4 ms·
Wonder if Atlassian still thinks it was a good idea to retire Hipchat and take 10% of Slack.
by robertlagrant 5y ago
Wonder if Atlassian still thinks it was a good idea to retire Hipchat and take 10% of Slack.
- davidjgraph 5y agoWouldn't they have been paid for their 10% and made a profit?
- stonemetal12 5y agoProbably, but going forward they no longer have a chat app and more or less either have to revive HipChat or refer people to a competitor.
- bachmeier 5y agoThey had already announced the retirement of Hipchat long before they made their deal with Slack. They agreed to stop development of their Slack competitor named Stride. IMO Stride had a lot of potential.
- dannyw 5y agoStride wasn't really going well. There were some obvious, serious issues with technical and product management, for example the insistence on using Jira's javascript-based textbox rewrite ("Fabric") as the GUI element for composing messages, under the spiel of 'unification'. As "Fabric" is a web component, this meant that mobile apps for Stride essentially had webviews for their input box, introducing additional latency (critical for typing), bugs, inconsistent editing behavior to native iOS/Android UIs, and performance issues especially across lower-spec android devices (it's just a textbox FFS!). Plus, engineers were wasting time figuring out how to sort browser and webview compatibility issues, instead of, you know, implementing a textbox using native technologies which isn't rocket science. To make matters worse, the team working on Fabric was juggling between fixing INPUT BOX bugs on Jira, on Confluence, and on Stride, making it a gigantic mess that kept multiple product managers distracted for the wrong reasons.
- robertlagrant 5y agoIs the input box technology choice really that insurmountable a problem to fix in the future?
- dbbk 5y agoWYSIWYG composers are hard to do well, though. They're certainly not trivial.
- pram 5y agoHipChat was an unmitigated dumpster fire. The backend literally couldn't scale for many reasons, and it was a rats nest of different databases and random libraries. For example, each connection required a discrete process for managing the BOSH session (as in, one process had to be dedicated for the entire time someone was connected to HipChat) Stride wasn't really any better, their whole strategy was to outsource the backend to PubNub
- dbbk 5y agoStride had some really nice features though, like highlighting the final decision coming out of a discussion thread, and rolling those up into daily highlights. No one else seems to have carried that on.