Обсуждение: Beta4 release announcement
Hi! I've hacked up a release announcement: http://wiki.postgresql.org/wiki/9.0beta4announcement Did I miss anything major? -selena -- http://chesnok.com/daily - me
Hi, 01.Ağu.2010 tarihinde 18:39 saatinde, Selena Deckelmann <selenamarie@gmail.co m> şunları yazdı: > Hi! > > I've hacked up a release announcement: > > http://wiki.postgresql.org/wiki/9.0beta4announcement > > Did I miss anything major? Looks good. Who is sending e-mail to -announce? -- Devrim GÜNDÜZ PostgreSQL DBA @ Akinon/Markafoni, Red Hat Certified Engineer devrim~gunduz.org, devrim~PostgreSQL.org, devrim.gunduz~linux.org.tr http://www.gunduz.org Twitter: http://twitter.com/devrimgunduz
Hi! 2010/8/1 Devrim GÜNDÜZ <devrim@gunduz.org>: > Who is sending e-mail to -announce? I can do it! -selena -- http://chesnok.com/daily - me
Ok, so the announce message was approved prematurely, after I sent an email asking folks not to. Ugh. Now what? 2010/8/1 Selena Deckelmann <selenamarie@gmail.com>: > Hi! > > 2010/8/1 Devrim GÜNDÜZ <devrim@gunduz.org>: > >> Who is sending e-mail to -announce? > > I can do it! > > -selena > > > -- > http://chesnok.com/daily - me > -- http://chesnok.com/daily - me
2010/8/1 Selena Deckelmann <selenamarie@gmail.com>: > Ok, so the announce message was approved prematurely, after I sent an > email asking folks not to. OK, so that answers my question I just emailed you :-p Who did you ask not to approve the message? I didn't see anything about that (not that I approved it, but I could have done). Another possibility... are you one of the folks we've whitelisted for sending to -announce because we like you so much? :-) -- Dave Page Blog: http://pgsnake.blogspot.com Twitter: @pgsnake EnterpriseDB UK: http://www.enterprisedb.com The Enterprise Postgres Company
2010/8/1 Dave Page <dpage@pgadmin.org>: > 2010/8/1 Selena Deckelmann <selenamarie@gmail.com>: >> Ok, so the announce message was approved prematurely, after I sent an >> email asking folks not to. > > OK, so that answers my question I just emailed you :-p > > Who did you ask not to approve the message? I didn't see anything > about that (not that I approved it, but I could have done). > > Another possibility... are you one of the folks we've whitelisted for > sending to -announce because we like you so much? :-) I got the stall message, so no, I am not whitelisted :) I sent the message asking for the message to not be approved to -www. Is there a better forum? -selena -- http://chesnok.com/daily - me
> I sent the message asking for the message to not be approved to -www.
> Is there a better forum?
Not really, and besides people approving messages to -announce should
know not to approve a release message without checking on it. I think
we need to find out who approved that and take them off approvals for
-announce permanently. What's the full list of people who have the
ability to approve -announce messages?
I'll add the release announcement to News for the home page. I believe
that the WWW approvers know not to approve it early.
-- -- Josh Berkus PostgreSQL Experts Inc.
