55 Bug Tracking Jokes Every QA Engineer Has Lived

The ticket was opened in 2019. It is now P2. It was P2 when I joined. The reporter no longer works here. The component no longer exists.

The bug is reproducible. It requires a full moon, the staging database, and a user from a country we do not yet support.

I marked the ticket as won't fix. It was reopened. I closed it again. It was reopened again. The reporter is my own past self from 2018. I cannot stop them.

Triage is on Thursday. There are 217 tickets in the queue. There is one hour blocked. We have done this every Thursday for four years. The queue has never been empty.

The bug only happens on the user's machine. The user is the CEO.

Severity 1. Priority 4. Both labels are on the same ticket. They are both correct. Nobody has been able to explain why.

I added needs more info. The reporter added a screenshot. The screenshot was a photo of their monitor. The photo was taken at an angle. The screen had glare. The error message was illegible.

I labeled a ticket cannot reproduce. Three weeks later it reproduced. The ticket had been closed. I did not see the comment. The user filed a new ticket. We now have two.

The ticket said the homepage was broken. No URL. No browser. No screenshot. No steps. Just broken. I marked it needs more info. Nine months later it is still needs more info.

I once filed a bug against myself. I assigned it to myself. I asked myself for clarification. I marked it as blocked on me. I closed it as won't fix. The retrospective was tense.

The bug was a feature in the spec. The spec was wrong. The PM filed the bug. The PM had written the spec.

Severity is the customer's word. Priority is ours. Customer says critical. We say P3. Both are documented. Both will be cited in the postmortem.

The ticket is on the backlog. The backlog has 4,800 items. We groom it every two weeks. The grooming has been moving the same six tickets since the spring.

I once closed a ticket as fixed. The fix had not shipped. It was in a PR that had been open for a month. The reporter ran into the bug again. The reopened ticket cited my closure as a learning moment.

A ticket was duplicated as a duplicate. The duplicate was also a duplicate. The original duplicate had been closed. The duplicate of the duplicate is now the canonical ticket. Nobody can find it.

I added a comment that said I think I see what's happening. The comment was from 2021. The ticket is still open. We do not know what I had seen.

The ticket title was URGENT. The description was lowercase. The reporter was the VP. The fix took two hours. The communication took six days.

I once filed a P0. It was the wrong P0. The real P0 was a different bug, the next day. By then the team had stopped reading P0 alerts. The pager was now folklore.

The ticket has 47 comments. None of them describe the bug. They are all about ownership. The bug, in the first comment, is one sentence.

I closed a ticket as expected behavior. The reporter replied that the behavior was not expected by anyone outside the team. They were right. We reopened it. We renamed it spec gap. The spec is now under review.

Someone added the label needs design. Design added the label needs engineering. Engineering added the label needs product. Product added the label needs triage. Triage was on Thursday.

The reproducer required clicking 14 times in a specific order. The 14th click crashed the app. The fix was a one-line null check. The ticket was open for three years.

I once asked a customer for a video. They sent a 45 minute screen recording. The bug appeared in minute 38. I rewound. I confirmed. I closed the ticket as known issue. The known issue ticket had been opened in 2020.

Linear has cycles. Jira has sprints. GitHub has milestones. The bug is in none of them. The bug has been in none of them since 2021.

A bug was filed against the design system. The design system is owned by a team that does not exist anymore. The bug is now everyone's, which means it is no one's. The bug is still there. We have learned to live with it.

I once closed a ticket and the reporter sent a thank you. The thank you was a slack message. The slack message ended in a screenshot. The screenshot showed the bug was still there. They had thanked me for the comment, not the fix.

Severity 5 exists. It is for bugs that are technically wrong but nobody will ever notice. We have 800 of them. They are the team's true backlog. They will outlive the company.

The bug is in production. The staging environment does not have the bug. The staging environment is also six months behind production. The bug is the difference.

I once spent an afternoon writing a perfect reproducer. The ticket was duplicated to an existing ticket the next morning. The duplicate said see other ticket. The other ticket said cannot reproduce.

Triage assigned a bug to the wrong team. The wrong team triaged it and assigned it to a different wrong team. The bug has been triaged seven times. Each team has marked it not us. It is now the bug's tour of the org chart.

