HipChat Replacement?

[ylluminate]

Have you considered emulating the HipChat API so as to allow HipChat clients to use Kandan server? Further, have you considered allowing importation of HipChat server data into the system to pre-populate it with HipChat data for would-be converts?

[scouttyg]

This would be interesting, the only thing I might worry about is that if we get too "close" to HipChat in structure that there is some sort of infringement, but at best this would probably fall under reverse engineering which should be ok.

[jrgifford]

It would probably be like any of the varied twitter-compatible services that there were for a period of time, where you just point it to a diff endpoint.*

I am not a lawyer.

[ylluminate]

Well, they give the data out so I don't really see this being a super serious issue. Further there are plenty of other use cases where we have servers and clients that emulate protocols for communicative purposes. Ie, Red5 for an Adobe Media Server replacement, etc.

Frankly, the following preference option seems to be their indirect outreach for someone to do this by opening up the server field:

Ultimately HipChat itself needs to remain agile and offer more of a reason to keep its customers vs just beating down competition. I personally love HipChat, but their policy change last year where they now have unlimited free accounts with limitations is a problem. If they allowed users to have a few paid for pro accounts while allowing the rest of the other users to be free, that would be great. They ultimately will probably do this, but there is no reason to not have competition that fills voids, especially when they have opened up their client to connect to other servers.

It might be wise to make this an option for Kandan such that it has to be enabled and isn't right there out of the box, but frankly it is a killer feature. If it existed, I'd switch over in a heartbeat as a stopgap to solve immediate issues.

This has the added benefit of allowing importation of HipChat data for would-be converts. Further it gives Kandan time to beef up the client side solutions in such a way that it immediately provides what is needed and can then just simply grow out and become a superior product in terms of UI/UX. Personally I could give a number of feedback suggestions to improve upon the experience that HipChat currently has and figuring out new features and functionality would be no harder than strolling through their feedback area.

[scouttyg]

It might be wise to make this an option for Kandan such that it has to be enabled and isn't right there out of the box, but frankly it is a killer feature. If it existed, I'd switch over in a heartbeat as a stopgap to solve immediate issues.

Kandan is a great open source chat app. Period. But my guess would be, if you switched over if we implemented this feature, you'd find we still lack many of the features people have come to expect in a decent chat application. So you might solve one problem, and create five more by switching.

Right now, I can think of about 15-20 bugs / fixes / additions we need to do so we can be a complete, functional chat client. Once those are in, we'll probably upgrade our tech to Rails 4.2 and Ruby 2.2 so that way we can not only be a feature complete app, but a stable one as well.

After that, this is where in my mind we'd start to venture into features like this, but even then I get slightly nervous. Even outside of legality, do we want to be a drop in replacement for HipChat? My biggest problem with dokku was that it was similar enough to heroku, but I'd keep hitting edge cases with features that worked perfectly on heroku but not on dokku, which led to me giving up. Will we attract more people to the client by emulating HipChat, or will we drive people away when we are not feature for feature compatible with them?

[XVilka]

Related to #302 (if you decide to implement HipChat API).