Content/getting started with conversation js sdk tweaks

[adambutler]

• Change UUIDs - Changed to match write the docs guide, prefixes kept in place. Remove username - It wasn’t being used anyway, most implementations would use the session. Plus it was confusing that the name that came back wasn’t the username the user entered but rather the one setup with the API.
• Simplify HTML - Previously there was a mix of elements that shared the same names, some as class names some as IDs. Was confusing.
• Reference but remove the call to app.getConversations in favour of simpler singular getConversation given that we know the ID anyway.
• Replace confusing series of inline anonymous for ConversationClient promises in favour of named functions.
• Adds datetime of message.
• More descriptive variable names.
• Remove need to bind conversation to global scope, also removes the need to setup an unassigned variable at the start.
• Modernise JavaScript. Tested working in Chrome / Firefox without transpiling.
• Added a final result at the end.

[AlexLakatos]

I like the class approach, even though it requires considerable extra work for QS 2 and 3. I guess I'd better get to it then. On the whole "Modernize JavaScript" bit though, there is the ongoing discussion of targeting a larger audience with ES5 or staying relevant and on the edge with ESx. Looking through the building blocks in NDP it seems we're already using some ES6, so I guess I could live with having ES6 here as well.

[leggetter]

On the whole "Modernize JavaScript" bit though, there is the ongoing discussion of targeting a larger audience with ES5 or staying relevant and on the edge with ESx.

@AlexLakatos as our JS Advocate this is your call.

We do also need to remember that we need to update the Node library (beta) and add additional commands to the CLI (beta) for managing conversations, users, members etc. Does the class approach provide enough benefit to mean the rework outweighs the benefits of doing these other updates sooner?

[AlexLakatos]

That's where QS3 is anyway, using the ES5 function as class. It's not rewriting everything, it's just mashing my changes with @adambutler 's. And I think it's worth taking the extra time to do these here and now, it makes the code look cleaner and loose the "work in progress" approach in favor of a "production" approach.