Try Gitter as a beta test for possible replacement for our Slack in the future

[LearningNerd]

Please discuss in the thread the pros/cons you see with possibly switching away from Slack and using Gitter (https://gitter.im/) instead for our community's chatroom.

Update: our trial Gitter is here if you want to test it out: https://gitter.im/LearnTeachCode/Lobby

On the topic of how to go about switching over or testing if we like it, I totally agree with @TheBeege's suggestion here:

...based on my last company's move to Slack. They identified people likely to be early adopters and got them in on it as kind of a closed beta. They helped flesh out the channels, figure out a channel taxonomy, setup custom emotes, and all that sort of stuff. Once the dust settled from all the customizations they'd done, they started communicating out a date to move everyone to Slack and allowed teams to move over early. I think that strategy would be helpful here, too.

Any other suggestions, post em here too!

[rouzbeh84]

I've advocated this a bit and think the only barrier would be getting to switch over. It could be a good filter for people who are serious vs those just getting their toes wet though? Wouldn't be too difficult to run concurrently with the plethora of all in one messaging services these days (Franz, Rambox, All-in-One Messenger).

Definitely like the idea of making it more of an 'exclusivity' type thing though!

[TheBeege]

I volunteer as an early adopter :3

[TheBeege]

Sorry for the double-post. I think that in order for us to figure out pros + cons, we'll have to start using it. I imagine some of us, like myself, haven't used it yet, so we need a chance to try. I get the sense an early adopter thing is kind of inevitable for us to figure out good strategies

[rouzbeh84]

Sure, I'm pretty familiar with it as I use it fairly regularly. I personally use the desktop app but the in browser experience is just as good. I mean, when it comes down to it, its almost identical to slack. I don't think it has all those shiny just released features like video/audio stuff though

[TheBeege]

Here's my assessment so far

The bad:

• The emoji selection, while extensive, is not customizable.
• There's no here or channel mentions
• Names and icons are bound to your Github account (this could also be considered good)

The good:

• Dark theme
• Custom integrations are super easy
• Github integration is super nice

@x2adrew and I were talking about trying Mattermost. It is self-hosted though, not provided as a service. I'm gonna give setting it up a shot before I go back to Seoul

[armaneous]

Honestly, the no reactions to messages really changed the feel of conversations and I didn't even realize how much we used them until they weren't there. The chat looked so... flat.

[armaneous]

And I'm actually a huge fan of a self-hosted solution, especially one that's open-source. We get to own the data and could do some cool data science shit with it.

[rouzbeh84]

@/all is 'channel/room' based and I think the twitter login would probably use the twitter info but I've already linked github and failed miserably testing this out as I have the same pic for both.

I'm also pretty open to any and all self hosted solutions as I've suggested a self hosted gitlab/mattermost a couple times before haha

https://github.com/Kickball/awesome-selfhosted has solutions for all types of stuff.

[TheBeege]

So I have a MatterMost server setup at http://learn-teach-code-mattermost.us-west-1.elasticbeanstalk.com. This would not be a final production instance. It's just one I threw together on Amazon EBS for testing.

Do we have any specific criteria for accepting a new chat solution?

[ghost]

Hey, y'all. If there are no significant differences in the way people can communicate with each other, and, Slack is an incredibly well established platform (acknowledging that all platforms have their faults), are there any specifically important reasons to even consider a change to a new messaging platform?

[ghost]

e.g., if Slack can only support a finite amount of users, if we're hitting a threshold towards paid service, then we can look into covering those costs, without moving the entire community to a "new" platform.

[ghost]

THIS:

---

Bryan "Beege" Berry @TheBeege Dec 20 2016 15:20 "all of the built-in integrations appear to be more dev focused than chat focused"

---

Though there is something to be said about having a dev focused messaging platform, which does familiarize group members with a dev-oriented tool, if the goal of the messaging platform is to merely promote more communication between members of the groups outside of the meetings, then Slack provides that. A more dev oriented platform can leave much to be desired when on-boarding new members who are beginners while having the UX be even a little too.. devvy? haha

[armaneous]

@easyspirit We're coming up on 1100 Slack members now and, on the free tier, that means we hit up against the 10,000 archived messages limit very easily. The 10,000 count includes all the channels and all the private messages happening. We're now getting to the point where you can't see messages more than 2 weeks old (and this is only going to get worse). We get some really helpful conversations happening, but searching for those again is almost pointless past about 12 days.

We considered the paid route for a whole 5 seconds or so. The problem is the pricing structure is $7/user/month - $7700/mo for us (and growing) wouldn't be very feasible. Slack isn't designed for this sort of community, it's meant for organizations or small teams and does really well at that. From what I've read online when we were looking for alternatives, Slack actually degrades in performance when you approach thousands of members.

Mattermost is the most Slack-like and almost indistinguishable from Slack, it's just a self-hosted solution which means we'll likely spend a fraction of what we would for using Slack.

[ghost]

Ah, gotcha'. And, likewise, Gitter is $5/user/month for private channel usage... That's quite the quandary. :)

What about moving the separate Channels to their own separate Teams? That may alleviate most, if not all, of the current message limit issues. It'll make the navigation between conversations a little different (hitting the team buttons on the left menu bar), but it'll do the job.

[armaneous]

@easyspirit Many of us subscribe to more than one Slack community (I'm at 5, personally, though I know some here are at over 20). Breaking up the channels into their own Slack communities is a lot of clutter for people in that situation. Also, there is a (small-ish) barrier to joining another Slack community, the whole registration process (submitting/approving emails).