Is it really a bad "beta", or has your preconceived notion on what a "beta" should be been radically altered by the sheer number of startups that throw the word around now?
Gmail was in "beta" for about a million years, and was incredibly stable for most of that. I think it has become the allegorical metric these days. But I remember back in the late 90's and early 2000's when making the choice to use beta software was a decision not to be taken lightly because you probably would find tons of very easily reproducible, obvious, critical bugs.
Don't forget that before "beta" was widely used as an adjective, it was a noun and it was immediately followed by the word "test".
I'd agree in principle, but did the email announcement might have been a bit more toned down if that was what they were thinking. I'm pretty sure they wanted some buzz.
I used it for five minutes and found: There is no way to search for a person in the Lobby and you can't reorder the tabs. These are two things I do a lot.
Interestingly the first thing I did was re-order the tabs. Now I went back to check and I can't re-order them. Maybe that info helps the Hipchat devs out - I just trashed prefs and then you can re-order one tab. Then it stops allowing you to do this.
The tab thing is a bug, but it's unclear from jedberg's wording. The actual problem is that you can reorder the tabs, but only once, after which reordering stops working.
Slow, crashed within minutes of using it. It's not very pretty either. I'll stick with Adium/Jabber hooks but was really hoping for something good here.
HipChat uses XMPP, so you can already use it with Adium/Pidgin without much trouble. The only things that are really missing by default are the syntactic sugar for @-mentioning people, HipChat's custom emoticons, and inlining of image/link previews (many of which you can fix with already-existing plugins).
You don't need to identify as "bot", just send a keepalive message (can be a single space, XMPP ping, whatever) every 60 seconds or so. You can also request no history when joining rooms (most bot frameworks do this). It's part of the XMPP spec as well. Our roster pushes are also as in the spec, or have you found an oddity we should correct?
Joining a room with your full name can be an unexpected requirement but it really aligns with the way we enforce names on our service.
I've switched from using HipChat on the web (because I didn't like the AIR version's scrolling behaviour) to the native Mac client beta and I'm really happy with it so far.
I'm on the most recent OSX version in case this might make a difference. No bugs or crashes for me so far.
There's a difference between a new beta and an app that has been in closed beta for months already and only today has been made available to the public. Just because it's 'beta' doesn't mean there should be obvious trivial bugs present.
Just to run off a few issues we found within literally minutes of using the app:
1. Resizing window to a small size then making it large again causes the entire chat view to become very narrow.
2. Messages sent by one user, often get displayed next to another person's username (so it looks like someone else wrote the message).
3. Notification popups appears even when the app is actively being viewed.
4. If you go away from keyboard, the app sets you as idle, but when you return, the idle indicator doesn't get removed
++ on the notifications. my biggest beef with hipchat is it will send emails of conversations as i am having the conversations. email notifications when I am away is nice, kind of annoying when you are actively typing in the window.
Maybe I am being naive(going to assume I am), and i admittedly have never used HipChat. But could someone explain to me what this offers over say setting up your own XMPP or IRC server and using that?
The cost to setting up and maintaining your own server and emulating many of the more modern HipChat features (drag+drop file sharing, archived/searchable history, mobile push notifications) is not trivial. Most companies find it far cheaper to spend money on a service than roll their own in house (see also: GitHub, UserVoice, hosted WordPress, etc).
Also, nontechnical people are often not very comfortable using IRC/XMPP.
Your account manager is an exception, and your argument is a false equivalency: just because you know of a non-technical person who has experience with IRC does not mean that all nontechnical people have experience with IRC. Indeed, the trend is most likely that people in non-technical roles have less experience and exposure to such tools.
HipChat is more than an IRC server - it's a fully searchable web-based chat archival system, as well as a file transmission and sharing app backed by S3. It's also fully mobile-compatible and will do proper message routing to multiple devices owned by the same user, as well as temporary message caching/retry if your device is unreachable when a message is sent. etc etc. The list goes on.
None of it impossible to replicate in-house, but I dare a non-technical person to replicate HipChat's stack.
