I remember mIRC at this moment. It let an entirely new internet generation connect easier, compared to the chat clients of its' day.
Slack still strikes me as a modernized IRC client at heart. Today's IRC client. Discovered by a new internet generation with today's integration trimmings.
Let's look at IRC rooted technologies as waves of improvement for a moment. Platforms are guaranteed to come and go, no matter how big. Protocols seem to survive.
The need to chat will remain. Protocols like XMPP are important for this reason. Developing XMPP quickly as private protocol builds should be a worthwhile undertaking. XMPP innovation is something I should go learn about more, though.
No one platform can be the entire ecosystem. They all try. Maybe it's vanity, or a little insulting to create digital wells of content that the odds say will shut down one day.
A step like disconnecting is a challenge for that reason. It will invite lots of people to look elsewhere, trying to find how it should be.
It's not unreasonable to provide a limited subset of messaging features with via IRC/XMPP. It doesn't seem unreasonable to improve XMPP itself either.
I'm perfectly comfortable with a business strategy argument and wonder if Slack isn't looking long enough. Taking a 50-year vision might actually be better if something like XMPP interoperability and evolution remained to outlast everyone. Leaving the garden nicer than you found it, when possible, isn't always a bad thing.
Due to this walling of gardens, my phone has no less than 8 or 10 chat apps silo's I can't avoid. (Skype, Hangouts, WhatsApp, Slack, SMS, Messenger, etc.). To all chat visionaries, none of you are that important. I would replace you all in a heartbeat with something that had a longer shelf life.
The tech that keeps integration and interoperability are the ones that I care about most.
Tools like Pidgin or Trillian connected the worlds of MSN Messenger, ICQ, aim and IRC. Connection and connectedness is blissful for a reason - including across platforms.