Sorry, typo. I meant to say OnDemand is the old name of Confluence Cloud. This issue is not invalid, you just changed the name of the product.

Editing the source is a fundamental feature without which Confluence is unusable for many purposes. In particular, it's required to work around WYSIWYG editor bugs you will probably will never fix.

We should not have to pay a third party for a fundamental feature.

Hi Robert, the CONF project was split into CONFSERVER and CONFCLOUD. Some tickets were cloned which may not have affected both platforms.
OnDemand is the former name for Cloud deployments (not Server). That is, instances with a URL <your_instance>.atlassian.net.
There are a number of Source Editor plugins available on the Atlassian Marketplace:

• Cloud and Server (paid) - Source Editor for Confluence
• Server (free) - Confluence Source Editor

This issue was only about Confluence Cloud (formerly named Confluence OnDemand). It's as valid as it ever was.

What was invalid was moving it to the CONFSERVER project instead of the CLOUD project.

As this Suggestion has been split between in to separate Cloud and Server tickets, this ticket is no longer valid. Please refer to CONFCLOUD-24922

Seems to me that this would provide a work-around for so many other issues, the real value won't be reflected in votes purely on this one task, it's a structural capability rather than a final pain-point, hope that's taken into account in the voting system.

Our team at Mohamicorp has just released a plugin, Source Editor for Confluence, that provides Confluence XHTML source editing for Confluence Cloud.

We're looking at performing a migration of multiple wikis from another platform to Confluence, but migrations of this sort are never perfect and if we can't clean up the resulting storage data, life is going to be hard.

• I cannot paste basic safe html eg. table from another source and have it interpreted correctly.
• I cannot find and replace in the editor when a table is on the page as it ignores all except the first scope hit.
• I can view and edit the storage content, do the find and replace myself, but pasting it back doesn't parse and there is no option for replace content, only view it.
  • That is a particularly unfunny joke at my considerable expense.
• Telling me how fantastic the editor and macros are, and how we are listening and created 'something for the section of the userbase not using our paid hosting service' is not just unfunny, it is rude.
• Years of 'pain point dev days' where your significant pool of talented staff are encouraged to pick something to fix - but parsing unsafe html is too hard?

K15t is making Scroll PDF Exporter available as a Confluence Cloud plugin, but that's useless to me without a source editor.

It's a bad idea to not allow editing of the source directly.

If the worry is support calls from people that edited the source then simply notify the user that source editing is an unsupported feature.

If the worry is security then strip all "bad stuff" from the html before it is sent to the database.