Google doesn't exactly have an "IRC on crack" product though? Also, I find GTalk's routing to be spotty when it comes to multiple devices - it's quite common for me to receive messages from coworkers at night and have them route to my work machine instead of one that's active on the GTalk account at home.
Nowadays we use GroupMe for things that must be read in a timely manner. It'd be nice if we could use GTalk for it (this is a Google Apps GTalk account, not a public one).
It makes everyone feel they're a hip startup. I mean, my IRC client has been doing drag and drop file sharing and image URL inline thumbnailing for... 10 years. The HipChat client is nice, but let's be real, it's just a nice client for a handicapped implementation of XMPP. I'm forced to use HipChat at work and use Adium to connect. I've even got the emoticon set, and message styles to match.
They can't even program a mobile app that reconnects me automatically (unless that's changed recently).
Not having to maintain your own server. Slightly better integration with other things (e.g. messages sent when you're offline can be emailed to you) - nothing you couldn't do with a bunch of client plugins, but it's there and set up for you already. A very nicely polished in-browser client, nicer even than irccloud, which makes it much easier for non-technical folks to use.
There are lots of open IRC servers on which you can create password protected rooms. That plus Google Docs /Drive works great for small companies. For larger companies, however, I can see why something like Hipchat could be useful.
I've never really had a thorough introduction to IRC. Is it correct to assume that it's 100% ephemeral, that all conversations, questions, answers are gone when you log out?
If that's the case, I'm puzzled as to why it would maintain so much traction when the record of communication can be so valuable (e.g. people searching the log vs asking a question that has been answered many times... )
It's common for people to log channels - either directly via the server or via a bot that sits in the room and just records everything. Logs can be exposed via various web frontends, most of which are pretty limited feature-wise.
Most IRC clients do not log out of the box, so direct messages from other users are easily lost unless you go out of your way to make sure - the server does not do this logging for you.
The Quassel IRC client is actually split into two pieces.
The server module which actually connects to, and interacts with, the IRC server.
The client portion which connects to, and interacts with, the quassel server.
Most people just use "Quassel" as a simple desktop app in which case both modules are merged and it acts just like a normal IRC client.
However you can also install the server module on a server somewhere in which case it stays connected to the IRC server, buffering chats etc. You can then connect to it with the client part of quassel which is installed on your local desktop, which you close it all you are doing is closing the client module.
I really like that design, a good separation of concerns allows it.
I appreciate this and the previous comments, but I gotta say: this sounds like a lot of schlep.
I may be wrong, but I feel like the right persistent chat service can/should become a 21st century usenet/google group, where the conversation is always happening, IFFT-style hooks help you filter/act on info, and history is Gmail-simple.
I'm investigating a group-chat solution right now for an organization of several thousand non-technical members, and my intuition tells me IRC would simply not be adopted by most users.
If you can't setup an IRC or XMPP server, then I don't think you should call yourself a software engineer. If you're a scrappy startup, you sure as hell better be able to setup an IRC server. And no... it doesn't take more than an hour to spin up an EC2 instance and setup inspircd as a private server.
5 minutes of conf file tweaking and you're done. Seriously that's it! I did this about 8 years ago and haven't touched my setup since. There's pretty much zero maintenance to do with an XMPP server.
Do you get file sharing, easy group chat, @mentions across rooms, public temporary invites to customers into internal channels, very easy github and zendesk integration and, the killer feature for me, a searchable archive over all conversations?
I'm sure you can do all of these with ejabberd and some additional external tools, but setting all these components up goes way beyond just apt-get'ing the software.
Now, you might not need some (or all) of the additional features, so ejabberd is the perfect solution for you, but you can't jump to the conclusion that because you don't need a feature, nobody does.
I'm a quite new HipChat customer - we were using skype before, but the indexed archive is a total killer feature for me and already helped me and my coworkers a lot as it helps us to learn from conversations between other people, thus reducing the amount of stuff that has to be asked multiple times.
For me, that's totally worth the $2 per user per month.
