Hacker News Reader

Comment by Vinnl | Hacker News Reader

Having worked at a large company before that had their own instance of Confluence Questions (i.e. Atlassian's SO clone): it was incredibly useful. As a new employee, it allowed me to relatively quickly find the relevant people, or even to come across answers by them to questions people kept running into over and over again.

As I became responsible for an internal library used company-wide, I pushed for the Q&A site to be our primary means of communication, and would always refer to it after answering questions from people contacting me directly. I saw the same pattern repeat itself for many new hires.

Otherwise I think your observations are spot on.

View on HN

It is simply not possible for Confluence Questions to be considered a useful product by even the most generous definitions.

The reason is simple: it only supports exact-matching text search _on titles_. There are a host of other usability problems that are so bad that they render a giant negative impact on information retrieval and actively steer Q&A traffic to repetitious and ephemeral tools like Slack.

But the lack of a functional search engine indexing the Q&A content with a high-performing notion of search relevancy is such a show-stopping, egregious defect that even all the other egregious flaws can be ignored by comparison.

It’s not a subjective debate at that point: just a factual defect of such severity that any misguided reports of the tool somehow miraculously being even marginally useful for someone have to be ignored.

I thank my lucky stars that Stack Overflow offers competition to this with a real search engine.

[0]: < https://community.atlassian.com/t5/Questions-for-Confluence/... >

I am on a committee in my workplace that deals directly with Confluence customer support to roll out changes to our internally deployed Confluence & Jira instances, and I know first hand from dealing directly with Confluence that they _intentionally_ only support exact matching _on the titles_ for searching in Questions. Really appalling.

I haven't used Confluence Question, but I use Jira, Confluence and Bitbucket every day, and search is also the limiting factor of all of those product.

Jira search is useless, you have to use filters and memory of dates to get a set of issues small enough for you to search manually.

The search in Confluence can best be summed up as "Here's a PDF". For some weird reason Confluence can search PDF sort of okay. It's typically not what you want, but because Confluence can search PDFs better than normal Confluence pages, PDFs are what you get in 75% of your searches.

Atlassians continued failure to deliver a half decent search feature in any of their products is fascinating. I honestly don't mind their products, they're fine for our needs, except you can't depend on search.

Disclaimer: long time Atlassian employee but I do not work on Jira/Conf/Bitbucket & all words here are my own opinions.

A major problem is that many deployments of Atlassian products are Server deployments (i.e. not Cloud), which means there's basically the embedded Lucene engine and that's it; anything else starts hurting either performance or deployment options at scale. And most Server customers have a much greater interest in scaling (while keeping servers performant & easy to run) than getting tweaks to search algorithms - in particular many customers refuse to upgrade unless there is a major bug fix or performance boost they're going to get.

On the Cloud side, for better or worse the focus has mostly been on letting people find their work, not find new content they haven't seen before. For example, Confluence's My Work tab (or similar - I forget the exact name), or the search in the cross-product https://start.atlassian.com/ are both useful for getting back to things you've seen before. Hopefully some day we'll finally make some headway on the "I don't know what I'm looking for but I know how to describe my problem" use case.

And just to buck the trend, search works as advertised in the Atlassian product I work on, statuspage.io ;) (but then our search use case is straightforward and rare enough to just use elasticsearch with some minor tweaking).

Search on Bitbucket is similarly broken for any cases more exotic than an exact word match. So in practice, it's useless for source code.

I agree that there was a lot that was lacking about Confluence Questions; my main point was that a StackOverflow-like product is incredibly useful. SO itself would probably do a better job at fulfilling that role than Confluence Questions.

The company I work for uses internal SO and I think it’s pretty useful. There’s a culture to push people to turn useful slack answers into SO q&a so that they are useful for others. Also, our internal search engine searches over it (as well as the wiki, jira, etc.)

Yeah my experience is that large places that use things like Yammer it ends up getting used as basically a Q&A site.

Yammer is more like a private Facebook. StackOverflow is a technical trove. Distinctly different goals.

Right, I understand that, but my point is that how it actually gets used is in largely a private-stackoverflow-substitute capacity, even though that's not really its intended use case. That's where the 'desire lines' go.

Perhaps for a large company with multiple offices etc, sure.

For a smaller company though, other than as a search tool (previously asked questions) I don't see what advantage it has over just using Slack/Teams etc. in a relevant channel to your question.

Because even in a small company, searching past content and conversations in Slack is a pain. The information gets lost. There is no reference back, so people just end up asking the same questions again, which is both super annoying and a gigantic waste of time.

I mean, I assume that they don't actually care about small companies, they don't have much money anyways. In these scenarios, they tend to act as a "loss-leader" for marketing purposes, mostly.

FWIW Atlassian's perspective is generally more like "they don't have much money yet, so let's get them while it's easy to get them". Small companies grow into big companies, and small single-team deployments tend to spread into big whole-company deployments - but small customers are much easier to get than large customers :)

Disclaimer: long time Atlassian employee but all words are my own.

Yes, companies large enough where you might not know who you'd need for help with a problem within your first week is exactly the use where SO shines, in my opinion.

Reply on news.ycombinator.com