I labeled a ticket flaky. It was a real bug, intermittently. The label has been on the ticket for two years. The label is now an ownership transfer. Anyone who sees the bug ignores it. The bug is now a feature of the ticket.

A ticket title said it shouldn't do this. I asked what it was doing. I asked what it should do instead. I asked which page. The reporter answered all three questions with yes. I closed the ticket as needs more info, again.

The ticket is two years old. The PR that closed it was merged last week. The release notes will say bug fix. The bug had a name in the codebase. The name was Bob.

I once filed a ticket on a Friday at 4:45pm. Monday morning it had three comments, two labels, a priority, and an owner. None of them were mine. The ticket had outgrown me over the weekend.

There is a ticket I keep meaning to close. It has been closed once. I reopened it because I had a thought. The thought is gone. The ticket is open. It will probably outlive me.

I once added a comment that said see Slack thread. The Slack thread was in a channel that was archived a year later. The link is dead. The comment is now mystery scripture. Engineers cite it. Nobody knows what it said.

Atlassian shipped a new Jira UI. The filters I had spent six years tuning no longer work. I am rebuilding them. The tickets, of course, are unchanged.

I once asked for a postmortem on a bug. The postmortem became a ticket. The ticket became a sub-task. The sub-task became a blocker on a roadmap item. The roadmap item was deferred. The postmortem was never written.

The bug has a workaround. The workaround is documented in a Confluence page from 2020. The Confluence page links to a Slack thread. The Slack thread links to a JIRA ticket. The JIRA ticket says see Confluence.

I closed a ticket as obsolete. The component had been removed. The bug had not been fixed. The bug is no longer reachable. The bug is still there, in the code, somewhere. We call this a soft fix.

There are 14 fields on the ticket form. Only two are required. The team has agreed to fill in five. We actually fill in two. Sometimes one.

I once added a watcher to a ticket. The watcher was on PTO. They came back to 312 ticket emails. They unsubscribed from the project. The project lost its memory.

I labeled a ticket good first issue. A junior picked it up. The fix touched seven systems and required a database migration. The junior shipped it. I have stopped labeling tickets good first issue.

I once closed every ticket older than 18 months. The team called it bug bankruptcy. The customer support team called it Monday.

Linear's Cmd-K is faster than Jira's search. Jira's reports are deeper than Linear's. The team picked Jira. The individuals all use Linear's keyboard shortcuts in their muscle memory anyway. The mismatch is the new bug.

A bug is filed. A PR is opened. The PR is merged. The bug is closed. The user files the bug again the next day. The fix had not shipped. The release was held. The release notes were already drafted. The drafted release notes are now archaeology.

I once added a comment that said this is intentional. The comment was on a bug from before I joined. The comment was wrong. The bug was not intentional. The bug got promoted to a feature on my word alone. The feature is now load bearing.

I have a saved filter called my tickets. It has 89 results. Three are in progress. The rest are in a quantum state. Looking at them collapses the wavefunction into more work.

Someone asked which bug tracker we used. I said all of them. We use Jira for engineering. Linear for design. GitHub Issues for the open source repo. Trello for marketing. Notion for the staff engineering group. Email for the CEO. The CEO is the busiest tracker.

I marked a ticket as needs repro. The reporter sent a repro. I ran the repro. It worked on my machine. I closed the ticket. The reporter sent a video of the repro failing on their machine. I had not asked about their machine. The bug was their browser, from 2017.

The estimate on the ticket was 2 points. The fix took three weeks. The retrospective said estimates were aspirational. The next sprint had a ticket with the same shape. It was estimated at 2 points. We have learned nothing. We are at peace.

I once filed a ticket called the website is slow. The ticket is now an epic. The epic has 47 sub-tasks. The website is still slow. The epic has its own steering committee.

Two reporters filed the same bug independently on the same day. The first reporter was thanked. The second was marked as duplicate. The second was the one who had written the affected code. The dedup is technically correct. The blame is misallocated.

I once added a custom field called actual root cause. It has been filled in twice across 4,000 tickets. Both times by me. Both wrong. I am the only consumer of the field. I have not deleted it. It is a monument.

The ticket I opened a month ago is closed. The closing comment says fixed in main. The fix is not in main. Main has not been deployed in a week. The bug is still in production. The ticket, however, is closed.