At least you can still get AIR for OSX. Trying running the HipChat client on Linux... it's pretty painful, and getting more painful as time goes on (and we get further and further from Adobe's last release of AIR for Linux). It's even more painful to get it running on a 64 bit Linux distribution...
- AIR doesn't play nicely with dynamically switching graphics chips (read: all MacBook Pros and iMacs). The app itself, AIR's installer, and updater, will all randomly crash due to this, with no hint at what's wrong. The only way to fix this (if you've somehow realized this is the problem) is to lock your graphics card in System Preferences.
- It will peg your graphics card, just like Flash on Mac. You can be looking at a plain chat window and hear your MacBook Pro spinning up like a jet engine, as if you're about to do something, well, more substantial.
- It crashes, and worse, it has a really bad habit of corrupting itself when crashing. To-date I've reinstall HipChat at least 10 times on my machine. The "AIR installer crashes randomly because of graphics card problem" undocumented bug makes this even more annoying.
- It's just slow.
johnnydoe is right. AIR is like client-side Java, or any other write-once-deploy-many products. It's lowest-common-denominator crap. It has all of the limitations of its target platforms, none of its strengths, and all of the laziness that comes with trying to design a one-size-fits-all UI for everyone.
Sweet, this means my Atlassian-loving company might finally start using a web-based chat room system. For some reason our teams love email, wiki, IMs, but won't even countenance the benefit of realtime chat.
We have a permanent jabber room on our server that all but one dev uses. No one was ever told they needed to be there.
Just start using a chatroom and invite people in. Remind a few to add it to the auto-join list. If your coworkers find a benefit in it, there doesn't need to be an official policy; the usage will just grow organically. If they don't see a benefit, why force it on them?
Push notifications. They’re confusing. All the in-person communication is pushed, however room conversation is not unless you are @mentioned.
If you would like to make an announcement to 10 people, you have to mention them one by one manually.
why is atlassian acquiring commodity tech? I don't understand their strategy.
edit: as opposed to just building it themselves, what's so hard about a chat app? maybe i underestimate the amount of iteration in the hipchat product and they're buying that product knowledge, not the tech.
Buying the tech helps them get there faster. Also they don't just buy another product, HipChat has a very fanatical user base. It will likely be easy for Atlassian to get it adopted by the rest of their user base.
Crucible/fisheye - slow. I mean really slow. Unusably slow both with respect to the UI and back end. Half our team use IE which it just doesn't work in at all. Chrome is the only thing it just about crawls along in. Incredibly difficult keeping it alive with 50 users. Crashes once a week entirely and sometimes refuses to start with no error messages at all.
JIRA - workflow crash took our team of 30 devs at the time out entirely for 2 days. Overcomplicated administrative mess especially with respect to plugins. Reindexing takes out the entire JIRA instance until complete. Permission schemes and workflow is an epic mess of cludges.
Both products: Null pointer exceptions galore, scary memory ceiling. Integration sucks - all the horrible mappings to maintain.
They feel like bloated, slow and badly designed products which is worrying considering the cost.
TBH their support are good but I shouldn't have to use it on a product suite that costs that much.
How are any of these reasons that HipChat is "about to be ruined"? Both of those are huge products, and afaict were developed by Atlassian - not acquired.
Atlassian has made two big acquisitions that I know of (Bitbucket, SourceTree), and in both cases the product was better off after the acquisition.
From what I see, Atlassian goes in and provides cash and infrastructure, leaving the teams that were responsible for the product's success to continue making great things. I don't see any indication that things will be different for HipChat.
I agree with crucible/fisheye being slow, but they have greatly improved it lately.
As for the UI you basically have to run a recent firefox/chrome/chromium.
All your other points I somewhat agree with but to their credit they all improve steadily each release.
Be mindful that the products were source individually from various places and need a lot of work to look and work together nicely.
Let me finish on some faint praise: they are by FAR the least worse of all the 'enterprise' software we use :-)
Unfortunately, due this responsibility being approximately 5% of my time, I don't have time to invest in to-and-fro with Atlassian. I expect something to work (ironically like Trac did - its predecessor).
On the positive side, we're upgrading to latest versions soon.
My personal experience is that JIRA is an unmitigated UX design trainwreck. Most of the UI pages and workflows that I've seen are nightmarishly kludgey. It makes Bugzilla's UI seem decent by comparison.
I haven't used their other products (other than BitBucket, which is fine, but hasn't kept up with Github), but JIRA desperately needs a UX design vision.