From danstoner at gmail.com Mon Jul 2 21:09:01 2012 From: danstoner at gmail.com (Dan Stoner) Date: Mon, 2 Jul 2012 21:09:01 -0400 Subject: [geeklog-devel] postgresql and Geeklog 2.0.0b1 Message-ID: I just got back from a tech conference and I'm a little fired up about open source again. I allocated some time to test Geeklog with postgresql. Here is my environment (currently a RHEL 5.x clone): PHP 5.3.3 PostgreSQL(libpq) Version 8.1.23 The first two issues might be worth documenting in the release notes or something: 1. The default configuration in this packaged version of postgres creates databases with UTF8 encoding, so I checked the box for this during the Geeklog install. 2. Full privileges on the database itself do not seem to be sufficient. Without granting "Superuser" role to my postgres web application user account, the install script aborts with this message: Warning: pg_query(): Query failed: ERROR: must be superuser to create procedural language CONTEXT: SQL function "make_plpgsql" statement 1 in /var/www/private/pgblog_private/system/databases/pgsql.class.php on line 350 The Superuser role can be granted by issuing the following: ALTER USER geekloguser WITH superuser; If this is only needed for the initial database config / creation it might be worth making a note that the superuser role can be removed after the initial install. ALTER USER geekloguser WITH nosuperuser; The third and fourth issues preventing me from using the site: 3. The admin account with default password of "password" did not seem to exist or let me log in. 4. The main page seemed to render ok, as do some of the sub-pages, but many of them error out with "Unfortunately, an error has occurred rendering this page. Please try again later." Looking in my geeklog error log, I see: Mon 02 Jul 2012 20:51:34 EDT - 98.70.54.73 - 2 - pg_query(): Query failed: ERROR: column "e707dc1196bbd50f2cc5e1f51791c3e2d8862a75" does not exist @ /var/www/private/pgblog_private/system/databases/pgsql.class.php line 350 Mon 02 Jul 2012 20:51:48 EDT - 98.70.54.73 - Error, invalid username: 'admin' Mon 02 Jul 2012 20:51:55 EDT - 98.70.54.73 - 2 - pg_query(): Query failed: ERROR: column "034f3c84e263ad32a1523e0a6323f4f0e6ced60e" does not exist @ /var/www/private/pgblog_private/system/databases/pgsql.class.php line 350 Mon 02 Jul 2012 20:52:03 EDT - 98.70.54.73 - 2 - pg_query(): Query failed: ERROR: column "gl_stories.title" must appear in the GROUP BY clause or be used in an aggregate function @ /var/www/private/pgblog_private/system/databases/pgsql.class.php line 350 Mon 02 Jul 2012 20:52:14 EDT - 98.70.54.73 - 2 - pg_query(): Query failed: ERROR: column "gl_stories.date" must appear in the GROUP BY clause or be used in an aggregate function @ /var/www/private/pgblog_private/system/databases/pgsql.class.php line 350 Mon 02 Jul 2012 20:52:34 EDT - 98.70.54.73 - 2 - pg_query(): Query failed: ERROR: column "gl_stories.title" must appear in the GROUP BY clause or be used in an aggregate function @ /var/www/private/pgblog_private/system/databases/pgsql.class.php line 350 Mon 02 Jul 2012 20:55:33 EDT - 98.70.54.73 - Error, invalid username: 'admin' Mon 02 Jul 2012 20:55:40 EDT - 98.70.54.73 - 2 - pg_query(): Query failed: ERROR: column "67b8bf5150a006c12953adf82b28e7bbe2477e21" does not exist @ /var/www/private/pgblog_private/system/databases/pgsql.class.php line 350 Let me know what I can do to help. Regards, - Dan Stoner From yankees26an at gmail.com Mon Jul 2 23:51:04 2012 From: yankees26an at gmail.com (Stanislav Palatnik) Date: Mon, 2 Jul 2012 23:51:04 -0400 Subject: [geeklog-devel] postgresql and Geeklog 2.0.0b1 In-Reply-To: References: Message-ID: Hello Dan, Thanks for bringing this up. Can you verify if the admin account was added to PostgreSQL. There should be an entry for a username "Admin" in the "gl_users" table(if you used the default database name). I think the problem has something to do with the gl_topic_assignments table. I'm not sure why its id is a varchar(40). On Mon, Jul 2, 2012 at 9:09 PM, Dan Stoner wrote: > I just got back from a tech conference and I'm a little fired up about > open source again. I allocated some time to test Geeklog with > postgresql. > > Here is my environment (currently a RHEL 5.x clone): > > PHP 5.3.3 > PostgreSQL(libpq) Version 8.1.23 > > > The first two issues might be worth documenting in the release notes > or something: > > > 1. The default configuration in this packaged version of postgres > creates databases with UTF8 encoding, so I checked the box for this > during the Geeklog install. > > 2. Full privileges on the database itself do not seem to be > sufficient. Without granting "Superuser" role to my postgres web > application user account, the install script aborts with this message: > > > Warning: pg_query(): Query failed: ERROR: must be superuser to create > procedural language CONTEXT: SQL function "make_plpgsql" statement 1 > in /var/www/private/pgblog_private/system/databases/pgsql.class.php on > line 350 > > > The Superuser role can be granted by issuing the following: > > ALTER USER geekloguser WITH superuser; > > If this is only needed for the initial database config / creation it > might be worth making a note that the superuser role can be removed > after the initial install. > > ALTER USER geekloguser WITH nosuperuser; > > > > The third and fourth issues preventing me from using the site: > > 3. The admin account with default password of "password" did not seem > to exist or let me log in. > > 4. The main page seemed to render ok, as do some of the sub-pages, > but many of them error out with "Unfortunately, an error has occurred > rendering this page. Please try again later." > > Looking in my geeklog error log, I see: > > > Mon 02 Jul 2012 20:51:34 EDT - 98.70.54.73 - 2 - pg_query(): Query > failed: ERROR: column "e707dc1196bbd50f2cc5e1f51791c3e2d8862a75" does > not exist @ > /var/www/private/pgblog_private/system/databases/pgsql.class.php > line 350 > Mon 02 Jul 2012 20:51:48 EDT - 98.70.54.73 - Error, invalid username: > 'admin' > Mon 02 Jul 2012 20:51:55 EDT - 98.70.54.73 - 2 - pg_query(): Query > failed: ERROR: column "034f3c84e263ad32a1523e0a6323f4f0e6ced60e" does > not exist @ > /var/www/private/pgblog_private/system/databases/pgsql.class.php > line 350 > Mon 02 Jul 2012 20:52:03 EDT - 98.70.54.73 - 2 - pg_query(): Query > failed: ERROR: column "gl_stories.title" must appear in the GROUP BY > clause or be used in an aggregate function @ > /var/www/private/pgblog_private/system/databases/pgsql.class.php line > 350 > Mon 02 Jul 2012 20:52:14 EDT - 98.70.54.73 - 2 - pg_query(): Query > failed: ERROR: column "gl_stories.date" must appear in the GROUP BY > clause or be used in an aggregate function @ > /var/www/private/pgblog_private/system/databases/pgsql.class.php line > 350 > Mon 02 Jul 2012 20:52:34 EDT - 98.70.54.73 - 2 - pg_query(): Query > failed: ERROR: column "gl_stories.title" must appear in the GROUP BY > clause or be used in an aggregate function @ > /var/www/private/pgblog_private/system/databases/pgsql.class.php line > 350 > Mon 02 Jul 2012 20:55:33 EDT - 98.70.54.73 - Error, invalid username: > 'admin' > Mon 02 Jul 2012 20:55:40 EDT - 98.70.54.73 - 2 - pg_query(): Query > failed: ERROR: column "67b8bf5150a006c12953adf82b28e7bbe2477e21" does > not exist @ > /var/www/private/pgblog_private/system/databases/pgsql.class.php > line 350 > > > > Let me know what I can do to help. > > Regards, > > - Dan Stoner > _______________________________________________ > geeklog-devel mailing list > geeklog-devel at lists.geeklog.net > http://eight.pairlist.net/mailman/listinfo/geeklog-devel > > -- > -- > My Best, > > Stanislav Palatnik > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From dirk at haun-online.de Tue Jul 3 04:34:22 2012 From: dirk at haun-online.de (Dirk Haun) Date: Tue, 03 Jul 2012 10:34:22 +0200 Subject: [geeklog-devel] postgresql and Geeklog 2.0.0b1 In-Reply-To: References: Message-ID: <20120703103422.Horde.ISurBsL8999P8q6ObIQDfMA@webmail.df.eu> Quoting Dan Stoner : > I just got back from a tech conference and I'm a little fired up about > open source again. I allocated some time to test Geeklog with > postgresql. Thanks for doing that. Please note that Postgres support in 2.0.0b1 is very buggy and barely working. I made a few tweaks since then, so if you have the time, try a nightly tarball (or wait for b2). > 3. The admin account with default password of "password" did not seem > to exist or let me log in. Works for me (now - I think this was one of the post-b1 fixes that I had to make). > 4. The main page seemed to render ok, as do some of the sub-pages, > but many of them error out with "Unfortunately, an error has occurred > rendering this page. Please try again later." You may want to enable root debugging (in siteconfig.php). It'll not only print the error message directly in the browser window, but also displays a call stack so that you can see where the call is coming from (an error in pgsql.class.php isn't all that helpful - the faulty SQL was usually created elsewhere). > Looking in my geeklog error log, I see: (snip) These look familiar and should hopefully be fixed now. > Let me know what I can do to help. See above. Please note that I'm testing locally with Postgres 9.1, so it'll be interesting to see if there are other problems with 8.1. bye, Dirk From yankees26an at gmail.com Tue Jul 3 15:01:29 2012 From: yankees26an at gmail.com (Stanislav Palatnik) Date: Tue, 3 Jul 2012 15:01:29 -0400 Subject: [geeklog-devel] postgresql and Geeklog 2.0.0b1 In-Reply-To: <20120703103422.Horde.ISurBsL8999P8q6ObIQDfMA@webmail.df.eu> References: <20120703103422.Horde.ISurBsL8999P8q6ObIQDfMA@webmail.df.eu> Message-ID: > > The Superuser role can be granted by issuing the following: > ALTER USER geekloguser WITH superuser; I tried implementing that, and ran into an issue. Query failed: ERROR: must be superuser to alter superusers You must be a superuser to add another account as a superuser. I think you ran that command from pgadmin, which uses the default superuser account(postgres), so it let you run this command. However, most people installing geeklog will be on a shared hosting environment and will not have access to this. However, I ran installs on Windows(7 and 8), and Ubuntu(11), and PostgreSQL did not complain about needing superuser access to install the language. Additionally, there haven't been any reports of users running into this problem until now. Can you give more information about your PostgreSQL configuration? It seems to be running on paranoid mode :) On Tue, Jul 3, 2012 at 4:34 AM, Dirk Haun wrote: > Quoting Dan Stoner : > > I just got back from a tech conference and I'm a little fired up about >> open source again. I allocated some time to test Geeklog with >> postgresql. >> > > Thanks for doing that. > > Please note that Postgres support in 2.0.0b1 is very buggy and barely > working. I made a few tweaks since then, so if you have the time, try a > nightly tarball (or wait for b2). > > > > 3. The admin account with default password of "password" did not seem >> to exist or let me log in. >> > > Works for me (now - I think this was one of the post-b1 fixes that I had > to make). > > > > 4. The main page seemed to render ok, as do some of the sub-pages, >> but many of them error out with "Unfortunately, an error has occurred >> rendering this page. Please try again later." >> > > You may want to enable root debugging (in siteconfig.php). It'll not only > print the error message directly in the browser window, but also displays a > call stack so that you can see where the call is coming from (an error in > pgsql.class.php isn't all that helpful - the faulty SQL was usually created > elsewhere). > > > > Looking in my geeklog error log, I see: >> > (snip) > These look familiar and should hopefully be fixed now. > > > > Let me know what I can do to help. >> > > See above. Please note that I'm testing locally with Postgres 9.1, so > it'll be interesting to see if there are other problems with 8.1. > > bye, Dirk > > > > ______________________________**_________________ > geeklog-devel mailing list > geeklog-devel at lists.geeklog.