Bug 228098

Summary:	Corruption in svn.freebsd.org?
Product:	Services	Reporter:	Trond Endrestøl <Trond.Endrestol>
Component:	Core Infrastructure	Assignee:	Cluster Admin <clusteradm>
Status:	Closed Works As Intended
Severity:	Affects Only Me	CC:	david
Priority:	---
Version:	unspecified
Hardware:	Any
OS:	Any

Description Trond Endrestøl 2018-05-09 15:02:35 UTC

While investigating other issues in stable/11, I noticed svnweb.freebsd.org says the current revision of base/stable/11/etc/sendmail/freebsd.mc is 302408.

Yet, a fresh checkout of base/stable/11 using https://svn.freebsd.org/base/stable/11 gives me:

$FreeBSD: stable/11/etc/sendmail/freebsd.mc 285230 2015-07-07 03:00:57Z gshapiro $

This is the same result I get from my own private mirror of the base repo, which also syncs off https://svn.freebsd.org/base.

In case it's of interest, my traffic originates from 128.39.0.0/16 and 2001:700::/32.

According to svnweb.freebsd.org, it should read:

$FreeBSD: stable/11/etc/sendmail/freebsd.mc 302408 2016-07-08 00:04:57Z gjb $

Why do we see this discrepancy?

Comment 1 david 2018-05-09 16:26:11 UTC

FWIW, I also see r285230 reflected in src/etc/sendmail/freebsd.mc -- e.g., on my laptop (updated via private mirror):

g1-215(11.2-P)[7] grep '\$FreeBSD' /usr/src/etc/sendmail/freebsd.mc
VERSIONID(`$FreeBSD: stable/11/etc/sendmail/freebsd.mc 285230 2015-07-07 03:00:57Z gshapiro $')
g1-215(11.2-P)[8] ls -lT !$
ls -lT /usr/src/etc/sendmail/freebsd.mc
-rw-r--r--  1 david  wheel  4533 Jul  8 04:02:18 2016 /usr/src/etc/sendmail/freebsd.mc
g1-215(11.2-P)[9] uname -a
FreeBSD g1-215.catwhisker.org 11.2-PRERELEASE FreeBSD 11.2-PRERELEASE #617  r333377M/333397:1101515: Wed May  9 03:39:52 PDT 2018     root@g1-215.catwhisker.org:/common/S1/obj/usr/src/sys/CANARY  amd64
g1-215(11.2-P)[10]

Comment 2 Trond Endrestøl 2018-05-09 19:26:34 UTC

(In reply to david from comment #1)
I think I overreacted. It's clearly a minor bug in Subversion. Since there were no changes to the sendmail files when stable/11 was created, it retains the original revision number even though the repo-copy resulted in a new revision. I think it's safe to close the matter.

Comment 3 Trond Endrestøl 2018-05-09 19:27:57 UTC

Works as Subversion intended it, at least in the version current at the time.

Comment 4 david 2018-05-09 19:28:32 UTC

OK; fair enough.