext version of chat (which gitter already has we already have IIRCD in imported text) but that in turn needs a sanitizer for cleaning untrusted HTML or a sandbox where it can be displayed safely. )))
This is a very detailed response to one UI issue you mention. Apart from that, we are not reinventing the web, but very much working within its spirit to the functionality people expect from social networks now, as just one thing. And yes, we do need to think about the user experience from a distance before we jump in, to make sure it is powerful and consistent and adds new value. It would be best to talk about the chat UX though in a forum topic about solid chat, not this general one about Solid world.
I think the way I understood the comment is from a skeuomorphic perspective of not reinventing the same things in the new thing and to imagine new things that flow from the proposed new directions .
From this perspective, I think it may be boiled down to avoid getting caught up in “here’s a ‘hello world’ and ‘todo list’ in SOLID” direction .
These implications for UI and “human readability (UX)” are interesting for me, as I’ve always been fascinated by the opinionated implementations using rdf label in place of URI/URL whereby the opinion is normally system level imposed on the users, rather than user declared/discoverable preferences .
BUT those things are presenting DIFFERENT information from the actual identifier of what is being referred to as well all know, so what matters is authenticity and preference .
Someone’s WebID shouldnt he assumed to be replaceable with someone “name” .
Similar to storage / footprint discoverability, this will also be interesting what patterns emerge .
Are there recordings of the talks and will they be made available?
The recordings are added to the This Month in Solid newsletter as well, so I recommend following that as well.
Sorry , I didn’t want to hurt anyone and I’m really thankfull to all you did & glad to all the work is done around Solid.
What I wanted to say is that we know now that we can reproduce linear Chat mechanisms on Solid, but perhaps we should start to imagine what a linked Chat could be where we could not only post message but also link pods resources, fragments, meetings into that Chat, ,…
Something like Dokieli aim to propose, a graph Chat…
And this is not only for Chat but that could appli to other tools like inbox or Contacts… That’s why I thought it could be integrated to roadmap… At least it is in mine… Think of new decentralized tools, not reproduce linear classic tools…