
Hmm. Barry, in current CVS, mailmanctl and qrunner both call LogStdErr with a parameter of tee_to_stdout. but in the library they include, that parameter is tee_to_real_stderr. Something's a bit out of sync in CVS
so does cron/gate-news
everything else seems okay, just those three files.
-- Chuq Von Rospach, Architech chuqui@plaidworks.com -- http://www.plaidworks.com/chuqui/blog/
The Cliff's Notes Cliff's Notes on Hamlet: And they all died happily ever after

"CVR" == Chuq Von Rospach <chuqui@plaidworks.com> writes:
CVR> Hmm. Barry, in current CVS, mailmanctl and qrunner both call
CVR> LogStdErr with a parameter of tee_to_stdout. but in the
CVR> library they include, that parameter is
CVR> tee_to_real_stderr. Something's a bit out of sync in CVS
CVR> so does cron/gate-news
CVR> everything else seems okay, just those three files.
But I think it's your cvs tree!
-------------------- snip snip -------------------- % find -type f | xargs grep tee_to ./Mailman/Logging/Utils.py:def LogStdErr(category, label, manual_reprime=1, tee_to_real_stderr=1): ./Mailman/Logging/Utils.py: If tee_to_real_stderr is true, then the real standard error also gets ./Mailman/Logging/Utils.py: if tee_to_real_stderr: -------------------- snip snip -------------------- % find -type f | xargs grep LogStdErr ./Mailman/Logging/Utils.py:def LogStdErr(category, label, manual_reprime=1, tee_to_real_stderr=1): ./bin/mailmanctl:from Mailman.Logging.Utils import LogStdErr ./bin/mailmanctl:LogStdErr('error', 'mailmanctl', manual_reprime=0) ./bin/qrunner:from Mailman.Logging.Utils import LogStdErr ./bin/qrunner:LogStdErr('error', 'qrunner', manual_reprime=0) ./contrib/auto:from Mailman.Logging.Utils import LogStdErr ./contrib/auto:LogStdErr('auto', 'auto') ./cron/gate_news:from Mailman.Logging.Utils import LogStdErr ./cron/gate_news:LogStdErr('error', 'gate_news', manual_reprime=0) ./scripts/bounces:from Mailman.Logging.Utils import LogStdErr ./scripts/bounces:LogStdErr('error', 'bounces') ./scripts/confirm:from Mailman.Logging.Utils import LogStdErr ./scripts/confirm:LogStdErr('error', 'confirm') ./scripts/join:from Mailman.Logging.Utils import LogStdErr ./scripts/join:LogStdErr('error', 'join') ./scripts/leave:from Mailman.Logging.Utils import LogStdErr ./scripts/leave:LogStdErr('error', 'leave') ./scripts/owner:from Mailman.Logging.Utils import LogStdErr ./scripts/owner:LogStdErr('error', 'mailowner') ./scripts/post:from Mailman.Logging.Utils import LogStdErr ./scripts/post:LogStdErr("error", "post") ./scripts/request:from Mailman.Logging.Utils import LogStdErr ./scripts/request:LogStdErr('error', 'mailcmd') [skipping the build subdir] -------------------- snip snip --------------------
What happens when you do "cvs -q up -P -d -A" ?
-Barry

well, that's bloody weird. CVS doesn't think so.
plaidworks.com 162# cvs -z3 -d:pserver:anonymous@cvs.mailman.sourceforge.net:/cvsroot/mailman update mailman cvs server: Updating mailman cvs server: Updating mailman/Archiver cvs server: Updating mailman/Bouncers cvs server: Updating mailman/Cgi cvs server: Updating mailman/Commands cvs server: Updating mailman/Gui cvs server: Updating mailman/Handlers cvs server: Updating mailman/Logging cvs server: Updating mailman/MTA cvs server: Updating mailman/Queue
here we go again, Barry. rmemeber when my cvs didn't udate the mail packages?
weirdo.
On Wednesday, October 30, 2002, at 10:50 PM, Barry A. Warsaw wrote:
-- Chuq Von Rospach, Architech chuqui@plaidworks.com -- http://www.plaidworks.com/chuqui/blog/
Someday, we'll look back on this, laugh nervously and change the subject.

