On Mon, May 11, 2009 at 12:41:09AM -0700, Steve Langasek wrote:
> I am reopening this bug because I consider the new behavior a significant
> and very annoying regression.  The definition of the --release argument is
> "Finalize the changelog for a release."  Nothing in that implies that I
> should have to edit and save changes in order for the tool to take the
> action specified on the commandline,

The fact that you're presented with an editor should be implication
enough that you need to save your changes.  This is consistent with the
other invocations of dch that present the user with an editor.

Since dch makes changes based purely on heuristics, it is up to the user
to indicate back to dch whether they want to keep the changes or not.
The method of indicating such affirmation is saving the file.

> Please revert this change.

As pointed out on IRC, the next upload will allow one to set the
DEBCHANGE_FORCE_SAVE_ON_RELEASE config value such that dch will always
save the changes when calling `dch -r'.

-- 
James
GPG Key: 1024D/61326D40 2003-09-02 James Vega <[email protected]>

Attachment: signature.asc
Description: Digital signature

Reply via email to