**net > http://eight.pairlist.net/**mailman/listinfo/geeklog-devel > > -- > -- > My Best, > > Stanislav Palatnik > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From danstoner at gmail.com Wed Jul 4 11:09:00 2012 From: danstoner at gmail.com (Dan Stoner) Date: Wed, 4 Jul 2012 11:09:00 -0400 Subject: [geeklog-devel] postgresql and Geeklog 2.0.0b1 In-Reply-To: <20120703103422.Horde.ISurBsL8999P8q6ObIQDfMA@webmail.df.eu> References: <20120703103422.Horde.ISurBsL8999P8q6ObIQDfMA@webmail.df.eu> Message-ID: I tried the nightly tarball with root debugging enabled. I had a few issues related to filesystem permissions (not postgres), once those were fixed things got better. I was able to log in with Admin and default password. After clicking around a bit, most of the functionality seems to be ok (the admin panels are accessible). However, the main index page fails with the following: An error has occurred: This is being displayed as "Root Debugging" is enabled in your Geeklog configuration. If this is a production website you must disable this option once you have resolved any issues you are investigating. 2 - pg_query(): Query failed: ERROR: column "gl_stories.date" must appear in the GROUP BY clause or be used in an aggregate function @ /var/www/private/pgblog_private/system/databases/pgsql.class.php line 350 Call Stack # Function File Line 1 pg_query /var/www/private/pgblog_private/system/databases/pgsql.class.php 350 2 dbQuery /var/www/private/pgblog_private/system/lib-database.php 213 3 DB_query /var/www/private/pgblog_private/system/lib-syndication.php 99 4 SYND_feedUpdateCheckAll /var/www/private/pgblog_private/system/lib-syndication.php 210 5 SYND_feedUpdateCheck /var/www/sites/thatlinuxbox.com/pgblog/lib-common.php 2662 6 COM_rdfUpToDateCheck /var/www/sites/thatlinuxbox.com/pgblog/index.php 154 array(3) { ["sql"]=> string(229) "SELECT sid FROM gl_stories, gl_topic_assignments ta WHERE draft_flag = 0 AND date <= NOW() AND ta.type = 'article' AND ta.id = sid AND (ta.tid IN ('General','Geeklog')) AND perm_anon > 0 GROUP BY sid ORDER BY date DESC LIMIT 10" ["ignore_errors"]=> int(0) ["matches"]=> array(0) { } } In this case I didn't drop my database, so the install script did an Upgrade, so I didn't get any issue about Superuser role. However, it seems that the install script is adding the PLPGSQL stored procedure language to the database (to the schema?), which does require superuser role. I guess some people's databases are created with that language by default, in my stock config it is not. Regards, - Dan Stoner On Tue, Jul 3, 2012 at 4:34 AM, Dirk Haun wrote: > Quoting Dan Stoner : > >> I just got back from a tech conference and I'm a little fired up about >> open source again. I allocated some time to test Geeklog with >> postgresql. > > > Thanks for doing that. > > Please note that Postgres support in 2.0.0b1 is very buggy and barely > working. I made a few tweaks since then, so if you have the time, try a > nightly tarball (or wait for b2). > > > >> 3. The admin account with default password of "password" did not seem >> to exist or let me log in. > > > Works for me (now - I think this was one of the post-b1 fixes that I had to > make). > > > >> 4. The main page seemed to render ok, as do some of the sub-pages, >> but many of them error out with "Unfortunately, an error has occurred >> rendering this page. Please try again later." > > > You may want to enable root debugging (in siteconfig.php). It'll not only > print the error message directly in the browser window, but also displays a > call stack so that you can see where the call is coming from (an error in > pgsql.class.php isn't all that helpful - the faulty SQL was usually created > elsewhere). > > > >> Looking in my geeklog error log, I see: > > (snip) > These look familiar and should hopefully be fixed now. > > > >> Let me know what I can do to help. > > > See above. Please note that I'm testing locally with Postgres 9.1, so it'll > be interesting to see if there are other problems with 8.1. > > bye, Dirk > > > > _______________________________________________ > geeklog-devel mailing list > geeklog-devel at lists.geeklog.net > http://eight.pairlist.net/mailman/listinfo/geeklog-devel From danstoner at gmail.com Thu Jul 5 09:54:11 2012 From: danstoner at gmail.com (Dan Stoner) Date: Thu, 5 Jul 2012 09:54:11 -0400 Subject: [geeklog-devel] postgresql and Geeklog 2.0.0b1 In-Reply-To: References: <20120703103422.Horde.ISurBsL8999P8q6ObIQDfMA@webmail.df.eu> Message-ID: I thought about this some more and poked a bit. Newly created databases in PostgreSQL inherit from the template1 database. On my server, template1 does not include the plpgsql language (this has been the default on all postgres database servers I have ever administered). Maybe most hosting environments do add that language to the template database, so most users would not see this issue. The two likely alternatives to meeting this requirement and allow the install to go through are: 1. When the geeklog postgres database is created, also add the plpgsql language with: CREATE LANGUAGE plpgsql; 2. Before creating a new geeklog database, make sure that the template1 database contains the plpgsql language. Same command as above, just on the template1 database. All future databases that are created would have this language available. Here's some SQL... template1=# select * from pg_language where lanname='plpgsql'; lanname | lanispl | lanpltrusted | lanplcallfoid | lanvalidator | lanacl ---------+---------+--------------+---------------+--------------+-------- (0 rows) template1=# CREATE LANGUAGE plpgsql; CREATE LANGUAGE template1=# select * from pg_language where lanname='plpgsql'; lanname | lanispl | lanpltrusted | lanplcallfoid | lanvalidator | lanacl ---------+---------+--------------+---------------+--------------+-------- plpgsql | t | t | 17625 | 17626 | (1 row) Regards, - Dan Stoner On Wed, Jul 4, 2012 at 11:09 AM, Dan Stoner wrote: > I tried the nightly tarball with root debugging enabled. > > I had a few issues related to filesystem permissions (not postgres), > once those were fixed things got better. > > I was able to log in with Admin and default password. After clicking > around a bit, most of the functionality seems to be ok (the admin > panels are accessible). > > > However, the main index page fails with the following: > > An error has occurred: > > This is being displayed as "Root Debugging" is enabled in your Geeklog > configuration. > > If this is a production website you must disable this option once you > have resolved any issues you are investigating. > > 2 - pg_query(): Query failed: ERROR: column "gl_stories.date" must > appear in the GROUP BY clause or be used in an aggregate function @ > /var/www/private/pgblog_private/system/databases/pgsql.class.php line > 350 > > Call Stack > # Function File Line > 1 pg_query /var/www/private/pgblog_private/system/databases/pgsql.class.php 350 > 2 dbQuery /var/www/private/pgblog_private/system/lib-database.php 213 > 3 DB_query /var/www/private/pgblog_private/system/lib-syndication.php 99 > 4 SYND_feedUpdateCheckAll /var/www/private/pgblog_private/system/lib-syndication.php 210 > 5 SYND_feedUpdateCheck /var/www/sites/thatlinuxbox.com/pgblog/lib-common.php 2662 > 6 COM_rdfUpToDateCheck /var/www/sites/thatlinuxbox.com/pgblog/index.php 154 > array(3) { > ["sql"]=> > string(229) "SELECT sid FROM gl_stories, gl_topic_assignments ta > WHERE draft_flag = 0 AND date <= NOW() AND ta.type = 'article' AND > ta.id = sid AND (ta.tid IN ('General','Geeklog')) AND perm_anon > 0 > GROUP BY sid ORDER BY date DESC LIMIT 10" > ["ignore_errors"]=> > int(0) > ["matches"]=> > array(0) { > } > } > > > In this case I didn't drop my database, so the install script did an > Upgrade, so I didn't get any issue about Superuser role. However, it > seems that the install script is adding the PLPGSQL stored procedure > language to the database (to the schema?), which does require > superuser role. > > I guess some people's databases are created with that language by > default, in my stock config it is not. > > > Regards, > > - Dan Stoner > > > On Tue, Jul 3, 2012 at 4:34 AM, Dirk Haun wrote: >> Quoting Dan Stoner : >> >>> I just got back from a tech conference and I'm a little fired up about >>> open source again. I allocated some time to test Geeklog with >>> postgresql. >> >> >> Thanks for doing that. >> >> Please note that Postgres support in 2.0.0b1 is very buggy and barely >> working. I made a few tweaks since then, so if you have the time, try a >> nightly tarball (or wait for b2). >> >> >> >>> 3. The admin account with default password of "password" did not seem >>> to exist or let me log in. >> >> >> Works for me (now - I think this was one of the post-b1 fixes that I had to >> make). >> >> >> >>> 4. The main page seemed to render ok, as do some of the sub-pages, >>> but many of them error out with "Unfortunately, an error has occurred >>> rendering this page. Please try again later." >> >> >> You may want to enable root debugging (in siteconfig.php). It'll not only >> print the error message directly in the browser window, but also displays a >> call stack so that you can see where the call is coming from (an error in >> pgsql.class.php isn't all that helpful - the faulty SQL was usually created >> elsewhere). >> >> >> >>> Looking in my geeklog error log, I see: >> >> (snip) >> These look familiar and should hopefully be fixed now. >> >> >> >>> Let me know what I can do to help. >> >> >> See above. Please note that I'm testing locally with Postgres 9.1, so it'll >> be interesting to see if there are other problems with 8.1. >> >> bye, Dirk >> >> >> >> _______________________________________________ >> geeklog-devel mailing list >> geeklog-devel at lists.geeklog.net >> http://eight.pairlist.net/mailman/listinfo/geeklog-devel From yankees26an at gmail.com Thu Jul 5 12:13:18 2012 From: yankees26an at gmail.com (Stanislav Palatnik) Date: Thu, 5 Jul 2012 12:13:18 -0400 Subject: [geeklog-devel] postgresql and Geeklog 2.0.0b1 In-Reply-To: References: <20120703103422.Horde.ISurBsL8999P8q6ObIQDfMA@webmail.df.eu> Message-ID: Hello Dan, 1. When the geeklog postgres database is created, also add the > plpgsql language with: > CREATE LANGUAGE plpgsql; Hmm why would this work compared to what already exists? Doesn't this command also need superuser privileges to execute? On Thu, Jul 5, 2012 at 9:54 AM, Dan Stoner wrote: > I thought about this some more and poked a bit. > > Newly created databases in PostgreSQL inherit from the template1 database. > > On my server, template1 does not include the plpgsql language (this > has been the default on all postgres database servers I have ever > administered). Maybe most hosting environments do add that language > to the template database, so most users would not see this issue. > > > The two likely alternatives to meeting this requirement and allow the > install to go through are: > > 1. When the geeklog postgres database is created, also add the > plpgsql language with: > > CREATE LANGUAGE plpgsql; > > 2. Before creating a new geeklog database, make sure that the > template1 database contains the plpgsql language. Same command as > above, just on the template1 database. All future databases that are > created would have this language available. > > > Here's some SQL... > > template1=# select * from pg_language where lanname='plpgsql'; > lanname | lanispl | lanpltrusted | lanplcallfoid | lanvalidator | lanacl > ---------+---------+--------------+---------------+--------------+-------- > (0 rows) > > template1=# CREATE LANGUAGE plpgsql; > CREATE LANGUAGE > > template1=# select * from pg_language where lanname='plpgsql'; > lanname | lanispl | lanpltrusted | lanplcallfoid | lanvalidator | lanacl > ---------+---------+--------------+---------------+--------------+-------- > plpgsql | t | t | 17625 | 17626 | > (1 row) > > > Regards, > > - Dan Stoner > > > On Wed, Jul 4, 2012 at 11:09 AM, Dan Stoner wrote: > > I tried the nightly tarball with root debugging enabled. > > > > I had a few issues related to filesystem permissions (not postgres), > > once those were fixed things got better. > > > > I was able to log in with Admin and default password. After clicking > > around a bit, most of the functionality seems to be ok (the admin > > panels are accessible). > > > > > > However, the main index page fails with the following: > > > > An error has occurred: > > > > This is being displayed as "Root Debugging" is enabled in your Geeklog > > configuration. > > > > If this is a production website you must disable this option once you > > have resolved any issues you are investigating. > > > > 2 - pg_query(): Query failed: ERROR: column "gl_stories.date" must > > appear in the GROUP BY clause or be used in an aggregate function @ > > /var/www/private/pgblog_private/system/databases/pgsql.class.php line > > 350 > > > > Call Stack > > # Function File Line > > 1 pg_query > /var/www/private/pgblog_private/system/databases/pgsql.class.php 350 > > 2 dbQuery /var/www/private/pgblog_private/system/lib-database.php > 213 > > 3 DB_query > /var/www/private/pgblog_private/system/lib-syndication.php 99 > > 4 SYND_feedUpdateCheckAll > /var/www/private/pgblog_private/system/lib-syndication.php 210 > > 5 SYND_feedUpdateCheck /var/www/sites/ > thatlinuxbox.com/pgblog/lib-common.php 2662 > > 6 COM_rdfUpToDateCheck /var/www/sites/ > thatlinuxbox.com/pgblog/index.php 154 > > array(3) { > > ["sql"]=> > > string(229) "SELECT sid FROM gl_stories, gl_topic_assignments ta > > WHERE draft_flag = 0 AND date <= NOW() AND ta.type = 'article' AND > > ta.id = sid AND (ta.tid IN ('General','Geeklog')) AND perm_anon > 0 > > GROUP BY sid ORDER BY date DESC LIMIT 10" > > ["ignore_errors"]=> > > int(0) > > ["matches"]=> > > array(0) { > > } > > } > > > > > > In this case I didn't drop my database, so the install script did an > > Upgrade, so I didn't get any issue about Superuser role. However, it > > seems that the install script is adding the PLPGSQL stored procedure > > language to the database (to the schema?), which does require > > superuser role. > > > > I guess some people's databases are created with that language by > > default, in my stock config it is not. > > > > > > Regards, > > > > - Dan Stoner > > > > > > On Tue, Jul 3, 2012 at 4:34 AM, Dirk Haun wrote: > >> Quoting Dan Stoner : > >> > >>> I just got back from a tech conference and I'm a little fired up about > >>> open source again. I allocated some time to test Geeklog with > >>> postgresql. > >> > >> > >> Thanks for doing that. > >> > >> Please note that Postgres support in 2.0.0b1 is very buggy and barely > >> working. I made a few tweaks since then, so if you have the time, try a > >> nightly tarball (or wait for b2). > >> > >> > >> > >>> 3. The admin account with default password of "password" did not seem > >>> to exist or let me log in. > >> > >> > >> Works for me (now - I think this was one of the post-b1 fixes that I > had to > >> make). > >> > >> > >> > >>> 4. The main page seemed to render ok, as do some of the sub-pages, > >>> but many of them error out with "Unfortunately, an error has occurred > >>> rendering this page. Please try again later." > >> > >> > >> You may want to enable root debugging (in siteconfig.php). It'll not > only > >> print the error message directly in the browser window, but also > displays a > >> call stack so that you can see where the call is coming from (an error > in > >> pgsql.class.php isn't all that helpful - the faulty SQL was usually > created > >> elsewhere). > >> > >> > >> > >>> Looking in my geeklog error log, I see: > >> > >> (snip) > >> These look familiar and should hopefully be fixed now. > >> > >> > >> > >>> Let me know what I can do to help. > >> > >> > >> See above. Please note that I'm testing locally with Postgres 9.1, so > it'll > >> be interesting to see if there are other problems with 8.1. > >> > >> bye, Dirk > >> > >> > >> > >> _______________________________________________ > >> geeklog-devel mailing list > >> geeklog-devel at lists.geeklog.net > >> http://eight.pairlist.net/mailman/listinfo/geeklog-devel > _______________________________________________ > geeklog-devel mailing list > geeklog-devel at lists.geeklog.net > http://eight.pairlist.net/mailman/listinfo/geeklog-devel > > -- > -- > My Best, > > Stanislav Palatnik > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From danstoner at gmail.com Thu Jul 5 12:38:04 2012 From: danstoner at gmail.com (Dan Stoner) Date: Thu, 5 Jul 2012 12:38:04 -0400 Subject: [geeklog-devel] postgresql and Geeklog 2.0.0b1 In-Reply-To: References: <20120703103422.Horde.ISurBsL8999P8q6ObIQDfMA@webmail.df.eu> Message-ID: Hi Stanislav, According to postgresql docs: "To create a database, you must be a superuser or have the special CREATEDB privilege." http://www.postgresql.org/docs/8.1/static/sql-createdatabase.html "The user must have the PostgreSQL superuser privilege to register a new language." http://www.postgresql.org/docs/8.1/static/sql-createlanguage.html When I create a new database for geeklog I am doing that step as a Superuser. Creating the language in that particular db would be just one additional step as Superuser. If template1 already has the language then this additional step is not needed. By the time I run the geeklog