"CVR" == Chuq Von Rospach <chuqui@plaidworks.com> writes:
CVR> well, that's bloody weird. CVS doesn't think so.
I just tried an anonymous cvs co on MacOS 10.2.1 and got the expected files. I thought maybe it could be a OSX thing, or maybe a pserver thing (remember that I always work in an authenticated checkout). But nope, I got the expected files.
"JCL" == J C Lawrence <claw@kanga.nu> writes:
JCL> <Elide rant on the gloriously broken state and quality of
JCL> CVS>
Sigh. Greg Wilson said at a Python conference many years ago (paraphrasing): cvs is the worst tool all of us use on a daily basis.
subversive-ly y'rs, -Barry

On Thu, 31 Oct 2002 09:04:04 -0500 Barry A Warsaw <barry@python.org> wrote:
JCL> <Elide rant on the gloriously broken state and quality of CVS>
Sigh. Greg Wilson said at a Python conference many years ago (paraphrasing): cvs is the worst tool all of us use on a daily basis.
Far be it from em to disagree.
subversive-ly y'rs, -Barry
I've been there. No thanks.
--
J C Lawrence
---------(*) Satan, oscillate my metallic sonatas.
claw@kanga.nu He lived as a devil, eh?
http://www.kanga.nu/~claw/ Evil is a name of a foeman, as I live.

On Thu, 31 Oct 2002, Dan Mick wrote:
It is expected that cvs without -A might not get everything. It is unexpected that -A won't solve that.
And just for the edification of the list, an explanation of the other two flags Barry suggested: "-P" says "give me new directories (and their contents) if any have been created." "-d" says "remove unmodified files (and resulting empty directories) that are no longer relevant."
Order should not matter, and I never tracked it down as far as I should, but let me mention that I've witnessed people having cvs problems when the flags are ordered "-P -d" instead of "-d -P"
from cvs --help update : -A Reset any sticky tags/date/kopts. -P Prune empty directories. -d Build directories, like checkout does.
-Dale

Sure. It finds stuff that's otherwise missing...
and my problem went away. Sigh. i'll fix my CVS commands. Thanks, all.
cvs -q up -P -d -A ? misc/email-2.3 ? misc/email-2.4.3 ? misc/JapaneseCodecs-1.4.8 ? misc/KoreanCodecs-2.0.5 P ACKNOWLEDGMENTS P NEWS P README.EXIM U README.MACOSX P TODO P UPGRADING P admin/www/MMGenerator.py P admin/www/download.ht P admin/www/download.html U admin/www/faq.ht P admin/www/faq.html P admin/www/i18n.ht P admin/www/i18n.html P admin/www/index.ht P admin/www/index.html P admin/www/mailman.html P admin/www/todo.ht P admin/www/todo.html P bin/add_members P bin/arch P bin/change_pw P bin/check_db P bin/check_perms P bin/cleanarch P bin/clone_member P bin/config_list P bin/dumpdb P bin/find_member P bin/genaliases P bin/inject P bin/list_admins P bin/list_lists P bin/list_members P bin/mailmanctl P bin/mmsitepass P bin/newlist P bin/pygettext.py P bin/qrunner P bin/remove_members P bin/rmlist P bin/sync_members P bin/transcheck P bin/unshunt P bin/update P bin/withlist P cron/bumpdigests P cron/checkdbs P cron/disabled P cron/gate_news P cron/mailpasswds P cron/nightly_gzip P cron/senddigests U doc/mailman-admin.tex P messages/mailman.pot U messages/cs/LC_MESSAGES/mailman.mo P messages/cs/LC_MESSAGES/mailman.po U messages/de/LC_MESSAGES/mailman.mo P messages/de/LC_MESSAGES/mailman.po U messages/es/LC_MESSAGES/mailman.mo P messages/es/LC_MESSAGES/mailman.po U messages/fi/LC_MESSAGES/mailman.mo P messages/fi/LC_MESSAGES/mailman.po U messages/fr/LC_MESSAGES/mailman.mo P messages/fr/LC_MESSAGES/mailman.po U messages/hu/LC_MESSAGES/mailman.mo P messages/hu/LC_MESSAGES/mailman.po U messages/it/LC_MESSAGES/mailman.mo P messages/it/LC_MESSAGES/mailman.po U messages/ja/LC_MESSAGES/mailman.mo P messages/ja/LC_MESSAGES/mailman.po U messages/ko/LC_MESSAGES/mailman.mo P messages/ko/LC_MESSAGES/mailman.po U messages/nl/LC_MESSAGES/mailman.mo P messages/nl/LC_MESSAGES/mailman.po U messages/no/LC_MESSAGES/mailman.mo P messages/no/LC_MESSAGES/mailman.po U messages/pt_BR/LC_MESSAGES/mailman.mo P messages/pt_BR/LC_MESSAGES/mailman.po U messages/ru/LC_MESSAGES/mailman.mo P messages/ru/LC_MESSAGES/mailman.po U messages/sv/LC_MESSAGES/mailman.mo P messages/sv/LC_MESSAGES/mailman.po U misc/email-2.4.3.tar.gz P misc/paths.py.in P src/common.h P templates/it/article.html P templates/it/postauth.txt P templates/ja/listinfo.html P templates/ja/options.html P templates/ja/roster.html U templates/no/archidxfoot.html U templates/no/archidxhead.html U templates/no/archliststart.html U templates/no/archtoc.html U templates/no/archtocentry.html P templates/no/article.html P templates/no/postack.txt P templates/ru/options.html P tests/test_bounces.py U tests/bounces/dsn_10.txt U tests/bounces/microsoft_01.txt U tests/bounces/microsoft_02.txt
# cvs -q up -d -P -A ? misc/email-2.3 ? misc/email-2.4.3 ? misc/JapaneseCodecs-1.4.8 ? misc/KoreanCodecs-2.0.5
On Thursday, October 31, 2002, at 12:14 PM, Dan Mick wrote:
Dale Newfield wrote:
What happens when you do "cvs -q up -P -d -A" ?
What he said; Chuq, can you soothe my mind about the panacea for CVS's brokenness, and try those flags?
-- Chuq Von Rospach, Architech chuqui@plaidworks.com -- http://www.plaidworks.com/chuqui/blog/
Stress is when you wake up screaming and you realize you haven't fallen asleep yet.

