Hacker News new | past | comments | ask | show | jobs | submit login
Show HN: Hosted IRC for Teams (my first startup) (teamrelaychat.nl)
74 points by pepijndevos on April 21, 2013 | hide | past | favorite | 69 comments



There is an opportunity for IRC and XMPP protocols to co-exist in the same group chat and solve customer needs better.

One of our customers is a YC company, that moved their group chat from a IRC based system to our XMPP based group chat product[1], because IRC did not work well for them. After they started using us, we figured that some people still preferred IRC and some wanted the ease of use of XMPP. The solution was a simple bot that will bridge the messages between the IRC channel and the group chat. This is now working out very well for them.

They now have benefits of both IRC and XMPP on their website. I'll be happy to talk more details and possibly collaborate. My email in profile.

[1] - https://groups.gaglers.com/


You can also have an irc interface to xmpp group chats using bitlbee.


Perhaps there is something to learn from grove.io (I'm not saying it's a bad idea, only that you should learn more about what didn't work for them -- which you presumably already have)

https://grove.io/blog/grove-shutting-down-october-13


Encourage you to read the more recent blog posts at https://grove.io/blog/ as well. My company acquired Grove and are continuing to run it profitably I might add.


That's great! I just found Grove in this comments thread and it does look very promising. Best of luck with the service.


Pownce, Convore, Grove, all shared similar fates, I think its more indicative of the founder not the business.


Personal opinions of said founder aside, all three were bringing arguably trivial products to market where there already was a large presence of two or more players dominating (Twitter/Facebook/status.net for Pownce, HipChat/Campfire for Grove) and with a lack of game-changing innovation to separate them.

Makes less sense to blame the people, rather than lay blame to what they did (or did not) do.


> bringing arguably trivial products to market ... with a lack of game-changing innovation to separate them.

It sounds to me like you are judging the founder. Not that I would disagree with you.

Don't get me wrong, I think Leah is awesome. She just builds up companies for soft landings/aquihires and pockets some money along the way. There is no reason a viable company couldn't exist on the grove model.


I did enjoy Pownce back in the day. It worked and I loved the ease of connecting files with certain people. But I think my bias was greatly attributed to adoring their funky layout design: http://www.bomega.com/wp-content/uploads/2007/07/pownce.png


It didn't take over the world, and I don't expect to either.

Some people go live on a farm and herd sheep, I herd servers.

Another thing I took away from that is that closed source software dies with their author. Open source lives forever.


I need it to be dead simple for people to join the chat, using existing chat clients. i.e. whatever they have installed, I can't ask them to install something new.

What I mean is this: A lot of people I want to invite into chat, it's an experiment for them and they aren't really committed.

Oh, I have to install IRC software? Sorry, too much work.

Oh, I can't access it from AIM/gchat/etc? Sorry, too much work.

These are the people I want to onboard, so they can see the value in participating in my chat room.

What is the tool that allows me to attract anyone to the chat room, without them installing new software?


Would the best solution be a web-based client? I have XChat on all my Windows machines and use Colloquy on my Macbook so I never worry about IRC support. But it's obvious that not everybody uses IRC on a daily basis.

I would wonder if it's possible to support people using their own desktop client, while others alternatively use a webpage on the site which connects into the same IRC channel.


> What is the tool that allows me to attract anyone to the chat room, without them installing new software?

I think web based, non IRC, solutions have that covered really well.


I don't want IRC. I want HipChat (inline images, ability to directly upload & share media, persistence, good searching, integrated video, HTTP API) but with non-shit native clients. IRC is a fucking dinosaur, I feel something like this doesn't have the capability to really improve the situation.


1. Many people like the minimalism of IRC.

2. Being a standard open protocol makes IRC better than any alternative anyway. Can you use Hipchat on a graphing calculator? You can use IRC on one.

Though I love IRC, I do wish it has some features present in things like Hipchat.


I love companies that try to make modern use of IRC and monetize that :-) It's a dinosaur but it works and somehow still has a lot of users. Chicken egg problem? Network effect? Nerd 1337ness?

shameless plug: We're offering hosting for Quassel(cores), a distributed IRC client that enables access to the same user/backlog from multiple different devices: http://woboq.com/quassel.html


IRC is just a protocol, just like email, there's nothing precluding better clients. In fact Grove.io ("Hosted IRC and so much more.") offers many of the features you listed.

Whether you market it as "IRC++" (like Grove), or "HipChat with IRC support", is another question, but I'd love a service that had a great IRC (and maybe XMPP) interface, plus all of those features.


There is nothing about IRC that prevents it from providing those features, and many of them already exist in various clients. Ircle had video support in the 90s, Linkinus supports embedded media, and persistence is simply a matter of using a ZNC proxy (IRCcloud has it built in). P2P file transfers have been in IRC for decades, but it would be a piece of cake to integrate a media host into an IRC client.

