Engineering / Field note
A WordPress draft’s date is not a publication receipt
WordPress can refresh a floating draft’s date during an update. The core conditions show why original dates, migration times and actual publication need separate evidence.
A WordPress draft can acquire a fresh-looking date during an update without becoming a published article. The 7.1 branch of WordPress core contains an explicit date-refresh path for particular undated drafts. That makes a date displayed beside draft content a limited piece of evidence about the content’s history. [2]
The conditions are specific: the existing status is draft, pending or auto-draft; the existing post_date_gmt contains the all-zero date; and the update’s edit_date value is empty. In that case, wp_update_post replaces the local post date with the current site time and clears the GMT input before passing the update onward. [2]
This is an analysis of the public core source checked on November 18, 2026. It is not a new vulnerability announcement or a claim that every WordPress date changes on every edit. The exact status, stored fields and update inputs matter to the behavior.
Our interpretation is that content history needs names for different events. When an author wrote something, when it first became public, when a migration copied it and when someone last edited it are separate facts. A convenient date column cannot establish all four merely because it contains a valid-looking time.
The draft clock can move while publication stands still
WordPress’s reference says a draft does not need an assigned date. Its API fields distinguish post_date from post_date_gmt. In the pinned core, statuses registered with date_floating enabled can retain the zero GMT date through the insertion/update path. [1] [2]
Consider a hypothetical knowledge base copied into a new CMS as drafts. A later editorial pass corrects a paragraph while leaving its status unchanged. For a draft meeting the documented conditions, the local date can move to the update time. That is an illustrative consequence of the core path, not a report of an import or repair we performed.
If another part of that hypothetical system later interprets the local date as the article’s original publication time, it has changed the meaning of the evidence. The editor may have intended a text correction while a downstream report appears to describe a different publishing history. The problem is the interpretation attached to the field.
In our view, an original source date deserves its own provenance: which record supplied it, what the source called it and how confidently it can be interpreted. A migration timestamp answers when a copy was made. It does not supply missing evidence about when readers first saw the original.
A local time needs its original context
WordPress documents get_gmt_from_date as converting a date in the site’s timezone into UTC. Its source uses the site’s timezone for that interpretation. A conversion therefore depends on more than the digits in a local date string. [3]
RFC 3339, published in July 2002, defines Internet timestamps with an explicit relationship to UTC, through a numeric offset or the UTC designator. It also discusses the interoperability problems of unqualified local times. These are longstanding format considerations rather than a recent change to WordPress. [4]
For an old record without an offset, choosing the current website’s timezone is an assumption about the past. Our reading is that the assumption should stay visible until evidence supports it. Preserving an uncertain raw value can be more honest than presenting a neat conversion whose historical basis has disappeared.
A successful write answers a narrower question
The wp_update_post reference documents a post identifier as the success return value. That return contract does not say every existing field is immutable. The date-handling source is a concrete reason to distinguish an accepted update from proof that the complete earlier record stayed unchanged. [1]
For a migration or editorial review, the useful acceptance question is which properties were meant to change and which were meant to survive. Text, status, date interpretation and original provenance can each have their own expected outcome. A check concerned only with the visible paragraph leaves the other questions unanswered.
The same distinction matters when reconstructing a timeline later. A credible publication record identifies the event it observed and the evidence supporting that event. A source-supplied historical date may belong alongside it, clearly labelled, without replacing the actual record of first publication on the destination website.
A CMS date remains useful. Its meaning becomes clearer when the surrounding system explains whether it represents a floating draft, an explicit editorial date, an inherited source value or an observed publication event. The clock can be correct while the story attached to it is wrong; preserving that distinction makes the history worth trusting.
Sources + further reading
Source dates describe the original material. Publication above is the date this Zyel article went live.
- wp_update_post: draft dates, fields and return valuesWordPress Developer Resources · Checked Nov 18, 2026
- WordPress 7.1 branch: pinned core date-handling sourceWordPress core project · Checked Nov 18, 2026
- get_gmt_from_date: converting a site-local date to UTCWordPress Developer Resources · Checked Nov 18, 2026
- RFC 3339: timestamps and local offsetsIETF / RFC Editor · Checked Nov 18, 2026
Reader discussion
What would you ask next?
Questions, useful corrections and thoughtful disagreement are welcome. Comments are reviewed before they appear.
No reader comments yet.