http://www.pgexperts.com
Hey! On Sun, Aug 1, 2010 at 12:52 PM, Josh Berkus <josh@agliodbs.com> wrote: > >> I sent the message asking for the message to not be approved to -www. >> Is there a better forum? > > Not really, and besides people approving messages to -announce should know > not to approve a release message without checking on it. I think we need to > find out who approved that and take them off approvals for -announce > permanently. That seems a bit harsh. People do lots of worse things than mistakenly approve email, and don't get responsibility removed. Let's make the process for this a bit more clear. Where should the authoritative discussion be had about timing for release? Is it here? Or on -packagers, or somewhere else? -selena -- http://chesnok.com/daily - me
On 08/01/2010 08:39 AM, Selena Deckelmann wrote: > Hi! > > I've hacked up a release announcement: > > http://wiki.postgresql.org/wiki/9.0beta4announcement > > Did I miss anything major? This paragraph is confusing: Note that, due to a system catalog change, an initdb and database reload will be required for upgrading from 9.0Beta1. We encourage users to use this opportunity to test pg_upgrade for the upgrade from beta3 or an earlier version of 9.0. Please report your results. It should read instead: Unlike previous betas, there has been no system catalog change since Beta3, so an initdb should not be required to upgrade test databases from Beta3. Upgrading test databases from earlier versions or betas will require an initdb, however, so we urge users to take the opportunity to test upgrading their database using 9.0's pg_upgrade and report the results. ... as I believe we did *not* bump the system catalogs this time. -- -- Josh Berkus PostgreSQL Experts Inc. http://www.pgexperts.com
On Sun, Aug 1, 2010 at 12:58 PM, Josh Berkus <josh@agliodbs.com> wrote: > On 08/01/2010 08:39 AM, Selena Deckelmann wrote: >> >> Hi! >> >> I've hacked up a release announcement: >> >> http://wiki.postgresql.org/wiki/9.0beta4announcement >> >> Did I miss anything major? > > This paragraph is confusing: > > Note that, due to a system catalog change, an initdb and database reload > will be required for upgrading from 9.0Beta1. We encourage users to use this > opportunity to test pg_upgrade for the upgrade from beta3 or an earlier > version of 9.0. Please report your results. > > It should read instead: > > Unlike previous betas, there has been no system catalog change since Beta3, > so an initdb should not be required to upgrade test databases from Beta3. > Upgrading test databases from earlier versions or betas will require an > initdb, however, so we urge users to take the opportunity to test upgrading > their database using 9.0's pg_upgrade and report the results. > > ... as I believe we did *not* bump the system catalogs this time. Ok :) That text was in from the last release. Mea culpa. -selena -- http://chesnok.com/daily - me
On Sun, 2010-08-01 at 12:52 -0700, Josh Berkus wrote: > > I'll add the release announcement to News for the home page. I > believe that the WWW approvers know not to approve it early. :) and I'll also commit things to -www tomorrow. I'll also load the docs... -- Devrim GÜNDÜZ PostgreSQL Danışmanı/Consultant, Red Hat Certified Engineer PostgreSQL RPM Repository: http://yum.pgrpms.org Community: devrim~PostgreSQL.org, devrim.gunduz~linux.org.tr http://www.gunduz.org Twitter: http://twitter.com/devrimgunduz
Selena Deckelmann wrote: > I sent the message asking for the message to not be approved to -www. I'm with Dave here: I didn't see anything on pgsql-www related to this. I just checked the archives too; nothing I see matches. Think there were two problems here: that warning message didn't make it through for whatever reason, and someone moderating announce who should have known to check this sort of announcement more carefully goofed. -- Greg Smith 2ndQuadrant US Baltimore, MD PostgreSQL Training, Services and Support greg@2ndQuadrant.com www.2ndQuadrant.us
2010/8/1 Devrim GÜNDÜZ <devrim@gunduz.org>: > On Sun, 2010-08-01 at 12:52 -0700, Josh Berkus wrote: >> >> I'll add the release announcement to News for the home page. I >> believe that the WWW approvers know not to approve it early. > > :) > > and I'll also commit things to -www tomorrow. I'll also load the docs... Thanks Devrim - that'll be a big help! -- Dave Page Blog: http://pgsnake.blogspot.com Twitter: @pgsnake EnterpriseDB UK: http://www.enterprisedb.com The Enterprise Postgres Company
