Introducing Kikuchat: The Self-hosted Team Chat Solution | Hacker News Reader

Introducing Kikuchat: The Self-hosted Team Chat Solution

kikuchat.com

18 pointsvlucas22 comments

It makes me sad that people are reinventing the IRC and XMPP wheel, and doing it poorly. Want your own version of this, with a bit less effort, but free?

1. Install an XMPP server

  $ sudo apt-get install prosody

2. Install a web frontent

  There are dozens. Here's one for example: http://amiadogroup.github.com/candy/

3. Congratulations, you just saved yourself $199.

Don't be sad. Competition and innovation are good things. XMPP and IRC aren't perfect solutions. I haven't seen this app in action so I can't comment on the "doing it poorly" part, but can you elaborate on that? What is poor about the implementation?

Ummm... I just tried the 'candy' demo - there's no options for file sharing from the candy interface. Maybe there are in others, but I'd have to go test/try out there - this one says it does it already.

You could easily write a plugin which enables file sharing.

Interesting, are there any user limits on the purchased version?

Also, do you have any data on hardware requirements? The software requirements are only PHP and MySQL so I'm guessing it's not doing anything like in-memory caching and it'll need to hit MySQL quite a bit to generate the chat. How does it perform under 100 users? Or 1000? or more?

There are no software limits on users. It all depends on what your hardware and connection can handle.

The Groupon dev team helped us beta-test it a bit, and it handles 100 users well with decent hardware on Apache (they were using an EC2 instance, though I am not sure what size). You could probably go up to around 300 or so without issues on Nginx and PHP-FPM, but I'm not sure you would be able to get to 1000 without adding memcache or load balancing.

We are looking into making a pure node.js version with websockets if there is enough interest, though the price point for something like that would likely be higher. We are happy to provide support and additional feature development (like memcache integration) as needed. Just send us an email if you require anything custom, and we can work out the details.

I'm really interested in this kind of "view source" business model. Is it working for you?

There's a strong case for self-hosted apps, particularly in enterprise and school environments, and it feels like open source shouldn't be the only model for that (it's not easy to run a business based on open source software). I'm glad to see that people are trying different approaches.

Best of luck!

The jury's still out on whether or not it will work for us - we just launched today! :)

This software sales model isn't new, but at the same time it is not widely used. Most people either go for hosted offerings or encrypt the source code. The viewable source code model is the same one that vBulletin and ActiveCollab use, for instance.

I would love to see some kind of live demo.

Why would one want to spend $200 on this instead of using IRC or Jabber?

Same reason people use Campfire and Hipchat?

Er.. and what is that? Campfire and Hipchat both take 5 minutes to set up and are cheap enough that they save you money by not having to set up your own system.

Yours requires setup and costs a ton. How is it like Campfire/Hipchat?

I guess time will tell if it was a good move or not. You don't have to be so cynical. ActiveCollab did this for Basecamp and it panned out pretty well for them.

I'd like to point out that this doesn't answer my question. I mean, you're not selling Campfire or Hipchat are you?

I think you're vastly underestimating the market for this software.

If it's even almost as good as Campfire or Hipchat, people will buy it because most companies don't want their employees communicating via third party services where shit is archived.

I've been looking for something like this for a while now. I won't use Campfire or Hipchat due to the reasons outlined above (and I don't trust 37s or Hipchat to know how to secure their customer data), and I don't want to use Hipchat's shitty AIR apps.

TL;DR: This product deserves to exist. :)

100% agreement. I think there is a huge market for this, I hate Air apps, and the main thing that bothers me is manually managing users in Campfire. Haven't dug in enough yet, but if Kiku has some LDAP/AD integration... specifically CAS integration, then I'm interested.

It is like them in terms of features and functionality. A self-hosted option is appealing to companies that want something like Campfire/Hipchat, but want more control/security over their own data (they can host themselves behind their own VPN/firewall, etc).

Had me until I read "PHP".

Unfortunately, PHP is still the easiest thing for most people to setup and use, so that's what we went with for a self-hosted option. We are strongly considering a node.js version that would allow a lot more users. Is that something you might be interested in?

Shouldn't be 'unfortunate' - PHP is a perfectly fine for this sort of thing. You can address the 'ultrascalable/node/nginx/caching/i've-got-2-million-users-on-one-server-running-on-my-ipod' type of user demands after you're profitable with this first version.

The 'self-hosting' thing shouldn't even be just about saving money - there's a strong case to be made for security/privacy by self-hosting this sort of software internally.

I say "unfortunately" not because of PHP, but because other languages like Python and Ruby are still not as widely used. PHP's "upload-and-go" approach is a big reason for its wide success.

Cool :)

Reply on news.ycombinator.com