install script, the database should already exist (created by a Superuser). The database should also already have plpgsql added (added by a Superuser either to the individual database, or to template1 prior to creating the geeklog database). - Dan Stoner On Thu, Jul 5, 2012 at 12:13 PM, Stanislav Palatnik wrote: > Hello Dan, > >> 1. When the geeklog postgres database is created, also add the >> plpgsql language with: >> CREATE LANGUAGE plpgsql; > > > Hmm why would this work compared to what already exists? Doesn't this > command also need superuser privileges to execute? > > On Thu, Jul 5, 2012 at 9:54 AM, Dan Stoner wrote: >> >> I thought about this some more and poked a bit. >> >> Newly created databases in PostgreSQL inherit from the template1 database. >> >> On my server, template1 does not include the plpgsql language (this >> has been the default on all postgres database servers I have ever >> administered). Maybe most hosting environments do add that language >> to the template database, so most users would not see this issue. >> >> >> The two likely alternatives to meeting this requirement and allow the >> install to go through are: >> >> 1. When the geeklog postgres database is created, also add the >> plpgsql language with: >> >> CREATE LANGUAGE plpgsql; >> >> 2. Before creating a new geeklog database, make sure that the >> template1 database contains the plpgsql language. Same command as >> above, just on the template1 database. All future databases that are >> created would have this language available. >> >> >> Here's some SQL... >> >> template1=# select * from pg_language where lanname='plpgsql'; >> lanname | lanispl | lanpltrusted | lanplcallfoid | lanvalidator | lanacl >> ---------+---------+--------------+---------------+--------------+-------- >> (0 rows) >> >> template1=# CREATE LANGUAGE plpgsql; >> CREATE LANGUAGE >> >> template1=# select * from pg_language where lanname='plpgsql'; >> lanname | lanispl | lanpltrusted | lanplcallfoid | lanvalidator | lanacl >> ---------+---------+--------------+---------------+--------------+-------- >> plpgsql | t | t | 17625 | 17626 | >> (1 row) >> >> >> Regards, >> >> - Dan Stoner >> >> >> On Wed, Jul 4, 2012 at 11:09 AM, Dan Stoner wrote: >> > I tried the nightly tarball with root debugging enabled. >> > >> > I had a few issues related to filesystem permissions (not postgres), >> > once those were fixed things got better. >> > >> > I was able to log in with Admin and default password. After clicking >> > around a bit, most of the functionality seems to be ok (the admin >> > panels are accessible). >> > >> > >> > However, the main index page fails with the following: >> > >> > An error has occurred: >> > >> > This is being displayed as "Root Debugging" is enabled in your Geeklog >> > configuration. >> > >> > If this is a production website you must disable this option once you >> > have resolved any issues you are investigating. >> > >> > 2 - pg_query(): Query failed: ERROR: column "gl_stories.date" must >> > appear in the GROUP BY clause or be used in an aggregate function @ >> > /var/www/private/pgblog_private/system/databases/pgsql.class.php line >> > 350 >> > >> > Call Stack >> > # Function File Line >> > 1 pg_query >> > /var/www/private/pgblog_private/system/databases/pgsql.class.php 350 >> > 2 dbQuery /var/www/private/pgblog_private/system/lib-database.php >> > 213 >> > 3 DB_query >> > /var/www/private/pgblog_private/system/lib-syndication.php 99 >> > 4 SYND_feedUpdateCheckAll >> > /var/www/private/pgblog_private/system/lib-syndication.php 210 >> > 5 SYND_feedUpdateCheck >> > /var/www/sites/thatlinuxbox.com/pgblog/lib-common.php 2662 >> > 6 COM_rdfUpToDateCheck >> > /var/www/sites/thatlinuxbox.com/pgblog/index.php 154 >> > array(3) { >> > ["sql"]=> >> > string(229) "SELECT sid FROM gl_stories, gl_topic_assignments ta >> > WHERE draft_flag = 0 AND date <= NOW() AND ta.type = 'article' AND >> > ta.id = sid AND (ta.tid IN ('General','Geeklog')) AND perm_anon > 0 >> > GROUP BY sid ORDER BY date DESC LIMIT 10" >> > ["ignore_errors"]=> >> > int(0) >> > ["matches"]=> >> > array(0) { >> > } >> > } >> > >> > >> > In this case I didn't drop my database, so the install script did an >> > Upgrade, so I didn't get any issue about Superuser role. However, it >> > seems that the install script is adding the PLPGSQL stored procedure >> > language to the database (to the schema?), which does require >> > superuser role. >> > >> > I guess some people's databases are created with that language by >> > default, in my stock config it is not. >> > >> > >> > Regards, >> > >> > - Dan Stoner >> > >> > >> > On Tue, Jul 3, 2012 at 4:34 AM, Dirk Haun wrote: >> >> Quoting Dan Stoner : >> >> >> >>> I just got back from a tech conference and I'm a little fired up about >> >>> open source again. I allocated some time to test Geeklog with >> >>> postgresql. >> >> >> >> >> >> Thanks for doing that. >> >> >> >> Please note that Postgres support in 2.0.0b1 is very buggy and barely >> >> working. I made a few tweaks since then, so if you have the time, try a >> >> nightly tarball (or wait for b2). >> >> >> >> >> >> >> >>> 3. The admin account with default password of "password" did not seem >> >>> to exist or let me log in. >> >> >> >> >> >> Works for me (now - I think this was one of the post-b1 fixes that I >> >> had to >> >> make). >> >> >> >> >> >> >> >>> 4. The main page seemed to render ok, as do some of the sub-pages, >> >>> but many of them error out with "Unfortunately, an error has occurred >> >>> rendering this page. Please try again later." >> >> >> >> >> >> You may want to enable root debugging (in siteconfig.php). It'll not >> >> only >> >> print the error message directly in the browser window, but also >> >> displays a >> >> call stack so that you can see where the call is coming from (an error >> >> in >> >> pgsql.class.php isn't all that helpful - the faulty SQL was