2010/8/1 Greg Smith <greg@2ndquadrant.com>: > Selena Deckelmann wrote: >> >> I sent the message asking for the message to not be approved to -www. > > I'm with Dave here: I didn't see anything on pgsql-www related to this. I > just checked the archives too; nothing I see matches. Think there were two > problems here: that warning message didn't make it through for whatever > reason, and someone moderating announce who should have known to check this > sort of announcement more carefully goofed. > > -- I've already mentioned this on the moderators list, but repeating here. The "binaries for other platforms" link isn't yet available. Is someone already working on that? Thom
On Mon, 2010-08-02 at 10:49 +0100, Thom Brown wrote: > The "binaries for other platforms" link isn't yet available. Is > someone already working on that? Guess who? :) -- Devrim GÜNDÜZ PostgreSQL Danışmanı/Consultant, Red Hat Certified Engineer PostgreSQL RPM Repository: http://yum.pgrpms.org Community: devrim~PostgreSQL.org, devrim.gunduz~linux.org.tr http://www.gunduz.org Twitter: http://twitter.com/devrimgunduz
On Sun, 2010-08-01 at 23:04 +0300, Devrim GÜNDÜZ wrote: > and I'll also commit things to -www tomorrow. I'll also load the > docs... Josh, website updated. You can send updated announcement. Regards, -- Devrim GÜNDÜZ PostgreSQL Danışmanı/Consultant, Red Hat Certified Engineer PostgreSQL RPM Repository: http://yum.pgrpms.org Community: devrim~PostgreSQL.org, devrim.gunduz~linux.org.tr http://www.gunduz.org Twitter: http://twitter.com/devrimgunduz
-----BEGIN PGP SIGNED MESSAGE----- Hash: RIPEMD160 Selena wrote: > Ok, so the announce message was approved prematurely, after > I sent an email asking folks not to. Still don't see any such message, which by tradition is sent well before any post to announce. > I sent the message asking for the message to not be approved > to -www. Is there a better forum? That's a good place, but another one idea is to send another message to - -announce itself. I see those from time to time from other people (e.g. "please don't approve the earlier message I just sent") Josh Berkus wrote: > Not really, and besides people approving messages to -announce should > know not to approve a release message without checking on it. I think > we need to find out who approved that and take them off approvals for > -announce permanently. Wow, overreact much? It's only a beta announcement. I approved it because it was in the queue and there was no "please don't approve this upcoming message" email sent. If you want to take away my announce "privs", go ahead. I guess as a core member you have the authoritah. If I can raise a bigger issue, why are we using the list as a CMS and not a mailing list? In other words, what is the point of sending things to the list that might *not* get approved? - -- Greg Sabino Mullane greg@turnstep.com PGP Key: 0x14964AC8 201008021104 http://biglumber.com/x/web?pk=2529DF6AB8F79407E94445B4BC9B906714964AC8 -----BEGIN PGP SIGNATURE----- iEYEAREDAAYFAkxW3xkACgkQvJuQZxSWSshbywCgu9QLkPpG2Vfz4EaiEltNOD+U bJAAnA4t/mao9xpB+F0tDvBwiCzt1C2B =i5wm -----END PGP SIGNATURE-----
On 08/02/2010 05:07 PM, Greg Sabino Mullane wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: RIPEMD160 > > > Selena wrote: >> Ok, so the announce message was approved prematurely, after >> I sent an email asking folks not to. > > Still don't see any such message, which by tradition is sent well > before any post to announce. > >> I sent the message asking for the message to not be approved >> to -www. Is there a better forum? > > That's a good place, but another one idea is to send another message to > - -announce itself. I see those from time to time from other > people (e.g. "please don't approve the earlier message I just sent") > > Josh Berkus wrote: > >> Not really, and besides people approving messages to -announce should >> know not to approve a release message without checking on it. I think >> we need to find out who approved that and take them off approvals for >> -announce permanently. > > Wow, overreact much? It's only a beta announcement. I approved it because > it was in the queue and there was no "please don't approve this upcoming message" > email sent. If you want to take away my announce "privs", go ahead. I guess > as a core member you have the authoritah. I agree - revoking moderation privileges for approving a valid message there with no "don't do that yet" attached on any list seems completely wrong(are we going to revoke commit provs from every commiter who ever commited something that has a bug?). I would actually say that it is good that we have active list moderators... Stefan
