It's all emacslisp. Including the webserver (it's elnode). So there is an irc bouncer written in elisp tightly integrated with the elnode webapp. The bouncer uses a modified rcirc (built into emacs) to connect to the irc servers.
The robot framework is also written in emacslisp.
The irc servers are ngircd and the webserver is nginx.
Congratulations, Nic, on what is likely the world's first Emacs-hosted startup! For those who don't know, teamchat is implemented in Emacs Lisp using Nic's elnode (async web server) and shoesoff (IRC bouncer) libraries, among others.
it's simple to build a simple one... but the more polished takes more time. Why not just offload that work to someone else? That's our view anyway. We've chatted to some potential clients and other companies running their operations this way.
Minor point: It would be nice to be able to click those screenshots and get bigger versions. It's currently very difficult to see what the application looks like.
Sent out the link to several co-workers (since we are looking for this sort of tool right now) and then we all get the "We'll let you know when we go live" email.
If you ping me on twitter I'll make sure you're prioritized. We are obviously wanted teams to try this to tell us what is good and what isn't. So if you're a team we want you basically.
I would be really excited to try this. Is it live yet? I always think it's better to save Show HN posts for products that we can actually play with. Looking forward to seeing this in any case.
Can you publish your numbers afterward? Or just whether you thought it was worth it? (Or something along those lines)
I have seen this strategy recommended to gauge interest, but have always been wary of doing it myself (for exactly this reason: people being like "wtf? this isn't real?").
It is probably one of those things that should be done, even if it makes you uncomfortable.
That's basically it. We've worked on this for quite a while and I reckon we could just keep going... This way we can gauge interest and prioritize teams who sign up.
And yes. I will publish results later in the week.
I'm not going to advertize it that much (because I don't want to support people that way yet) but all the code behind teamchat.net is freely available on github.
Yes. We're veterans of previous video startups. This is video meeting done right. It's not a big part of the product but it is important coz it means you don't need skype or hangout or whatever in addition to this.
43 comments
[ 53.0 ms ] story [ 261 ms ] threadThe robot framework is also written in emacslisp.
The irc servers are ngircd and the webserver is nginx.
You guys at #emacs have helped this happen. An awesome community.
That says there might be a need for it, but on the downside it's really simple to build, so competition can spring up very quickly.
Sent out the link to several co-workers (since we are looking for this sort of tool right now) and then we all get the "We'll let you know when we go live" email.
So thanks for making me look like an idiot.
The trouble is this is for teams, you kinda have to have a team. We wanted to gauge interest so we've done it this way.
If we can we'll go live for collections of people (if there are 2 or 3 people from the same company email domain) this week or next.
I had three guys from my team sign up so we could test it out, and only after signing up did we find out that it's not live yet.
I have seen this strategy recommended to gauge interest, but have always been wary of doing it myself (for exactly this reason: people being like "wtf? this isn't real?").
It is probably one of those things that should be done, even if it makes you uncomfortable.
And yes. I will publish results later in the week.
Looking forward to seeing the product!
Mobile-friendly IRC web client, private rooms, and bots with commands/webhooks.
https://grove.io
I'm guessing WebRTC would be the way to go... once it is released at least