"bob" == <bob@nleaudio.com> writes:
bob> Perhaps the docs on the website should be updated to reflect
bob> these commands, so others don't have the same problems.
Thing is, -A usually /shouldn't/ be necessary unless folks are trying to checkout a specific branch, tag, or date. And unless you're doing maintenance on the 2.0 branch, or you want to see what 2.1b3 was like, you probably aren't doing that.
So I still want to know why Chuq finds it necessary to include the -A flag. Chuq, does your CVS directory have a Tag file in it?
-Barry

On Friday, November 1, 2002, at 06:46 AM, Barry A. Warsaw wrote:
So I still want to know why Chuq finds it necessary to include the -A flag. Chuq, does your CVS directory have a Tag file in it?
Nope. can't find one anywhere in the source structure.
-- Chuq Von Rospach, Architech chuqui@plaidworks.com -- http://www.plaidworks.com/chuqui/blog/
The Cliff's Notes Cliff's Notes on Hamlet: And they all died happily ever after

"CVR" == Chuq Von Rospach <chuqui@plaidworks.com> writes:
CVR> Hmm. Barry, in current CVS, mailmanctl and qrunner both call
CVR> LogStdErr with a parameter of tee_to_stdout. but in the
CVR> library they include, that parameter is
CVR> tee_to_real_stderr. Something's a bit out of sync in CVS
CVR> so does cron/gate-news
CVR> everything else seems okay, just those three files.
But I think it's your cvs tree!
-------------------- snip snip -------------------- % find -type f | xargs grep tee_to ./Mailman/Logging/Utils.py:def LogStdErr(category, label, manual_reprime=1, tee_to_real_stderr=1): ./Mailman/Logging/Utils.py: If tee_to_real_stderr is true, then the real standard error also gets ./Mailman/Logging/Utils.py: if tee_to_real_stderr: -------------------- snip snip -------------------- % find -type f | xargs grep LogStdErr ./Mailman/Logging/Utils.py:def LogStdErr(category, label, manual_reprime=1, tee_to_real_stderr=1): ./bin/mailmanctl:from Mailman.Logging.Utils import LogStdErr ./bin/mailmanctl:LogStdErr('error', 'mailmanctl', manual_reprime=0) ./bin/qrunner:from Mailman.Logging.Utils import LogStdErr ./bin/qrunner:LogStdErr('error', 'qrunner', manual_reprime=0) ./contrib/auto:from Mailman.Logging.Utils import LogStdErr ./contrib/auto:LogStdErr('auto', 'auto') ./cron/gate_news:from Mailman.Logging.Utils import LogStdErr ./cron/gate_news:LogStdErr('error', 'gate_news', manual_reprime=0) ./scripts/bounces:from Mailman.Logging.Utils import LogStdErr ./scripts/bounces:LogStdErr('error', 'bounces') ./scripts/confirm:from Mailman.Logging.Utils import LogStdErr ./scripts/confirm:LogStdErr('error', 'confirm') ./scripts/join:from Mailman.Logging.Utils import LogStdErr ./scripts/join:LogStdErr('error', 'join') ./scripts/leave:from Mailman.Logging.Utils import LogStdErr ./scripts/leave:LogStdErr('error', 'leave') ./scripts/owner:from Mailman.Logging.Utils import LogStdErr ./scripts/owner:LogStdErr('error', 'mailowner') ./scripts/post:from Mailman.Logging.Utils import LogStdErr ./scripts/post:LogStdErr("error", "post") ./scripts/request:from Mailman.Logging.Utils import LogStdErr ./scripts/request:LogStdErr('error', 'mailcmd') [skipping the build subdir] -------------------- snip snip --------------------
What happens when you do "cvs -q up -P -d -A" ?
-Barry

