The integrations are indeed a gimmick. They consume so much space that they end up being shoved into their own channel, where nobody ever looks... so what was the point of them again?
Completely disagree. There's a tons and tons of ways to use them.
Sure it can be gimmick or it can be an extremely valuable tool and usually they are somewhere in-between.
In my own experience I've found they are a great replacement for many different types of email notifications such are exceptions from an application. Things you want to glance at and maybe take action on. When these are an email I have to "manage" them, choose where to put them or to delete them or whatever. And I eventually end up just filtering them into an inbox I rarely look at.
As a channel though it's just the right amount of attention and time to them. Took some time adjusting settings to get the right level of exceptions to show up and we still have them all in the exception tracker itself.
The other thing is I only check my email a couple times a day. because most of what goes into my inbox is not time sensitive. Whereas Slack is always on if I'm working so I will instantly see if there's a major issue on the site.
Of course every company is different and has different needs. But writing off all integrations as a gimmick is narrow minded.
It's a gimmick based on why you came to use the tool. My understanding is that it's goal is for team communication and all these alerts are simply distracting.
But maybe slack should position itself as an advanced alert management tool (which is what many seem to use it for).
I agree that some integrations are a gimmick, but having content relevant integrations has helped out our team immensely. We'll have channels specific to a project, so notifications that JIRA stories have been resolved, knowing if a build fails, exception logs from production, are all intertwined with us discussing the project throughout the day so they get exposure.
One thing that we've loved is being able to log information into other tools when it is top of mind. Discussing a recent bug that you noticed "/jirio create bug Username can contain spaces" and it's logged in JIRA to be dealt with accordingly.
The potential for custom integrations is incredible but obviously keeping the noise level down is key.
I've asked that question myself - particularly considering how expensive slack is. Some differences include:
1. Secure usernames integrated properly into the protocol, instead of relying on nickserv and configuring your client to send a dm when you connect which is far from a simple/intuitive system. This includes options for google/corporate single-sign-on and 2-factor auth.
2. Infinite searchable scrollback which keeps position properly across multiple devices. As an experienced IRC user I can achieve something similar using ssh+screen+irssi - but it's hard to use even for advanced users, and I can't imagine trying to use it from my phone.
3. Offline messaging that doesn't rely on the user knowing the right magic commands to activate the bot. Bob is offline - do I need to !tell bob whatever or !ask bob whatever? Can I DM it to the bot to avoid spamming the channel? What's the help command? Can I use that over DM? Is there even a bot in this channel?
4. Integrations are exceptionally easy to write. You can post to a channel with a single curl command. Obviously writing IRC bots is possible, but it's a lot more complicated.
For me personally, even these differences taken together don't seem worth the expense of slack. But I would understand if other people see it differently, especially if they're not that experienced with IRC.
1. SASL auth – some servers even support Oauth via SASL, or certificate login.
2. That’s what Quassel and IRCCloud provide as a one-click solution.
3. With everyone using Quassel or IRCCloud, this also becomes easy.
4. There are many services that provide webhook -> IRC services ;)
As someone who has been using Quassel, where everyone else uses Quassel, and who uses SASL auth, all the issues you mentioned stopped existing long ago.
Add the expense of Slack, and one seriously wonders if it’s just discoverability.
And packaging. Just picking on one of the ones you've suggested: Quassel is an app, so I can't easily try it out. I certainly can't invite everyone in my company to use it by clicking a single link in an e-mail. It doesn't have an iPhone app, I think those are quite popular. And so on.
Seriously, all these things are pretty easy, as evidenced by the fact they've been working since the 80's in the form of IRC.
The value appears to be in packaging them up as a modern web app (and having a lot of success in tech-industry PR which is where Slack has excelled beyond e.g. HipChat. But even that is partly down to the good packaging of the onboarding process).
Interestingly, that's exactly what I personally am working on: mobile clients for Quassel, easy deployment (single docket container for everything), quassel-as-a-service for people to just invite others, etc.
A web app also exists, but the discoveravility is an issue.
Packing the things up properly, marketing them well, etc is easier, as we also have the advantage of being open source, which helps in open source communities.