What happens when ec2 shuts down the machine with jabber on it? When there's a memory spike in the middle of your vacation and no-one can send messages? Etc, etc, etc
You turn the machine back on? Switch to a server provider that doesn't randomly shut your machines down?
What do you do when hipchat's severs are down? Twiddle your thumbs and wait for it to magically come back? At least if it's your server you can actively fix it instead of waiting for someone else to do it…
The killer feature is that nontechnical Mac and Windows users can click a link emailed from anyone on their team who is already using the app, and get seamless group chat across platforms, without calling the IT department to set up a server and/or open firewall ports, or committing someone on the team to managing anything XMPP-related. It just works.
Even Linux folks can participate with additional work (e.g. dozens of packages to install if you're running Ubuntu 64, since Adobe AIR is 32-bit only; annoying).
That said, I agree that the barrier to entry is pretty low.
Its fairly stable, desktop and web clients, accessible for non-tech members of the company, decent UI, various API hooks and hubot access, easier and faster to get up and running/ maintain then a home rolled XMPP/IRC. If you just need a functional company chat client, its solid.
We switched from Campfire to HipChat about 6 months ago and it's been awesome - apart from the Adobe Air client crashes whenever someone posts a GIF. Hopefully this'll fix it :D
The gif issue was an Adobe AIR bug, so it's definitely gone now. It was incredibly frustrating having them break functionality that many of our users love, especially because it wasn't possible for us to work around the issue. Sorry about that!
We hired someone new, so had to upgrade from Campfire's free plan. HipChat would have been the same price, so we decided to give it a try, and never turned back.
> enjoy crystal clear display with Retina support.
I'd enjoy it a lot more if it were true. Still no Retina support for any of their icons or images. Sure, the text finally looks right, but doesn't Cocoa handle that for you?
Hey - Garret from HipChat here. We're still working on retina assets for some of the app icons, and haven't started on retina versions of the emoticons quite yet. Hang tight!
Garret are there any plans to offer a self-hosted HipChat solution similar to how Jira and Confluence have self-hosted pricing?
My company is limited to our chat products due to HIPAA compliance and we would love to use this product for company chat but host it on our own infrastructure.
Seconded. We can't use public services for internal messaging, but we do have JIRA + Stash inside the firewall. Hipchat server would be good, or is it XMPP compatible?
It's definitely XMPP compatible. We have an outside dev firm that does some work for us and they connected their own chat system via XMPP. It works fine for them as far as I can tell. Their help docs [1] [2] have all the details you need.
All I can surmise from this video is the that old HipChat must have really sucked. I have no idea what this product is or why I should care, because the video and webpage don't explain any of this. Maybe someone can explain what HipChat even is...??
Apologies for hijacking the thread, but I'm wondering:
We're using Campfire right now, with the Propane mac client, and we have a dedicated room for notifications (currently exceptions from Errbit, monitoring alerts from Nagios, CI build failures on Jenkins, app deployments from our own custom tools).
It works pretty well, but we're a team working on a whole bunch of unrelated but overlapping apps, and having everything in a single room is getting to be very messy, and at times the Propane icon is bouncing almost continuously. And Propane doesn't support controlling when to notify me when there are new unread messages, so I have no way of filtering out the noise. Splitting out into multiple rooms is not a workable solution as the "spheres of interest" are not mutually exclusive.
I'm curious if anyone is using Campfire or some other tool in a similar way, or if HipChat could solve the problem (on first glance it looks fairly simplistic). Email doesn't cut it and direct IM through some bot is not really an option. A dedicated notification app/gateway seems like a better fit, but I haven't found one since Notifio folded.
We pipe everything into Campfire, scoped into the relevant room (last time I checked we had about 80 separate rooms).
The topical bit is important- a particular app's notifications get dumped into that app's room, for example. That means I only get notified if I'm in that room. Conversely, that means I don't get excessive notifications if I'm not in the room, for apps I'm not interested in.
"Spheres of interest", as you mention them, become less important as you grow the company. We've found it important to not get too sucked into every little aspect of the company- it's too broad. If someone thinks you should know something that's happening in a room you're not in, they can just ping you and you can hop in and stay up-to-date. That's been a really good solution for us.