On Mon, 2010-08-02 at 15:07 +0000, Greg Sabino Mullane wrote: > Josh Berkus wrote: > > > Not really, and besides people approving messages to -announce should > > know not to approve a release message without checking on it. I think > > we need to find out who approved that and take them off approvals for > > -announce permanently. > This is bogus. You can not expect every moderator to know what is going on with timing of releases or if a message is ready to go. If the message wasn't ready to go out, it should have never been sent to announce in the first place. Joshua D. Drake -- PostgreSQL.org Major Contributor Command Prompt, Inc: http://www.commandprompt.com/ - 509.416.6579 Consulting, Training, Support, Custom Development, Engineering http://twitter.com/cmdpromptinc | http://identi.ca/commandprompt
On Aug 2, 2010, at 12:19 PM, Stefan Kaltenbrunner <stefan@kaltenbrunner.cc> wrote: > I agree - revoking moderation privileges for approving a valid message there with no "don't do that yet" attached on anylist seems completely wrong(are we going to revoke commit provs from every commiter who ever commited something that hasa bug?). Allow me to be the first to hope NOT. ISTM the proposed policy flies in the face of our often-lamented shortage of active moderators. Complain about moderatorswhen they don't moderate; punish them when they do. Go team! ...Robert
On Mon, Aug 2, 2010 at 9:50 AM, Robert Haas <robertmhaas@gmail.com> wrote: > On Aug 2, 2010, at 12:19 PM, Stefan Kaltenbrunner <stefan@kaltenbrunner.cc> wrote: >> I agree - revoking moderation privileges for approving a valid message there with no "don't do that yet" attached on anylist seems completely wrong(are we going to revoke commit provs from every commiter who ever commited something that hasa bug?). > > Allow me to be the first to hope NOT. > > ISTM the proposed policy flies in the face of our often-lamented shortage of active moderators. Complain about moderatorswhen they don't moderate; punish them when they do. Go team! Thank you, Robert. :) -selena -- http://chesnok.com/daily - me
On Mon, 2010-08-02 at 12:50 -0400, Robert Haas wrote: > On Aug 2, 2010, at 12:19 PM, Stefan Kaltenbrunner <stefan@kaltenbrunner.cc> wrote: > > I agree - revoking moderation privileges for approving a valid message there with no "don't do that yet" attached onany list seems completely wrong(are we going to revoke commit provs from every commiter who ever commited something thathas a bug?). > > Allow me to be the first to hope NOT. Well I will certainly second but frankly there is no need to. It isn't going to happen. It is unfortunate that tempers flared over such a simple thing. Perhaps we should instead be proud of the accomplishments of Beta4. Let us promote those instead of being pride full and making silly demands that stand to do nothing but deter would be contributors. Sincerely, Joshua D. Drake -- PostgreSQL.org Major Contributor Command Prompt, Inc: http://www.commandprompt.com/ - 509.416.6579 Consulting, Training, Support, Custom Development, Engineering http://twitter.com/cmdpromptinc | http://identi.ca/commandprompt
All, Procedure added to the news & announce policy page: http://wiki.postgresql.org/wiki/NewsEventsApproval#Procedure_for_Approving_pgsql-announce Hopefully a clear policy will help prevent similar issues in the future. -- -- Josh Berkus PostgreSQL Experts Inc. http://www.pgexperts.com
On Mon, 2010-08-02 at 10:13 -0700, Josh Berkus wrote: > All, > > Procedure added to the news & announce policy page: > > http://wiki.postgresql.org/wiki/NewsEventsApproval#Procedure_for_Approving_pgsql-announce > > Hopefully a clear policy will help prevent similar issues in the future. What does this mean? Consider if the announcement meets this policy and is appropriate for announcement * if necessary, check on pgsql-www list I do not agree that any moderator should have to check pgsql-www first. If it is appropriate, then it should be able to be approved without caveat. Joshua D. Drake -- PostgreSQL.org Major Contributor Command Prompt, Inc: http://www.commandprompt.com/ - 509.416.6579 Consulting, Training, Support, Custom Development, Engineering http://twitter.com/cmdpromptinc | http://identi.ca/commandprompt
Josh Berkus <josh@agliodbs.com> wrote: > Hopefully a clear policy will help prevent similar issues in the > future. Wasn't part of the problem that the post to -www didn't show up? Did anyone answer the question of why announcements are sent to the list before they should go out? -Kevin
"Kevin Grittner" <Kevin.Grittner@wicourts.gov> writes:
> Josh Berkus <josh@agliodbs.com> wrote:
>> Hopefully a clear policy will help prevent similar issues in the
>> future.
> Wasn't part of the problem that the post to -www didn't show up?
