![](https://secure.gravatar.com/avatar/334b870d5b26878a79b2dc4cfcc500bc.jpg?s=120&d=mm&r=g)
Joshua Cranmer writes:
On 10/24/2011 8:04 PM, Barry Warsaw wrote:
On Oct 13, 2011, at 11:41 PM, Murray S. Kucherawy wrote:
There's movement afoot to deprecate use of "X-" in header field names. Just call it "Mailman-Topic". And if it's worthwhile, consider registering it with IANA.
I wonder if we should remove the X- prefixes for Mailman 3. Here's a list of ones we still add or recognize (some might be used only in the test suite):
I would say that anything that is used only in the test suite should still get an X-, although I suppose you could use Mailman-Test- too.
I believe the rule of thumb is you're supposed to use the X- prefix if it's not registered, so until the header is registered at IANA, I would vote that the X- prefix stays retained.
What Murray is saying is that the rule of thumb is changing in response to experience. What has happened is that the experience with promoting an X-Foo header to just Foo has been poor, and the attendant confusion often hinders adoption. So many people have been in the habit of ignoring the X- namespace anyway (the most widespread example I know of is the adoption of Mail-Followup-To in mail, which has no[1] sanction in the RFCs, although it's a long-standard header in news).
Footnotes: [1] Last I checked, anyway, a couple years ago, but widespread usage dates back to at least 2003.