well, that's bloody weird. CVS doesn't think so.
plaidworks.com 162# cvs -z3 -d:pserver:anonymous@cvs.mailman.sourceforge.net:/cvsroot/mailman update mailman cvs server: Updating mailman cvs server: Updating mailman/Archiver cvs server: Updating mailman/Bouncers cvs server: Updating mailman/Cgi cvs server: Updating mailman/Commands cvs server: Updating mailman/Gui cvs server: Updating mailman/Handlers cvs server: Updating mailman/Logging cvs server: Updating mailman/MTA cvs server: Updating mailman/Queue
here we go again, Barry. rmemeber when my cvs didn't udate the mail packages?
weirdo.
On Wednesday, October 30, 2002, at 10:50 PM, Barry A. Warsaw wrote:
-- Chuq Von Rospach, Architech chuqui@plaidworks.com -- http://www.plaidworks.com/chuqui/blog/
Someday, we'll look back on this, laugh nervously and change the subject.

"CVR" == Chuq Von Rospach <chuqui@plaidworks.com> writes:
CVR> well, that's bloody weird. CVS doesn't think so.
I just tried an anonymous cvs co on MacOS 10.2.1 and got the expected files. I thought maybe it could be a OSX thing, or maybe a pserver thing (remember that I always work in an authenticated checkout). But nope, I got the expected files.
"JCL" == J C Lawrence <claw@kanga.nu> writes:
JCL> <Elide rant on the gloriously broken state and quality of
JCL> CVS>
Sigh. Greg Wilson said at a Python conference many years ago (paraphrasing): cvs is the worst tool all of us use on a daily basis.
subversive-ly y'rs, -Barry

On Thu, 31 Oct 2002 09:04:04 -0500 Barry A Warsaw <barry@python.org> wrote:
JCL> <Elide rant on the gloriously broken state and quality of CVS>
Sigh. Greg Wilson said at a Python conference many years ago (paraphrasing): cvs is the worst tool all of us use on a daily basis.
Far be it from em to disagree.
subversive-ly y'rs, -Barry
I've been there. No thanks.
--
J C Lawrence
---------(*) Satan, oscillate my metallic sonatas.
claw@kanga.nu He lived as a devil, eh?
http://www.kanga.nu/~claw/ Evil is a name of a foeman, as I live.

On Thu, 31 Oct 2002, Dan Mick wrote:
It is expected that cvs without -A might not get everything. It is unexpected that -A won't solve that.
And just for the edification of the list, an explanation of the other two flags Barry suggested: "-P" says "give me new directories (and their contents) if any have been created." "-d" says "remove unmodified files (and resulting empty directories) that are no longer relevant."
Order should not matter, and I never tracked it down as far as I should, but let me mention that I've witnessed people having cvs problems when the flags are ordered "-P -d" instead of "-d -P"
from cvs --help update : -A Reset any sticky tags/date/kopts. -P Prune empty directories. -d Build directories, like checkout does.
-Dale