> Did anyone answer the question of why announcements are sent to the
> list before they should go out?
Yeah, I don't think that policy statement would have changed a thing.
The correct procedure for sending time-sensitive announcements is
to send them when they should be posted, not to imagine that the
approval queue is a delay mechanism that's under your control.
If you want you can then send something to pgsql-www to pester the
moderators to move quickly.
regards, tom lane
> Wasn't part of the problem that the post to -www didn't show up?
> Did anyone answer the question of why announcements are sent to the
> list before they should go out?
Oh, that was a mistake; normally they don't. However, one of the things
we count on human list moderators for is to catch mistakes ... duplicate
postings, cross-postings, links that don't work in commercial
announcements, inappropriate content, etc.
I imagine that Greg simply assumed that if Selena was posting it, the
message must be ready to go out. And since her message to this list
*not* to post it didn't come though ...
It's kind of like a flight check: normally we have a procedure (when/how
to post to pgsql-announce, which could use some explication on the wiki,
as this has shown) and a safety mechanism (the human moderator of the
pgsql-announce). In this case both the procedure and the safety
mechanism failed, fortunately without horrible consequences.
So I'd like to improve both before an error does have horrible
consequences, such as a security release. Working on the procedure now.
-- -- Josh Berkus PostgreSQL Experts Inc.
http://www.pgexperts.com
Greg, > Wow, overreact much? It's only a beta announcement. BTW, my apologies for the tone of my original message. I do strongly feel that moderators of -announce should exercise far more caution in approving things even if it means delays. The way I originally expressed this was not helpful. I can completely understand you automatically assuming that if Selena is posting something, it must be both intentional and approved. > If you want to take away my announce "privs", go ahead. I guess > as a core member you have the authoritah. I think you overestimate my powers. ;-) -- -- Josh Berkus
On Mon, Aug 2, 2010 at 3:16 PM, Josh Berkus <josh@agliodbs.com> wrote: >> Wow, overreact much? It's only a beta announcement. > > BTW, my apologies for the tone of my original message. I do strongly > feel that moderators of -announce should exercise far more caution in > approving things even if it means delays. The way I originally expressed > this was not helpful. I can completely understand you automatically > assuming that if Selena is posting something, it must be both > intentional and approved. I'm lost. Can someone explain to me in short words what was so wrong with Selena's announcement? Even though one paragraph of text was from the previous release, I don't think anything in there was actually untrue. At most, it was slightly misleading or not quite up to date. I don't even understand why we bothered to send a correction. It's all fine and good to say that moderators on -announce should exercise caution, but what are they supposed to be cautious ABOUT? A minor wording problem with a release announcement that requires a microscope to detect? -- Robert Haas EnterpriseDB: http://www.enterprisedb.com The Enterprise Postgres Company
Robert Haas <robertmhaas@gmail.com> wrote: > I'm lost. Can someone explain to me in short words what was so > wrong with Selena's announcement? It was sent one day too soon? -Kevin
On Mon, Aug 2, 2010 at 4:20 PM, Kevin Grittner <Kevin.Grittner@wicourts.gov> wrote: > Robert Haas <robertmhaas@gmail.com> wrote: >> I'm lost. Can someone explain to me in short words what was so >> wrong with Selena's announcement? > > It was sent one day too soon? Hmm, OK. But the correction didn't really make that clear, did it? -- Robert Haas EnterpriseDB: http://www.enterprisedb.com The Enterprise Postgres Company
On 8/2/10 1:17 PM, Robert Haas wrote:
> I don't even understand why we bothered to send a
> correction. It's all fine and good to say that moderators on
> -announce should exercise caution, but what are they supposed to be
> cautious ABOUT? A minor wording problem with a release announcement
> that requires a microscope to detect?
The issue was that it went out when Beta4 wasn't yet available and the
home page wasn't updated. Otherwise it would have been fine, and there
would have been no 2nd announcement.
-- -- Josh Berkus PostgreSQL Experts Inc.
http://www.pgexperts.com
> Hmm, OK. But the correction didn't really make that clear, did it?