The CTCP portion of IRC makes it possible to do just about anything you want, as long as both parties are using clients that support the action.


Fair point.

TRC offers persistence via ZNC. Some of the other points are offered by some clients.

File sharing is a work in progress for TRC, but a very important feature to me.


Don't see this as me putting you down. I just wish people started working on the next generation of group chat (which is a HUGE field) rather than just piling more of the same on top of old stuff.


Developer of the kiwiirc.com IRC client here (Unaffiliated with teamrelaychat.nl). While I agree that there are far better protocols than IRC for todays modern world, IRC still has a long way to go and is currently being extended to push it further in that direction. It still has millions of active users today so it's clear it has done something right.

Inline media, file uploading/sharing, mobile messaging, persistence, log searching, APIs, etc, etc can all be achieved with IRC and is actively being developed with the kiwiirc.com client that this startup is using giving it a great start.


Agree - for me, the persistence is the most important, something which IRC doesn't offer. I'm not always on, and in a business setting it's crucial to be available for messages at all times, even if I'm going to actually read the messages later. I can also find out what the others have been discussing.


Check out IRCCloud, they keep a client connected on your behalf 24/7 then let you connect to it when you open the web app.


TRC runs ZNC, which will save and replay messages you have missed.


Have you seen http://trash.io

http://trash.io/room/whatever_you_want to make a new room

realtime, drag and drop images, youtube, vimeo and multiplayer html5 games.

ws, redis, on top of a rails app on heroku.

enjoy.


The linux client that they released recently is decent


Why would I pay for IRC?


I know right? Am I missing something here? What does this server offer extra that would explain the hefty $20p/mo pricing? I can just open a free private IRC channel on any server and use that?


Some measure of privacy & control.

Of course, if you care enough about that you shouldn't be using any exterior service.


Let's say you have a developer whose time is worth $120/hr to you. If that developer spends two hours futzing with configuring an irc channel in a year and no time futzing with this service then it has paid for itself without needing to offer anything extra. It seems like they're attempting some extra features to sweeten the deal anyway.


Describe "on any server". Like a $20/m AWS micro?


On any IRC server. There's absolutely no reason to host your own.


This looks cool. Nice job.

I like the IRC style of chatrooms as it's familiar to me, but as others have already stated, I would prefer a service that provides additional features (ala HipChat or Campfire) such as file sharing, searchable history, and allows attendees to jump on a video call. Bonus points if you allow users to join over the phone as well.

We use Google Hangouts for video conferencing with my team of 3 (all remote), but we have to keep a campfire session open for sharing files and links relevant to the discussion. Even with this set up I'm missing a way to join our ad-hoc conference calls over the phone when I'm away from my computer.


Hey, this looks like it would go well with what I'm currently developing: http://www.getinstabot.com/. I'll give it a shot, congrats!


That looks great. How does it compare to, say Hubot? That's what is running on TRC.


Well, the idea is that it will provide more functionality without the hassle of hosting it yourself. As to the specific commands, Hubot has more, at the moment, but Instabot's list is growing.


Perhaps creating a thin layer of middleware so that your work supports hubot scripts is a good idea, then!


Yep, but hubot scripts are already very short and easy to implement, so it shouldn't be much trouble either way!


It's a bad idea to use "I'll" and "Forever" in the same 6" radius of each other.


"I'll publish the deploy scripts and help you migrate."

This might actually be useful sooner rather than later.


How does it compare to grove for example?


Good question. They are obviously similar in that they are both IRC and more. The key is in the 'more' part.

One thing I'm working on is file sharing. This already works in the web client. Desktop is going to be tricky, involving DCC probably.

Currently my search/history story is not so strong, but that's also being worked on: https://github.com/znc/znc/pull/325


I'm in here and on the demo server to answer questions and take feedback.

Word of advice: don't all pick demo1.


If you control the IRC daemon is there any reason we can't have full server-side logging of all channels and 1:1 interaction in a similar way to HipChat, et al?


Congratulations on the launch. We never met, but I recognized your name from the PUN mailing list/meetup.


Congrats! It’d be cool if I could collect uploaded files and logs into S3.


You can! Well, sortof.

On the web client you can upload files to S3. Logs are stored on local disk though.

Web interface for logs is being worked on: https://github.com/znc/znc/pull/325


Why is it soo expensive? Otherwise relatively good idea.


I find it rather annoying that someone on HN will complain about the price of a service no matter how low it is. $20 per month is absolutely nothing. Most of the other team chat services cost more, and for good reason. Businesses are more than willing to pay a lot of money ($20 is not a lot) for software. Broke college students are not a good target demographic.