And so you have set up Nagios and whatever to always include the appropriate recipients in the messages? Where do you centralize the information about who should be @-mentioned? Edit: Seems like a lot of noise if there are a lot of recipients (some of our apps or libs have 8+ people).
HipChat has the concept of a "notify" flag for every message you send via the API.
Notified messages make a ding sound in chat, flash the task bar, etc.
For my HipChat setup, build logs are silent (and don't disturb anyone), but a notification that a deployment failed has the "notify" flag set to gain everyone's attention.
Sure, I get that. But what if a deployment failed for an app that you are not interested in (because you're not working on it) and don't want notification about it?
You have the notification goto the room for the relavent Team/Project. Then only people who are in those rooms (even when offline!) get those notifications.
As I said, I work on about 8 codebases concurrently (shared across just two or three projects). I can't have 8 chat rooms open, that's just silly. Every time I want to find a window I will have to cycle through them.
"I can't have 8 chat rooms open, that's just silly"
Why is that silly? Why would messages about all 8 projects be going into the same space? Shouldn't they be separated? That's exactly how we filter out all the different messages. And Hipchat has shortcuts to switch between different rooms, and highlights rooms that have new messages. Works pretty well, from my experience.
Do you have your email program open with 8 different mailboxes, one for each project you want to receive communications about? I doubt you do.
Notifications like these are conceptually a queue of stuff that requires your attention. There is no point in separate notifications into multiple queues, they are all equally important. If project A is throwing a fit, it doesn't matter if I'm currently working on project B, it needs my attention.
I don't know Hipchat, but with Campfire and Propane, each chat is designed to be in a separate window/tab. If you bury the room behind a kind of modal selector (as seems to be the case, there's a sidebar with each room?), that's really no better. After all, if I've focused project A and project B starts blinking, I need to switch there. That's an unnecessary switch when all I want is a queue of items to read. (And no, email is way too heavy for this, and I don't want it interleaved with my regular email. We used to use email, those were bad times.)
It's not unusual for all projects to have incidents during a day, whether it be deployment information, exceptions or monitoring alerts.
We dont do that at all, though I'm 90% sure the hipchat API supports that.
What I meant was, when a team member wants to contact me personally, they put @pbiggar in front, and Hipchat will give me a notification for that, but not all the other messages.
What I'd like is to have a list of project names I am interested in, let the bots use the project names in their messages, and then have the client ping me when something interesting is mentioned.
Propane has custom mention support where you can enter a bunch of strings, but it doesn't really work.
Use different rooms (with relavent people in each room), or make your bot smarter (we do both). You could easily make it so a given user could 'subscribe' to a project by sending a message to the bot. i.e. '@companybot subcribe xyz-service'. Then have the bot keep track of those and automatically at-message the interested people upon failures etc. IMO opinion, different rooms are still better, because it brings everyone together when there's an issue also, but it depends how large an organization you have.
We don't have a bot. Every service that does notification uses the native plugin to do so; Errbit has a Campfire plugin, Jenkins has one, etc.
While we could possibly write a custom bot that centralized everything and tied all the services together, that is going to be a fair amount of work for something that ought to exist already.
Hipchat is incredible. Simple, but for our company - it basically is the MAIN form of communication internally. It keeps our team connected in real time, around the world, in the office and on the go. It's a rare day when I get an email from anyone on our team because we're in constant communication via Hipchat all day, every day. Very excited about this native version - much faster.
Immediately noticed one short coming: I can't resize the font any more. In the AIR version Cmd/Ctrl+Plus/Minus just worked (perhaps as a result of the AIR implementation, not a product decision), but in this native version it doesn't and there are no preferences for it, either.
If the HipChat guys see this, please consider it as feature for an update soon?
pay a monthly fee for group chat;
happens on someone else'e servers;
for the love of all that is holy, this exists already doesn't it?? and for free??
It seems "web 3.0" is all about making tarted up existing free software with hipster interfaces, selling it for $$$, and putting it "on the cloud" (i.e. on someone else's machines).
