Solid browser Long Chat questions

Basically, I found out most of what I was looking for by reading this great document!

Hi, I have been looking over the data / format etc when two people interact and have a conversation in long chat.

When I create a “chat channel” and a friend joins my chat - it looks like the chat gets stored in my chat.ttl I thought the concept was it stores my chat in my pod and my friends chat in theirs, could someone explain?

I note the chat “settings” in the browser UI for both users; how does the settings “Public link to this chat room” and “Personal note of this chat room” work in long chat in general, and for each user?

When the chat messages are stored in the chat owners folder they are stored within 3 folders; by year, month and day this actually makes sense to me as I did the same when I was playing around with a solid social app a little while back when I stored the posts; but looking at the container that holds the index.ttl for the chat there doesn’t seem to be any references to the 3 folders, or at least I couldn’t find any references; is the folder structure format of where the chat.ttl is stored defined anywhere?

Also, when my friend chats with me by visiting the chat channel url on my pod is their pod modified in any way - I can;t seem to find any changes ?


This is a conversation on this topic from the chat channel:

Julian Cole @Julian-Cole03:56

Does anybody know what format I would need to link to a message content in another pod rather than embed the message content in the chat creator’s pod file? for example, here is a truncated extract of the simple chat index.ttl with some chat flow and a message.


terms:created "2019-04-26T18:30:54Z"^^XML:dateTime;

n:content "how are you doing?";

n0:maker c:me.


a mee:Chat;

n1:author c0:me;

n1:created "2019-04-26T18:23:55Z"^^XML:dateTime;

n1:title "Chat";

flow:message :Msg1556303342376, :Msg1556303387381, :Msg1556303454449.

Basically what I’m after is to store each chat in each pod respectively, rather than all the chat in the person’s pod who created the chat - does that make sense?

Tom @sideshowtom 06:11

@Julian-Cole “Basically what I’m after is to store each chat in each pod respectively, rather than all the chat in the person’s pod who created the chat - does that make sense?” – does that mean that chat participants can change what they said?

Tom @sideshowtom 09:44

@Julian-Cole I wasn’t really kidding. One person has a memory of a conversation and another has a different one. The memories evolve independently. And who has the definitive record? Can they be trusted with that?

Kingsley Idehen @kidehen 11:03

@Julian-Cole :

Re April 27, 2019 3:56 AM, is it possible to move that to so it can be discussed in a more memorable manner that’s also useful for future recall?

The answer to this question will also help others understand what Linked Data enables implicitly, ditto Solid app dev best practices etc…

Julian Cole @Julian-Cole 11:11

@kidehen well, at this point I was just looking for loose opinions on what other people thought, I can move it to forum post once I have some more thoughts perhaps.

Julian Cole @Julian-Cole 11:17

@sideshowtom If i did a hash of your chat message and stored that in my pod against our conversation I would know if the chat had been tampered with; I was thinking that chat should work like social posts and comments, in that the message, post, comment belonged to you - not the person hosting the chat?

Julian Cole @Julian-Cole 11:26

@kidehen gitter chat is anything but “memorable”; a few months back i took the time to scroll though all the solid app-development messages from the beginning and I found some really great helpful tips on Solid which are buried amongst the other chat. If I was doing a ‘gitter’ style chat I’d want to add sticky questions / comments and an events board and maybe tags in there too :slight_smile:

Tom @sideshowtom 11:32

@Julian-Cole Well I dont know I guess it depends on who is hosting the chat and what their filters are. Maybe theyre just looking for information relevant to a product they sell and maybe only positive comments, and they filter out and forget negative ones. If on the other hand each participant hosts their own comment maybe they want to revise their comments for some reason, say maybe to claim that they made a prediction that came true, like with a stock. BTW, I agree with Kingsley that this might be better on the forum.

Tom @sideshowtom 11:46

@Julian-Cole I think its a big topic and it would be nice to have it there for future reference.

Some more brilliant :slight_smile: thoughts on a chat app…

Where things are stored and how they’re linked should be configurable, but the expected default probably should be that each participant keeps a copy of the entire chat on their pod, with links to each utterance of the other participants. Then each participant can evolve their own memory of the chat.

Some may translate it to another language, some may enhance it with links using different ontologies, some may condense parts of it, and some may delete or alter parts of it. If you have a link to something someone else said, following it would be like asking them if they remember having said it. They may have altered what they said or deleted it entirely and thats just like real life. For a definitive record of the chat, a listener agent that is trusted by all the participants could be invited. Each participant could configure what they remember, maybe just what they themselves or one or another of the participants said.

Chats could be filtered in real time or later. Everybody needs filters. The only people who don’t have them are infants, and they take everything in. By the time we’re old we have so many filters that only a little bit of reality gets through.

1 Like