usually >> >> created >> >> elsewhere). >> >> >> >> >> >> >> >>> Looking in my geeklog error log, I see: >> >> >> >> (snip) >> >> These look familiar and should hopefully be fixed now. >> >> >> >> >> >> >> >>> Let me know what I can do to help. >> >> >> >> >> >> See above. Please note that I'm testing locally with Postgres 9.1, so >> >> it'll >> >> be interesting to see if there are other problems with 8.1. >> >> >> >> bye, Dirk >> >> >> >> >> >> >> >> _______________________________________________ >> >> geeklog-devel mailing list >> >> geeklog-devel at lists.geeklog.net >> >> http://eight.pairlist.net/mailman/listinfo/geeklog-devel >> _______________________________________________ >> geeklog-devel mailing list >> geeklog-devel at lists.geeklog.net >> http://eight.pairlist.net/mailman/listinfo/geeklog-devel >> >> -- >> -- >> My Best, >> >> Stanislav Palatnik >> > > _______________________________________________ > geeklog-devel mailing list > geeklog-devel at lists.geeklog.net > http://eight.pairlist.net/mailman/listinfo/geeklog-devel > From jmucchiello at yahoo.com Thu Jul 5 21:34:05 2012 From: jmucchiello at yahoo.com (Joe Mucchiello) Date: Thu, 5 Jul 2012 18:34:05 -0700 (PDT) Subject: [geeklog-devel] postgresql and Geeklog 2.0.0b1 Message-ID: <1341538445.29200.YahooMailNeo@web121502.mail.ne1.yahoo.com> > "To create a database, you must be a superuser or have the special > CREATEDB privilege." > http://www.postgresql.org/docs/8.1/static/sql-createdatabase.html > > "The user must have the PostgreSQL superuser privilege to register a > new language." > http://www.postgresql.org/docs/8.1/static/sql-createlanguage.html > > When I create a new database for geeklog I am doing that step as a > Superuser. The Geeklog install script doesn't run the create database command in pgsql, mysql or mssql. It assumes the database is already created in order to avoid the problem of needing create database privilege. Most shared hosts provide a runnable script that creates databases outside of the normal admin programs for the database. (This is how they limit users to 1, 3, 5 or some other number of databases.) So perhaps instead the pgsql install just needs to check if the language is set on the database, attempt to create it if it isn't there, and failing that, explain what is needed to get it up and running. I really don't think Geeklog should take it on itself to modify "templates" in the database system. That is too far outside Geeklog's responsibilities to be mucking with. ? Joe -------------- next part -------------- An HTML attachment was scrubbed... URL: From danstoner at gmail.com Fri Jul 6 07:55:19 2012 From: danstoner at gmail.com (Dan Stoner) Date: Fri, 6 Jul 2012 07:55:19 -0400 Subject: [geeklog-devel] postgresql and Geeklog 2.0.0b1 In-Reply-To: <1341538445.29200.YahooMailNeo@web121502.mail.ne1.yahoo.com> References: <1341538445.29200.YahooMailNeo@web121502.mail.ne1.yahoo.com> Message-ID: I did not mean to suggest that the install script should be doing this. Noting the prerequisites in the INSTALL doc or release notes might be sufficient. The script could be extra nice and check to see if the language is in the database, and give a nice message rather than failing the SQL call (in the same way the script notices if a database does not exist). Cheers, Dan Stoner On Jul 5, 2012 9:39 PM, "Joe Mucchiello" wrote: > > "To create a database, you must be a superuser or have the special > > CREATEDB privilege." > > http://www.postgresql.org/docs/8.1/static/sql-createdatabase.html > > > > "The user must have the PostgreSQL superuser privilege to register a > > new language." > > http://www.postgresql.org/docs/8.1/static/sql-createlanguage.html > > > > When I create a new database for geeklog I am doing that step as a > > Superuser. > > The Geeklog install script doesn't run the create database command in > pgsql, mysql or mssql. It assumes the database is already created in order > to avoid the problem of needing create database privilege. Most shared > hosts provide a runnable script that creates databases outside of the > normal admin programs for the database. (This is how they limit users to 1, > 3, 5 or some other number of databases.) > > So perhaps instead the pgsql install just needs to check if the language > is set on the database, attempt to create it if it isn't there, and failing > that, explain what is needed to get it up and running. I really don't think > Geeklog should take it on itself to modify "templates" in the database > system. That is too far outside Geeklog's responsibilities to be mucking > with. > > Joe > > _______________________________________________ > geeklog-devel mailing list > geeklog-devel at lists.geeklog.net > http://eight.pairlist.net/mailman/listinfo/geeklog-devel > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From websitemaster at cogeco.net Thu Jul 12 09:59:00 2012 From: websitemaster at cogeco.net (Tom) Date: Thu, 12 Jul 2012 09:59:00 -0400 Subject: [geeklog-devel] Geeklog 2.0.0b2 Message-ID: <000001cd6036$7d226900$77673b00$@cogeco.net> Hey All, I will release Beta 2 tomorrow unless someone has a good reason to hold off. Tom -------------- next part -------------- An HTML attachment was scrubbed... URL: From dirk at haun-online.de Thu Jul 12 16:05:11 2012 From: dirk at haun-online.de (Dirk Haun) Date: Thu, 12 Jul 2012 22:05:11 +0200 Subject: [geeklog-devel] Geeklog 2.0.0b2 In-Reply-To: <000001cd6036$7d226900$77673b00$@cogeco.net> References: <000001cd6036$7d226900$77673b00$@cogeco.net> Message-ID: <8DA917C8-4F02-4729-92AB-4A3571C7393C@haun-online.de> Tom wrote: > I will release Beta 2 tomorrow unless someone has a good reason to hold off. Not a "blocker", but it seems we neglected the Changelog. I've made at least a quick update with the issues that were resolved according to our bugtracker. The level of detail is not up to our usual standards, though, and we should