The Legacy Editor is being deprecated in Confluence Cloud: Here’s what you need to know

[John Halbig]

We still need a way to disable the stupid and spammy "AI" functionality (https://jira.atlassian.com/browse/AI-930); Aside from the inexplicable markup changes from the prior version, broken tables, broken layouts and so forth, that is my biggest reason for not downgrading to the new version.

You heard me.

[Keith Furnell]

For those trying to keep track of all the unwanted ways that page content currently changes during conversion, there is a table here identifying types of content that "CONVERTS - MODIFICATION":

https://confluence.atlassian.com/confcloud/converting-pages-to-the-new-editor-993930226.html#Convertpagestotheneweditor-pageconversion

Unfortunately:

• It's a very big table and not sortable by "What happens when converted?"
• It seems to be incomplete, e.g.:
  • doesn't seem to mention https://jira.atlassian.com/browse/CONFCLOUD-67008
  • doesn't seem to mention loss of table/column widths (surely this can be improved?!)
• It references work items that are supposedly FIXED for "CAN'T CONVERT" rows, e.g.:
  • Nested tables - https://jira.atlassian.com/browse/CONFCLOUD-58656
  • Macros in tables - https://jira.atlassian.com/browse/CONFCLOUD-63350

[Dino Willy Baraldo]

In summary, the Cloud Editor represents a technological advancement, but the lack of an effective migration process risks turning a potential improvement into an operational obstacle. Atlassian will need to listen carefully to community feedback to ensure a transition that better respects the work already done.

[Rebekka Heilmann _viadee_]

@Dino Willy Baraldo very well said!

[Elle Ky]

@Martina Guth 

The legacy content macro is a long-term solution designed to support legacy content that isn’t natively supported in Confluence Cloud. It plays a critical role for users working with the legacy editor in Cloud, as well as for customers migrating from Data Center to Cloud. At this time, we have no plans to deprecate the legacy content macro, as it remains essential for bridging the gap between Data Center and Cloud experiences.

Regarding your image issues: we’ve made progress over the past few weeks on improving image conversions, but it’s not clear from your screenshot how your images are set up and whether it would apply to your situation. Could you try reverting the page and running the conversion again? If the issue persists, can you please file a support ticket so we can investigate further? We’re also continuing to work on additional image improvements over the next month.

Please note that we are still actively making improvements to the conversion process, and your feedback helps us prioritize and address the most impactful issues!

[Elle Ky]

@Keith Furnell Thank you for sharing your feedback. Here’s an update on some of the areas you raised:

• Changes in image size and alignment: We’ve recently shipped improvements here and will continue to roll out additional improvements over the next month.
• Header rows or columns getting un-bolded: This issue is now resolved. 
  • In legacy editor, the table header text doesn't actually have the bold property applied (verified by clicking on the text). They appear bold, but the bold property isn't applied to the page source. In legacy, you don't have the option to have un-bolded headers, whereas in cloud editor, you do. Cloud now preserves the intended bold styling for headers for converted pages.
• Note/Info/Warning panels being given Header 1 styling, causing them to be added to the Table of Contents: This is also fixed. Panel titles now convert to bolded normal text, so they no longer appear in the Table of Contents.
• Extraneous lines between paragraphs, tables, and images: We’re actively working on reducing unnecessary lines and improving overall formatting consistency.
• Full-width tables and panel conversion: We’re currently investigating and exploring improvements in this area.

[Elle Ky]

@s_weber Thank you for sharing your analysis and for highlighting your concerns. We absolutely share your goal of making this as seamless as possible.

On identifying affected pages:
Admins can already generate a CSV-formatted list of all legacy editor pages and their conversion eligibility using a script, as outlined in our FAQ and detailed instructions. We’re also working on an in-product solution to make this process even easier—this will be available in space settings (not content manager). You can follow our progress in the relevant JAC ticket. In the meantime, you’re not blocked from identifying affected content.

On improvements to the conversion process:
We hear your concerns about quality and manual effort. Improving the conversion experience is a top priority as we have communicated. Please see my comment above to Keith Furnell to understand recent improvements we made. While not every incremental improvement will be announced, rest assured that the work is ongoing and prioritized based on feedback.

The best way to share your feedback:
As noted in our FAQ, if there are specific features that are important to your workflows, please search for them on jira.atlassian.com. under Confluence Cloud suggestions. Voting for and commenting on suggestions directly helps us prioritize the improvements that matter most to you. Feedback posted in community forums is much harder for our product teams to track and action, so we strongly encourage you to use JAC to ensure your voice is heard and your needs are visible to our development team.

On escalation and feedback:
If you require an urgent response, we encourage you to open a support ticket so the Atlassian team can assist you as quickly as possible!

[Chris]

UNACCEPTABLE! Unless you have a solution to our image issue, that was brought up nearly a year ago. 
https://jira.atlassian.com/browse/CONFCLOUD-68501
https://community.atlassian.com/forums/Confluence-articles/Inline-images-coming-to-you/ba-p/2611260/page/2
100% of our documents will be effected by this forced change.

How do we get this:

To look like this:

[Chris]

"At this time, we have no plans to deprecate the legacy content macro, as it remains essential for bridging the gap between Data Center and Cloud experiences." - Elle Ky Atlassian Team

I'd have to go digging through old emails and posts, but I'm pretty sure the same thing was said about the actual legacy editor not going away, yet here we are.

[Laurie Hickey]

@Chris You have expressed the same fear that I have -- the Legacy Content macro getting deprecated at some future date, forcing us to identify (ha!) and edit all of those pages to make them completely compatible with the Cloud editor (similar to the fire drill we are dealing with now).  Unfortunately, our current focus is on identifying pages that are INELIGIBLE for conversion so that we can build new (replacement) pages in the Cloud editor prior to January 2026. :(  We will have to circle back around to the Legacy Content pages at a future date.

[Chris Jordan]

I've been reading through the comments and agree with the fact that the new editor is a serious regression from the Legacy editor.
From even something basic as the amount of font colours available is seriously lacking. It is hard to note what is contained in a section as there is no longer any dotted line to delineate between sections.
And as mentioned the increase to white space is ridiculous. It makes everything bigger where there is no need for it to be.

[Dino Willy Baraldo]

I’ve started using the new system, which replaces "Legacy Editor", and I’ve encountered a behavior that negatively affects the user experience. Specifically, the system displays contextual popups whenever the cursor hovers over a word or text block. This behavior is intrusive, especially in scenarios involving basic text editing, where AI assistance is not required.

Currently, there doesn’t appear to be an option to disable these features, which are unnecessary for my use case. It would be helpful to have the ability to configure the level of AI assistance or completely turn off automatic suggestions, in order to streamline the workflow in contexts that require minimal writing support

[Rick Westbrock]

@Keith Furnell in reference to https://jira.atlassian.com/browse/CONFCLOUD-63350 which you mentioned I tested the behavior of adding the Expand macro in a table cell.

On one of my pages which uses the new editor I created a new table, clicked inside a cell then typed /expand and it placed the macro in the selected cell as expected. I then added a nested table and was able to add the Expand macro inside a nested table cell. This was done on a page which was created in the new editor however, not a converted page so I don't know if a converted page displays the same behavior or not.