On Fri, 1 Jan 2021, Kevin A. McGrail wrote:
So I logged onto sa-vm and sudo'd to automc, when to svn/trunk and did svn
update in rules. See below. [1] Note the cron job shows: Checked out
revision 1885000. From looking at the script, it does a checkout of
rulesrc not rules so this might be "expected" behavior. NOT sure if things
were stale or if we use one revision for a week or something.
It should update rules/ every time as well, as those files are *not*
reliably static enough to be left alone for any length of time. They could
potentially change at any time, even though they generally haven't.
Is there any reason we should not be updating the entire trunk/ tree? Why
are we picking and choosing?
I think the run_nightly script in SVN should be updated to retrieve rules/
as well, or just all of trunk/ to avoid problems (e.g. with references to
modified plugins).
The URG_BIZ and the ADVANCED fee issues were something I saw in the crons
but they came in out of order and with no idea of the real dates so I was
waiting for the latest email with the output to check things.
Out of interest, did you make changes to those rules and possibly in 2
commits?
Nope. They were related so I kept them together in the same commit:
https://svn.apache.org/viewvc?view=revision&revision=1884468
Trying to figure out if something went wrong or not but my
eyesight is not good enough today to follow all the various cron jobs for
rules.
The __URG_BIZ change was the 15th, so a couple of weekly runs have
occurred since then and didn't automagically repair it.
I think that it's a design error in the script vs. something going wrong
in correct code.
Anyway, I ran the same command after the svn
up, ~/svn/trunk/build/mkupdates/run_nightly | /usr/bin/tee /var/www/
automc.spamassassin.org/mkupdates/mkupdates.txt and it passes now and it
published a ruleset that passes lint for me with 4.0.
Regards,
KAM
[1]
U 20_vbounce.cf
U 60_whitelist.cf
U 50_scores.cf
U 60_whitelist_auth.cf
U 20_phrases.cf
Updated to revision 1885008.
and in the root
U CREDITS
U rulesrc/sandbox/gbechis/20_freemail.cf
U rulesrc/sandbox/gbechis/20_misc.cf
U lib/Mail/SpamAssassin/Plugin/VBounce.pm
U lib/Mail/SpamAssassin/Plugin/DKIM.pm
U lib/Mail/SpamAssassin/Plugin/FreeMail.pm
U lib/Mail/SpamAssassin/Plugin/SPF.pm
U lib/Mail/SpamAssassin/Plugin/WLBLEval.pm
U lib/Mail/SpamAssassin/Conf.pm
A t/spf_welcome_block.t
A t/blocklist_autolearn.t
U t/data/01_test_rules.cf
A t/freemail_welcome_block.t
U MANIFEST
U NOTICE
Updated to revision 1885008.
[2]
automc@sa-vm:~/svn/trunk$ ~/svn/trunk/build/mkupdates/run_nightly |
/usr/bin/tee /var/www/automc.spamassassin.org/mkupdates/mkupdates.txt
{much snippage}
--
Kevin A. McGrail
Member, Apache Software Foundation
Chair Emeritus Apache SpamAssassin Project
https://www.linkedin.com/in/kmcgrail - 703.798.0171
On Fri, Jan 1, 2021 at 10:35 AM John Hardin <[email protected]> wrote:
On Fri, 1 Jan 2021, Kevin A. McGrail wrote:
Does anyone have some time to look into this error? It's why I wanted to
fix the server sending logs because I didn't think rules were being
published.
t/basic_lint.t .................. ok
t/basic_lint_without_sandbox.t .. ok
__ADVANCE_FEE_2_NEW depends on __URG_BIZ which is nonexistent
__URG_BIZ is defined in trunk/rules/20_phrases.cf and it's still there...
That test succeeds here in my scripted testing:
--- running basic_meta.t
1..2
ok 1
ok 2
All of my pre-commit lint checks are passing.
Here's the relevant part of a full "make test":
t/all_modules.t ................... ok
t/authres.t ....................... ok
t/autolearn.t ..................... ok
t/autolearn_force.t ............... ok
t/autolearn_force_fail.t .......... ok
t/basic_lint.t .................... ok
t/basic_lint_without_sandbox.t .... ok
t/basic_meta.t .................... ok
t/basic_meta2.t ................... ok
t/basic_obj_api.t ................. ok
Is trunk/rules/20_phrases.cf somehow stale in the masscheck SVN sandbox?
It should have been updated at the same time as the ADVANCE_FEE rules got
__URG_BIZ.
__ADVANCE_FEE_4_NEW depends on __URG_BIZ which is nonexistent
__ADVANCE_FEE_5_NEW depends on __URG_BIZ which is nonexistent
__ADVANCE_FEE_3_NEW depends on __URG_BIZ which is nonexistent
__ADVANCE_FEE_3_NEW depends on __URG_BIZ which is nonexistent
__ADVANCE_FEE_4_NEW depends on __URG_BIZ which is nonexistent
__ADVANCE_FEE_2_NEW depends on __URG_BIZ which is nonexistent
__ADVANCE_FEE_5_NEW depends on __URG_BIZ which is nonexistent
__ADVANCE_FEE_5_NEW depends on __URG_BIZ which is nonexistent
__ADVANCE_FEE_2_NEW depends on __URG_BIZ which is nonexistent
__ADVANCE_FEE_4_NEW depends on __URG_BIZ which is nonexistent
__ADVANCE_FEE_3_NEW depends on __URG_BIZ which is nonexistent
__ADVANCE_FEE_3_NEW depends on __URG_BIZ which is nonexistent
__ADVANCE_FEE_2_NEW depends on __URG_BIZ which is nonexistent
__ADVANCE_FEE_4_NEW depends on __URG_BIZ which is nonexistent
__ADVANCE_FEE_5_NEW depends on __URG_BIZ which is nonexistent
# Failed test at t/basic_meta.t line 93.
# Looks like you failed 1 test of 2.
t/basic_meta.t ..................
Dubious, test returned 1 (wstat 256, 0x100)
Failed 1/2 subtests
Test Summary Report
-------------------
t/basic_meta.t (Wstat: 256 Tests: 2 Failed: 1)
Failed test: 2
Non-zero exit status: 1
Files=3, Tests=6, 4 wallclock secs ( 0.02 usr 0.00 sys + 3.44 cusr
0.36
csys = 3.82 CPU)
Result: FAIL
Failed 1/3 test programs. 1/6 subtests failed.
make: *** [Makefile:1380: test_dynamic] Error 1
+ exit
--
John Hardin KA7OHZ http://www.impsec.org/~jhardin/
[email protected] pgpk -a [email protected]
key: 0xB8732E79 -- 2D8C 34F4 6411 F507 136C AF76 D822 E6E6 B873 2E79
-----------------------------------------------------------------------
This is a lame apocalypse. No zombies, and
the slow spiral to destruction has taken so long,
I think some people lost interest. -- STS
-----------------------------------------------------------------------
216 days since the first private commercial manned orbital mission (SpaceX)