revisit it when we enter the RC phase. bye, Dirk -- http://www.themobilepresenter.com/ From websitemaster at cogeco.net Sun Jul 15 09:58:16 2012 From: websitemaster at cogeco.net (Tom) Date: Sun, 15 Jul 2012 09:58:16 -0400 Subject: [geeklog-devel] Spam X SLV Message-ID: <00c901cd6291$e446d640$acd482c0$@cogeco.net> I notice that linksleeve.org and service has been down for the last couple of days. This means that the SLV module for SPAMX has not been able to do its job. Does anyone know what happen? (a quick Google search didn't reveal much) Tom From dirk at haun-online.de Sun Jul 15 11:52:14 2012 From: dirk at haun-online.de (Dirk Haun) Date: Sun, 15 Jul 2012 17:52:14 +0200 Subject: [geeklog-devel] Spam X SLV In-Reply-To: <00c901cd6291$e446d640$acd482c0$@cogeco.net> References: <00c901cd6291$e446d640$acd482c0$@cogeco.net> Message-ID: <833091E2-D247-46AC-A8B9-154F0E02930D@haun-online.de> Tom wrote: > Does anyone know what happen? (a quick Google search didn't reveal much) No idea. I sent an email to the whois contact address and to another address from a previous contact (but which is a few years old). We'll see ? bye, Dirk -- http://www.themobilepresenter.com/ From dirk at haun-online.de Sun Jul 15 12:44:39 2012 From: dirk at haun-online.de (Dirk Haun) Date: Sun, 15 Jul 2012 18:44:39 +0200 Subject: [geeklog-devel] Spam X SLV In-Reply-To: <00c901cd6291$e446d640$acd482c0$@cogeco.net> References: <00c901cd6291$e446d640$acd482c0$@cogeco.net> Message-ID: Tom wrote: > I notice that linksleeve.org and service has been down for the last couple > of days. This means that the SLV module for SPAMX has not been able to do > its job. Should be back up. bye, Dirk -- http://www.themobilepresenter.com/ From websitemaster at cogeco.net Sun Jul 15 20:04:09 2012 From: websitemaster at cogeco.net (Tom) Date: Sun, 15 Jul 2012 20:04:09 -0400 Subject: [geeklog-devel] Spam X SLV In-Reply-To: References: <00c901cd6291$e446d640$acd482c0$@cogeco.net> Message-ID: <00d901cd62e6$864bff20$92e3fd60$@cogeco.net> Good job :-) I never thought of checking the whois contact. Tom -----Original Message----- From: geeklog-devel-bounces at lists.geeklog.net [mailto:geeklog-devel-bounces at lists.geeklog.net] On Behalf Of Dirk Haun Sent: July-15-12 12:45 PM To: Geeklog Development Subject: Re: [geeklog-devel] Spam X SLV Tom wrote: > I notice that linksleeve.org and service has been down for the last > couple of days. This means that the SLV module for SPAMX has not been > able to do its job. Should be back up. bye, Dirk -- http://www.themobilepresenter.com/ _______________________________________________ geeklog-devel mailing list geeklog-devel at lists.geeklog.net http://eight.pairlist.net/mailman/listinfo/geeklog-devel From ironmax at spacequad.com Sun Jul 15 20:38:25 2012 From: ironmax at spacequad.com (ironmax at spacequad.com) Date: Sun, 15 Jul 2012 20:38:25 -0400 Subject: [geeklog-devel] geeklog-devel Digest, Vol 69, Issue 8 In-Reply-To: References: Message-ID: <81952CE96B044C88828254E598BDF6F2@spacequad.com> Tom, It seems to be working fine with all the domains I am hosting, in Canada Michael -----Original Message----- Message: 1 Date: Sun, 15 Jul 2012 09:58:16 -0400 From: "Tom" Subject: [geeklog-devel] Spam X SLV To: "'Geeklog Development'" Message-ID: <00c901cd6291$e446d640$acd482c0$@cogeco.net> Content-Type: text/plain; charset="us-ascii" I notice that linksleeve.org and service has been down for the last couple of days. This means that the SLV module for SPAMX has not been able to do its job. Does anyone know what happen? (a quick Google search didn't reveal much) Tom ------------------------------ _______________________________________________ geeklog-devel mailing list geeklog-devel at lists.geeklog.net http://eight.pairlist.net/mailman/listinfo/geeklog-devel End of geeklog-devel Digest, Vol 69, Issue 8 ******************************************** From dirk at haun-online.de Mon Jul 16 01:55:22 2012 From: dirk at haun-online.de (Dirk Haun) Date: Mon, 16 Jul 2012 07:55:22 +0200 Subject: [geeklog-devel] geeklog-devel Digest, Vol 69, Issue 8 In-Reply-To: <81952CE96B044C88828254E598BDF6F2@spacequad.com> References: <81952CE96B044C88828254E598BDF6F2@spacequad.com> Message-ID: <4FD0870D-0054-4B65-92B5-DB573720840E@haun-online.de> > It seems to be working fine with all the domains I am hosting, in Canada SLV was down (from the logs, it seems since Wednesday), but is now back up. bye, Dirk -- http://www.themobilepresenter.com/ From dirk at haun-online.de Fri Jul 27 16:28:49 2012 From: dirk at haun-online.de (Dirk Haun) Date: Fri, 27 Jul 2012 22:28:49 +0200 Subject: [geeklog-devel] "OAuth 2.0 is a bad protocol" Message-ID: <720D8837-7E8F-4823-8ADE-8DEEE9CC4601@haun-online.de> Interesting. Just the other day, I was thinking about OAuth 2.0 again and how we never got around to actually use the code we got for it (via GSoC). And today, I found this article: http://hueniverse.com/2012/07/oauth-2-0-and-the-road-to-hell/ Choice quotes: > Last month I reached the painful conclusion that I can no longer be associated with the OAuth 2.0 standard. I resigned my role as lead author and editor, withdraw my name from the specification, and left the working group. (?) > At the end, I reached the conclusion that OAuth 2.0 is a bad protocol. WS-* bad. It is bad enough that I no longer want to be associated with it. (?) > When compared with OAuth 1.0, the 2.0 specification is more complex, less interoperable, less useful, more incomplete, and most importantly, less secure. (?) > If you are currently using 1.0 successfully, ignore 2.0. It offers no real value over 1.0 Not sure which practical consequences to draw from this (how much of a real-world problem is not supporting OAuth 2.0, now that we allow authentication via Twitter and Facebook?). But at least I don't feel so bad about not using the code any more ? bye, Dirk -- http://www.themobilepresenter.com/