Possibly not, but I'm NOT sending a correction to the correction. ;-b
-- -- Josh Berkus PostgreSQL Experts Inc.
http://www.pgexperts.com
Josh Berkus <josh@agliodbs.com> wrote: >> Hmm, OK. But the correction didn't really make that clear, did >> it? > > Possibly not, but I'm NOT sending a correction to the correction. > ;-b I thought the new announcement did an admirable job of dancing around the issue: | (Note: This is a corrected version of yesterday's announcement | which went out prematurely) If I had tried and failed to find the download, reading this would make me think the announcement was "premature" because it went out before the beta was ready. If I hadn't tried to find the download, I might think it was premature because the language wasn't quite settled yet. Maximal coverage with minimal embarrassment. Brilliant! -Kevin
On Mon, 2010-08-02 at 15:07 +0000, Greg Sabino Mullane wrote: > Josh Berkus wrote: > > > Not really, and besides people approving messages to -announce should > > know not to approve a release message without checking on it. I think > > we need to find out who approved that and take them off approvals for > > -announce permanently. > This is bogus. You can not expect every moderator to know what is going on with timing of releases or if a message is ready to go. If the message wasn't ready to go out, it should have never been sent to announce in the first place. Joshua D. Drake -- PostgreSQL.org Major Contributor Command Prompt, Inc: http://www.commandprompt.com/ - 509.416.6579 Consulting, Training, Support, Custom Development, Engineering http://twitter.com/cmdpromptinc | http://identi.ca/commandprompt
On Mon, 2010-08-02 at 12:50 -0400, Robert Haas wrote: > On Aug 2, 2010, at 12:19 PM, Stefan Kaltenbrunner <stefan@kaltenbrunner.cc> wrote: > > I agree - revoking moderation privileges for approving a valid message there with no "don't do that yet" attached onany list seems completely wrong(are we going to revoke commit provs from every commiter who ever commited something thathas a bug?). > > Allow me to be the first to hope NOT. Well I will certainly second but frankly there is no need to. It isn't going to happen. It is unfortunate that tempers flared over such a simple thing. Perhaps we should instead be proud of the accomplishments of Beta4. Let us promote those instead of being pride full and making silly demands that stand to do nothing but deter would be contributors. Sincerely, Joshua D. Drake -- PostgreSQL.org Major Contributor Command Prompt, Inc: http://www.commandprompt.com/ - 509.416.6579 Consulting, Training, Support, Custom Development, Engineering http://twitter.com/cmdpromptinc | http://identi.ca/commandprompt
On Mon, 2010-08-02 at 10:13 -0700, Josh Berkus wrote: > All, > > Procedure added to the news & announce policy page: > > http://wiki.postgresql.org/wiki/NewsEventsApproval#Procedure_for_Approving_pgsql-announce > > Hopefully a clear policy will help prevent similar issues in the future. What does this mean? Consider if the announcement meets this policy and is appropriate for announcement * if necessary, check on pgsql-www list I do not agree that any moderator should have to check pgsql-www first. If it is appropriate, then it should be able to be approved without caveat. Joshua D. Drake -- PostgreSQL.org Major Contributor Command Prompt, Inc: http://www.commandprompt.com/ - 509.416.6579 Consulting, Training, Support, Custom Development, Engineering http://twitter.com/cmdpromptinc | http://identi.ca/commandprompt
-----BEGIN PGP SIGNED MESSAGE----- Hash: RIPEMD160 >> Did anyone answer the question of why announcements are sent to the >> list before they should go out? > Oh, that was a mistake; normally they don't. Is that true? I do seem to recall someone once advocating the idea of "queuing" messages that would get approved later. Can we officially can that idea here and now, and use one of the many other venues for storing drafts, timed messages, and the like? - -- Greg Sabino Mullane greg@turnstep.com PGP Key: 0x14964AC8 201008041016 http://biglumber.com/x/web?pk=2529DF6AB8F79407E94445B4BC9B906714964AC8 -----BEGIN PGP SIGNATURE----- iEYEAREDAAYFAkxZdi8ACgkQvJuQZxSWSsiifgCaAq8mF6eV2wMo6Qk0/JUBrWTQ 184AoIaU9YKoGfVkfxQvNMTLhAcIVO+y =fSnL -----END PGP SIGNATURE-----