I was trying to think of how to communicate this to a client, I’m still not sure how to communicate this in a non-technical way. I also think this is valuable for most people to know & think about.
Data, if it changes, has a relationship with time.
As a simple example, we do not refer to “census data”, we refer to “2022 census data” because it’s widely recognized that the 2022 data will differ from the 2020 data. So anytime you talk about that data you must provide a time-based reference.
If a user is added to a chat/messaging thread,
should they be able to see past messages?
The common example I like to use is social media chat messages. Let’s say for example you have a chat group called “silly things said on the internet”. In this chat group you complain about the “silly things Simon Le Pine says on his blog”. Then sometime in the future you add me, Simon Le Pine, to that chat thread; should I be able to see past messages that were posted before I joined the thread?
In social media, generally new users can’t see old messages. In email, being old-tech, no one thought of that so if you CC someone on an email reply they can easily scroll back through past messages in the thread.
On the other hand, when someone joins a LinkedIn group they can see old messages posted to the group. In the case of business-networking, it is generally accepted that new users will see past messages.
Messages, chats, and emails are fairly easy to wrap our heads around. What about when it comes to status updates? If you’re only recording a users current status we have no way to know when they set that status, what their previous status was, or a history of status’. If a new user connects with an old/existing user, should they be able to see the existing users’ status since it was updated prior to the new user joining?
Bottom line, it is very worthwhile to consider the relationship between data & time. Not only what data you’re recording and the timestamp associated but its also worth thinking through users’ expectations around privacy/accessibility in relation to historic data.