Re: RFC round 3: DEP-3: Patch Tagging Guidelines
On Fri, 03 Jul 2009, Jon Dowland wrote:
> One thing I would like to see patch metadata help
> facilitate is patch review. At the moment the "Reviewed-by"
> header proposed would allow a tool to ensure at least two
> sets of eyeballs had seen a patch; however, patches can go
> stale. I think some chronological information is needed
> alongside the review. I propose a "Last-Reviewed" header to
> capture this information.
Only Sean Penny acked this suggestion. What do other people think of it?
> I decided on a separate header for this for two reasons
> * adding the date info to the Reviewed-by header will make
> it quite a lot longer
Is that a problem? In fact, if we ask people to review our patches I would
rather record their comments and not only the time of their review.
In that case, adding a timestamp in a multi-line comments is not a big
> * generally you only need to know the date of the last
> review, not the date of last review per reviewer.
Why? You might not trust all reviewers...
> Secondly, why not use RFC 2822 for the date field used in
> Last-Update (and my proposed Last-Reviewed)? It has a
> better granularity including timezone support, is output by
> GNU date(1) easily with the -R argument and is already in
> use for email Date: headers and Debian .changes files
> amongst others.
We don't need that granularity and it's somewhate cumbersome to have to
include the ouptput of date -R when most of the content is typed directly
by the maintainer.
Note that the date field is always populated by the program in all
examples you give. There's no such program for patch headers... so I'd
rather keep it very simple.
Contribuez à Debian et gagnez un cahier de l'admin Debian Lenny :