Scheduled posts are now published more reliably

Accepted posts get published

• Automatic retries for about 22 hours: A scheduled post that fails to publish is retried automatically 7 times, after 1 minute, 5 minutes, 15 minutes, 1 hour, 3 hours, 6 hours and 12 hours (it used to be 2 times over about 6 minutes). Posts still go out when a server-side problem lasts a long time. Problems you have to fix yourself, such as an expired connection, are reported right away.
• See which posts are "waiting to retry": Posts being retried automatically show the next attempt time and the number of attempts. "Waiting to retry" is a new status in Post Management, and "Failed" and the Home failure list now show only posts you need to fix yourself.
• Checked again before the scheduled time: 24 hours and 1 hour before a post, we recheck the account connection, authorization and any suspension, and notify you in the app if the post cannot be published as is. If you fix it before the scheduled time, it is published as planned.

Problems are declined up front

When a post is accepted, we check the date and time (in the past, or without a time zone), the text length, an empty body, recent duplicates and the connection to X. Approving or posting directly to X in a team goes through the same checks as personal scheduling. The 24-hour post limits for Instagram and Threads and the monthly limit for posts with URLs on X are also checked at that point, and a post over the limit is declined with the reason and the time.

Fixes

• Instagram and Threads posts could be published twice when they were resent after the result did not reach us. Before resending, we now check whether the earlier submission went out, and mark the post as published if it did.
• An X thread could stall when its first post was temporarily unable to send. A resend now sends the whole thread again.
• Editing a scheduled X article could publish the old version. Editing is now declined while it is scheduled; cancel it first, then edit.