Before someone attempts to nitpick, 20 euros per month is nothing as well.


Because you get a full VPS, maintained by me.

At 10 people, it's already cheaper than other services, which usually charge a few bucks per user.


> Because you get a full VPS, maintained by me.

This is bad. You're throwing resources away and increasing your management overhead. It would make more sense to have some sort of bouncer-type proxy to determine the source then route through to a dedicated ircd instance on a shared box (e.g. x ircd instances, one per customer).


Your pricing is what failed it for me. I can get a 512MB VPS from DigitalOcean for $5/month.

I love this idea, though. I just feel that your approach of giving a full VPS is a bad idea. People who want a full VPS will buy a full VPS for much cheaper than the 20GBP you're offering it for.

What you need to do is multi-tenant your hosts, and isolate them from each other. Then you'll get costs savings that you can pass on to your customer; which will put you in line competitively with the likes of HipChat and Campfire.

Great idea, can't wait to see what comes of it!


The cheapest Campfire plan is $12 per month and the second cheapest is $24. HipChat is $2 per user per month. 20 euros per month with no user limit is entirely in line with this pricing.


Can I search chat history with this tool?


What are the advantages of this over say a hosted jabber or lync system?

also what authentication options are available I can't seem to find a mention on the site


Personal projects are not startups. I don't know where this falls in, but it bugs me when I see something where someone maybe spent a couple of weeks on it and call it a startup. Have you talked to investors? Have you quit your day job? Are you seeing startup like weekly growth?


"Have you talked to investors?"

A great many companies are "bootstrapped", which means the founders kick their own money in initially (or scrounge from friends and family). Many big companies, a recent example being GitHub, were started this way.

"Have you quit your day job?"

Many people wait until they see some traction before quitting. Doesn't make it any less of a startup. Consider it a consulting gig while launching the business (unless you don't consider it a startup if people consult on the side)

"Are you seeing startup like weekly growth?"

You are falling into this paradigm trap. Not all companies see hockey stick growth initially nor do they aim to. Most sustainable businesses I can think of started slow and gradually grew up. Nothing wrong with that.


Personal projects are not startups.

According to "The World Defined by Btipling". Most of us haven't read that book and don't care.

Have you talked to investors?

Who cares? Maybe he's self-funding it. Having outside investors isn't an absolute requirement to be a startup.

Have you quit your day job?

Who cares? Maybe he's bootstrapping the startup in his spare time while still working a dayjob. Maybe not. What f%!#ng difference does it make?

Are you seeing startup like weekly growth?

There is no such thing as "startup like weekly growth". Not all startups are going to fit the exact same model in terms of revenue growth, user growth, or whatever other metric you care to pick. And as far as that goes, he may be pre-revenue altogether. But none of that changes whether or not his thing is a "startup".


I'm pretty serious about this.


If I may ask, why? I don't personally think this is going to be successful, what with Hipchat and Campfire in the same space (which are very very popular) and grove.io having a good name for it and competing directly against you.

Edit: I think you also misunderstand the value proposition for your customers: you mentioned above that you should buy because you get "a full VPS, maintained by me", which is very odd.


He could win if he proves to be better at marketing. Outside of the HN / SV Startup World "echo chamber" most people in most companies in most of America (much less the world) have never heard of Hipchat, Campfire, Grove, Convore, or this guy.

You could argue that he isn't competing against Hipchat, etc., but competing to convert current non-users of this kind of technology.

If they come out targeting other startups and technology companies, then I'd agree the odds are long. But come out targeting textiles manufacturers or transformer manufacturers, or shipping / trucking / railroad companies, or any of a bazillion other industries, and they may just have a shot at establishing a beach-head. Especially if they are willing to provide some vertical specific aspects to specifically appeal to customers in specific niches.


He could, but based on his value prop above, I think there is much to learn before that's going to be possible. I'm not saying he can't, of course, rather that its going to be an uphill struggle that he might not get if he built a different product.


True. I don't mean to suggest that it's going to be easy.


I keep having to advise startups "it sounds like it might work, but that's going to be really hard. Maybe you should do something else".


There is also Flowdock[0], which wants to compete pretty aggressively here (disclaimer, I work for the company that bought Flowdock)

[0] http://www.flowdock.com


Yes indeed. I have a FlowDock account and my company (https://circleci.com) integrates with you guys.

I'm curious though: how big are you guys? I only mentioned HipChat and Campfire as they appear to be pervasive, and while I've been hearing more and more people mention Flowdock, it doesn't seem like you're an 800 pound gorilla that he should be afraid to compete against (yet!).


I don't know how much I can really say to be honest, but I do think Flowdock will compete more in the future.


Cool, looking forward to seeing it!




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: