Hey, Atlassian! How about fixing some bugs?

[scott.dudley]

Hi Atlassian,

I opened my email this morning to find 44 notifications from JAC documenting a large set of Confluence Cloud bugs and feature requests that have just been closed as “Won’t Fix”. I monitor only a small subset of issues in particular niche areas of the system, so I imagine many more issues were touched.

It would be nice if this trend could be going in the other direction…which is to say, it would be nice if Atlassian decided to fix 44 bugs rather than closing 44 bugs without action.

If the existing PMs are not motivated to allocate resources to this work (I would agree that fixing bugs is not usually considered “fun”), what about the idea of creating some “ecosystem embed” teams of developers within the actual product groups, that can prioritize work related exclusively to the ecosystem (of which fixing long-standing issues could be a component)?

With the increased revenue share that Atlassian is planning to take as of 2026+, it would be a much-needed show of support for the community to actually spend some of those extra funds to work on things that matter to the ecosystem, rather than just having the extra rev share vanish into the black hole.

From a user’s perspective, I see a large set of issues in the platform that range from slightly to significantly broken, but I rarely see significant effort from Atlassian put towards fixing non-crit/non-sev1 bugs. Over time, these problems accumulate and they lead to the feeling of “death by 1,000 paper cuts”.

So, my suggestion is that Atlassian consider creating some KPIs and allocating resources to execute, on regular basis, some sort of “quality sprint” where people work on nothing but correcting problems that have been generated in previous development iterations. It is really needed at this point, and it has to go beyond fixing critical-severity problems.

For example, in today’s batch of closed and “Won’t Fix” issues I see:

• a bug where a broken API causes permanent data loss for customers
• a bug impacting confidentiality and the potential unintended disclosure of information that was not meant to be shared
• a bug that prevent apps from doing things that seem somewhat essential (like uploading images to a page with an API and having the images actually be rendered for clients)
• problems and inconsistencies in API implementations that force effectively all Cloud developers to build workarounds

Again, if Atlassian categorically ignores most issues classified as “minor”, the aggregate eventually becomes a “major” issue.

[nathanwaters]

The competition:

[SilvreLestang]

I am working as a developer in the Atlassian ecosystem for more or less 3 years now, and I feel more and more the “death by 1,000 paper cuts”.

The time and energy to work around bugs, limitations, platform depreciation, weird behavior and inconsistency is tiresome. And it doesn’t seem it is improving at all.
Forge is just yet another things with its new bugs, inconsistency, baldy design architecture … that we have to learn and handle, on top of the others things. Yes, Connect will disappear soon, which will free a little bit our minds of all the quirks/bugs/… it contains.

As a developer, I am more and more conscious that most of the things I learned in this ecosystem are unique to it, and won’t be useful outside Atlassian, which from a knowledge and career perspective is scary.

One day, I will just gladly quit this ecosystem with relief.

[asridhara]

Thanks for sharing this, Scott. This is an area we definitely need to improve on. We’ve already started internal conversations on how we can better address bugs and assess their severity. I don’t have an immediate update, but I want to acknowledge that this is something we’re actively looking to resolve soon.

[MichaelAglas1]

Amen.

[BNormannPNielsen]

This also goes for Data Center (as not declared death so far…) and a lot of “gathering intrest” bug, wich may be a valid status, but Ive been here for +15 years, and knows that a “gathering intrest” bug will problably never be fixed, and I would assume a log of the 44 has been in that status. At our office, that status has almost been elevated to an official joke…unfortunately…