Hi, I posted this question to Gitter but I didn’t get any replies so I thought I’d post it here.
“I just wondered what people thought of the Solid long-chat format; i.e the 1 file with all the days chat in, can you see issues with 500+ messages in and scaling? i was looking at Gitter and they seem to grab bunches of messages around 50 odd on start, and then on each scroll up another bunch.”
Thanks.
Replying so you know I’ve read this as much anything. I’m not familiar with the format you mention or how gitter works, but it makes sense to divide things up because if they grow indefinitely eventually it will cause problems with performance, resources and so on. So I’m not surprised gitter would be doing so, and I think any cumulative timeline will have to address this in some way. Not much you didn’t realise I’m sure, but maybe of some help! 
Copying the popular chat apps makes a lot of sense because that is what we are all used to. If there are newer more adventurous features I think they will need to come in a second round.
Perhaps in Shighl lib that is used by Solidarity , the way I use to get data is not optimal. I’ve used @RubenVerborgh’ query-ldlex that not allow to group query as a high level lib.
Using a more low-level lib like rdflib could be more efficient I think. Help is welcome