I know this is the way the world seems to be moving (backwards, back to the 1970s if you ask me, to mainframes and dumb terminals (chrome book anyone?)) ...
but I for one would rather see a world where local users controlled their own machines, their own data, their own encryption, and a peer-to-peer system for communicating
Hipchat for 20 users is $40/month. If a single programmer on that team making $40/hr or more spends an hour setting up a "free" solution (an irc server, for example), that team loses money. Getting a free solution set up with all of hipchat's niceties across multiple platforms (and a web interface) will usually take longer than that.
And what about teams that aren't technical? Are you saying there still isn't a need for this type of software?
Besides, HipChat gives you a lot of nice features that you'd have to spend a lot of time trying to do yourself. Paying a small amount of money a month so that my whole team can chat far outweighs the pain of rolling my own solution and making sure everyone can use it.
PS I wholly reject the argument that setting up your own systems and maintaining your own machines, and being responsible for your own data is "too technical" for the average person.
If we are intelligent enough to raise babies, drive cars, operate propane barbecues, SHOOT GUNS for f*ck's sake (in the USA anyway), then come on people, it's time to man up.
Its cool and all, but I don't think I could convince my company (and probably any other) that its worth $2/user. I think maybe there should be plans that can be hosted on the company servers. Maybe a 1 time fee to license the software and then support plans or 1 time support costs?
You can't convince your company (who is paying you thousands of dollars a month in salary) to spend an extra $2 a month on something that will make you more efficient? What kind of company do you work for?
Its worth $2/user not to have to set stuff up on my company servers. We're a small company though (6 people now) so maybe the economics are different where you are.
If it improves your productivity more than $2 a month, then its more than worth it. For us, it was a no-brainer after we started using it. Giving everyone access to everyone both asynchronously and real-time is so powerful. We used IRC and Campfire before, and neither gained as broad adoption internally.
Can anyone comment on the battery life for the iOS client in comparison to skype? I am a heavy skype user, and it doesn't seem like hipchat has a lot to offer than skype does not. Not to mention the inertia that skype has.
however, I would totally consider it, if i could keep it on my phone all day and not have 10% battery by 2pm.
also, curious, does it do group video chat, ala skype?
Already seen a major issue pop up a few times where I type in a few different messages, minimize the app, come back to it, and none of my messages are there. My team is offline right now so not sure if they're getting the messages or not but damn is that annoying so far. Seems like a big deal if the core use of the app is flakey.
One thing I just noticed is that the native app uses a LOT less memory than the AIR client: 43 MB native vs 100-150 MB air. It definitely feels snappier too, from startup time to normal usage. Even with the limitations (no local Find functionality?!) I'm still very happy with this beta. I'm sure they will continue to improve.
I've been using HipChat for a distributed development project. I like this app. Incumbent Adobe Air app is nice enough for me though animations such as scrolling is not quite sophisticated.
Agreed. The air app was great, but had a lot of limitations. The mac beta doesn't have quite all the features the air app does yet, but its fast and supports retina so I'm happy.
Awesome - i have been using hipchat for years and haven't needed anything else. Most of my company is mac so this is a no brainer. Great work Hipchat team!
I do now. I thought Atlassian had built it in-house. My worry with these is that acquisitions change the product priorities. For example I was sceptical of the recent Crashlytics acquisition by twitter but they've already gone and improved the experience for everyone by offering enterprise features for free. Its hard to tell who's going to be absorbed and fade out in these kinda things.
We've been with Atlassian for almost a year now. I hope big initiatives like this Mac app show that we're still working hard on the core product. We've got some great stuff in the works, too.
It really is crappy. I fall offline all the time, with no explanation. At least I know I'm off, though - the guy who sits next to me continues to appear available on his machine, but is offline to the rest of us. It's a nightmare.
This mac app has been in development for ages, then been in closed beta and now finally it's available to everyone to try.
Our team just switched to the beta native app, and already found several critical bugs with the app within minutes.
Sorry guys, this is really poorly put together and the number of obvious bugs makes me think that you don't even use your own product!