Re: Losing Patches (was: embedded target breakage)

["Joseph S. Myers" <jsm28 at cam dot ac dot uk>]

On Sun, 1 Jul 2001, Gerald Pfeifer wrote:

>  o However, if a patch has not been reviewed/installed for, say, two
>    weeks, the submitter puts it into GNATS.

A related point: when putting a patch in GNATS, sending the URL to the
patch in the list archives should suffice (and means any subsequent
discussion on the lists can readily be reached from GNATS).  But we should
have a ready means of getting the plain text version of a patch (with tabs
intact, etc.) from the list archives.  That is, each list archive message
ought to have links to a plain text version of the original message, and a
plain text version but with MIME decoding done (to avoid problems with
quoted-printable, etc.).  Yes, this would double or triple the size of the
web list archives.

>  o As a release criterion for every major and minor release (that is,
>    GCC 3.0, 3.1, 3.2,...) we collectively try to process all such patches
>    that have been submitted before a given date (like one month before the
>    branch).

In principle I'd like something stronger: every PR, rather than just being
looked at to check whether it is a bug and whether it is a regression (and
some PRs never got looked at, or if they did then they didn't get marked
analyzed), gets investigated by someone to see if they can trace and fix
the bug (and, if they can't, gets a comment about where they got stuck).  
But as long as only a few of the hundred or so GCC developers do anything
much with GNATS, this isn't practical.

-- 
Joseph S. Myers
jsm28@cam.ac.uk