Sure. It finds stuff that's otherwise missing...
and my problem went away. Sigh. i'll fix my CVS commands. Thanks, all.
cvs -q up -P -d -A ? misc/email-2.3 ? misc/email-2.4.3 ? misc/JapaneseCodecs-1.4.8 ? misc/KoreanCodecs-2.0.5 P ACKNOWLEDGMENTS P NEWS P README.EXIM U README.MACOSX P TODO P UPGRADING P admin/www/MMGenerator.py P admin/www/download.ht P admin/www/download.html U admin/www/faq.ht P admin/www/faq.html P admin/www/i18n.ht P admin/www/i18n.html P admin/www/index.ht P admin/www/index.html P admin/www/mailman.html P admin/www/todo.ht P admin/www/todo.html P bin/add_members P bin/arch P bin/change_pw P bin/check_db P bin/check_perms P bin/cleanarch P bin/clone_member P bin/config_list P bin/dumpdb P bin/find_member P bin/genaliases P bin/inject P bin/list_admins P bin/list_lists P bin/list_members P bin/mailmanctl P bin/mmsitepass P bin/newlist P bin/pygettext.py P bin/qrunner P bin/remove_members P bin/rmlist P bin/sync_members P bin/transcheck P bin/unshunt P bin/update P bin/withlist P cron/bumpdigests P cron/checkdbs P cron/disabled P cron/gate_news P cron/mailpasswds P cron/nightly_gzip P cron/senddigests U doc/mailman-admin.tex P messages/mailman.pot U messages/cs/LC_MESSAGES/mailman.mo P messages/cs/LC_MESSAGES/mailman.po U messages/de/LC_MESSAGES/mailman.mo P messages/de/LC_MESSAGES/mailman.po U messages/es/LC_MESSAGES/mailman.mo P messages/es/LC_MESSAGES/mailman.po U messages/fi/LC_MESSAGES/mailman.mo P messages/fi/LC_MESSAGES/mailman.po U messages/fr/LC_MESSAGES/mailman.mo P messages/fr/LC_MESSAGES/mailman.po U messages/hu/LC_MESSAGES/mailman.mo P messages/hu/LC_MESSAGES/mailman.po U messages/it/LC_MESSAGES/mailman.mo P messages/it/LC_MESSAGES/mailman.po U messages/ja/LC_MESSAGES/mailman.mo P messages/ja/LC_MESSAGES/mailman.po U messages/ko/LC_MESSAGES/mailman.mo P messages/ko/LC_MESSAGES/mailman.po U messages/nl/LC_MESSAGES/mailman.mo P messages/nl/LC_MESSAGES/mailman.po U messages/no/LC_MESSAGES/mailman.mo P messages/no/LC_MESSAGES/mailman.po U messages/pt_BR/LC_MESSAGES/mailman.mo P messages/pt_BR/LC_MESSAGES/mailman.po U messages/ru/LC_MESSAGES/mailman.mo P messages/ru/LC_MESSAGES/mailman.po U messages/sv/LC_MESSAGES/mailman.mo P messages/sv/LC_MESSAGES/mailman.po U misc/email-2.4.3.tar.gz P misc/paths.py.in P src/common.h P templates/it/article.html P templates/it/postauth.txt P templates/ja/listinfo.html P templates/ja/options.html P templates/ja/roster.html U templates/no/archidxfoot.html U templates/no/archidxhead.html U templates/no/archliststart.html U templates/no/archtoc.html U templates/no/archtocentry.html P templates/no/article.html P templates/no/postack.txt P templates/ru/options.html P tests/test_bounces.py U tests/bounces/dsn_10.txt U tests/bounces/microsoft_01.txt U tests/bounces/microsoft_02.txt
# cvs -q up -d -P -A ? misc/email-2.3 ? misc/email-2.4.3 ? misc/JapaneseCodecs-1.4.8 ? misc/KoreanCodecs-2.0.5
On Thursday, October 31, 2002, at 12:14 PM, Dan Mick wrote:
Dale Newfield wrote:
What happens when you do "cvs -q up -P -d -A" ?
What he said; Chuq, can you soothe my mind about the panacea for CVS's brokenness, and try those flags?
-- Chuq Von Rospach, Architech chuqui@plaidworks.com -- http://www.plaidworks.com/chuqui/blog/
Stress is when you wake up screaming and you realize you haven't fallen asleep yet.

"bob" == <bob@nleaudio.com> writes:
bob> Perhaps the docs on the website should be updated to reflect
bob> these commands, so others don't have the same problems.
Thing is, -A usually /shouldn't/ be necessary unless folks are trying to checkout a specific branch, tag, or date. And unless you're doing maintenance on the 2.0 branch, or you want to see what 2.1b3 was like, you probably aren't doing that.
So I still want to know why Chuq finds it necessary to include the -A flag. Chuq, does your CVS directory have a Tag file in it?
-Barry

On Friday, November 1, 2002, at 06:46 AM, Barry A. Warsaw wrote:
So I still want to know why Chuq finds it necessary to include the -A flag. Chuq, does your CVS directory have a Tag file in it?
Nope. can't find one anywhere in the source structure.
-- Chuq Von Rospach, Architech chuqui@plaidworks.com -- http://www.plaidworks.com/chuqui/blog/
The Cliff's Notes Cliff's Notes on Hamlet: And they all died happily ever after
participants (6)
-
barry@python.org
-
Bob Puff@NLE
-
Chuq Von Rospach
-
Dale Newfield
-
Dan Mick
-
J C Lawrence