| 01:14:12 | CIA-6: | mcayland * r2884 /trunk/liblwgeom/Makefile.in: Fix -o option for flex; the output filename must be specified immediately after -o without any preceding white space. Per report from Jorgen Austvik. |
| 02:03:59 | CIA-6: | mcayland * r2885 /trunk/ ( 6 files in 4 dirs ): |
| 02:03:59 | CIA-6: | Allow Makefile substitutions for PGCONFIG which are required if pg_config is not |
| 02:03:59 | CIA-6: | located in the path. Note that as part of this fix, we now need to generate |
| 02:03:59 | CIA-6: | topology/Makefile using autoconf. Thanks to Jorgen Austvik for the report and |
| 02:03:59 | CIA-6: | partial patch. |
| 02:38:39 | arkygeek: | good morning :- ) |
| 02:40:08 | arkygeek: | I have a table with lat/lon coordinates. I added geom to it, and transformed it to UTM36N. This is working great. However, I want to add two new columns to the table ( UTMEasting and UTMNorthing ) Is this possible to do... ? |
| 02:41:29 | johto: | alter table tbl add column .. |
| 02:41:44 | johto: | Last login: Sun Jul 27 20:51:45 2008 from dsl-hkibrasgw2-ff75c300-9.dhcp.inet.fi |
| 02:41:47 | johto: | pointer% screen -r 30991 |
| 02:42:04 | johto: | err.. http://www.postgresql.org/docs/current/static/sql-altertable.html |
| 02:42:05 | sigq: | Title: PostgreSQL: Documentation: Manuals: PostgreSQL 8.3: ALTER TABLE ( at www.postgresql.org ) |
| 02:43:37 | arkygeek: | johto: hi. oh, I can add the column :- ) I was unclear... I want to populate the columns with calculated UTM coordinates |
| 04:08:16 | CIA-6: | robe * r2886 /trunk/doc/ ( reference.xml reference_new.xml ): Move over ST_GeomFromEWKT, ST_GeomFromEWKB and provide examples. Fix some miscellaneous references. |
| 04:57:22 | darkblue_B: | arkygeek: |
| 04:57:39 | darkblue_B: | if you know the formula, thats easy to do |
| 04:57:55 | darkblue_B: | I have to run out the door, but will be back later |
| 07:38:48 | GeomatiLux: | Hello freinds, |
| 07:39:02 | GeomatiLux: | I need help please ? |
| 07:39:47 | GeomatiLux: | I search the metadata mapping shema of postgresSQL/Postgis ? |
| 07:54:19 | jlivni: | GeomatiLux: Did you have a question? |
| 07:59:12 | GeomatiLux: | yes jlivni : ) |
| 07:59:33 | mloskot: | GeomatiLux: don't ask if you can ask, just ask |
| 08:00:15 | GeomatiLux: | I need the postgresSQL/postgis metadata mapping schema please |
| 08:00:41 | mloskot: | what you mean as "metadata mapping schema" |
| 08:02:17 | GeomatiLux: | It will be like that: <?xml version="1.0"?> |
| 08:02:18 | GeomatiLux: | <mapping> |
| 08:02:20 | GeomatiLux: | <dbms product="POSTGRESQL" id="postgres_id" label="Mapping Postgres" |
| 08:02:21 | GeomatiLux: | default="true"> |
| 08:02:23 | GeomatiLux: | <dbTypes> |
| 08:02:24 | GeomatiLux: | <dbType type="ABSTIME" ignoreLen="true" ignorePre="true" /> |
| 08:02:26 | GeomatiLux: | <dbType type="ACLITEM" ignoreLen="true" ignorePre="true" /> |
| 08:02:27 | GeomatiLux: | <dbType type="BIGSERIAL" ignoreLen="true" ignorePre="true" /> |
| 08:02:29 | GeomatiLux: | <dbType type="BIT" ignorePre="true" /> |
| 08:02:42 | mloskot: | GeomatiLux: if you don't mind, send paste's to http://rafb.net/paste |
| 08:02:43 | sigq: | Title: Nopaste ( at rafb.net ) |
| 08:02:58 | mloskot: | Anyway, there is no support for that out of the box |
| 08:03:04 | mloskot: | ORM is out of PostGIS scope |
| 08:03:17 | mloskot: | GeomatiLux: I'd ask on #postgresql channel |
| 08:04:36 | GeomatiLux: | Thank you mloskot, |
| 08:18:31 | darkblue_B: | arkygeek: hey |
| 09:12:10 | hobu: | what are some postgres configuration tweaks I can make to a stock install to make seqscans through a 11 million record table go much faster? |
| 09:12:41 | darkblue_B: | hi hobu |
| 09:12:56 | darkblue_B: | at OSCon I went to one of the Pg talks |
| 09:13:15 | darkblue_B: | I think I remember hearing that there are actually over 100 settings in Pg |
| 09:13:34 | darkblue_B: | most people are expected to touch at most about 5-8 |
| 09:13:51 | darkblue_B: | the speaker said he was "totally crazy" so he touches about 10 |
| 09:14:20 | darkblue_B: | there is an amount of memory allocated to a single query thats important |
| 09:14:43 | darkblue_B: | of course, there is a big diff is you are going to have dozens or hundred+ queries potentially at the same time |
| 09:14:58 | darkblue_B: | .. or if its just you doing something one at a time |
| 09:15:18 | darkblue_B: | in their data mining system they built a queue, so only 1 query can execute at a time |
| 09:15:38 | darkblue_B: | of course you have built your indexes |
| 09:16:07 | darkblue_B: | I'm not adept at reading the query planner but thats used heavily to optimize your query itself |
| 09:16:21 | hobu: | my usage is some one-time operations where I'm passing through the table and updating records |
| 09:16:36 | hobu: | only a couple of connections to the db active |
| 09:16:41 | darkblue_B: | yeah |
| 09:17:16 | darkblue_B: | it was really stunning to see the size of the data sets that are being handled by Pg |
| 09:17:43 | darkblue_B: | even though MySQL is getting huge resources.. Pg is actually continuing to hold its niche |
| 09:18:05 | darkblue_B: | .. due to its ability to handle these absolutely enormous data sets with 100% reliability |
| 09:22:24 | darkblue_B: | hobu: my guess is you do wht you can to load the whole thing into RAM |
| 09:22:36 | landonf: | Design debt and Brooks' law applied -- pushing more resources to MySQL doesn't result in linear scaling of development productivity. |
| 09:23:08 | jlivni: | hobu: i'd try #postgres ... |
| 09:23:14 | darkblue_B: | sure, but there are dozens of directions a project takes.. and each gets resources |
| 09:23:22 | darkblue_B: | at once |
| 09:23:28 | jlivni: | that said, away from a stock install i always up my shared buffers and effective_cache_size |
| 09:23:35 | landonf: | I still hold that both 'axioms' apply = ) |
| 09:47:02 | darkblue_B: | TheSteveMonkey: I saw you post earler |
| 09:47:05 | darkblue_B: | earlier |
| 09:47:12 | darkblue_B: | about that query |
| 09:48:13 | darkblue_B: | I believe I solved that problem without the sub-query or Union( ) |
| 09:48:51 | TheSteveMonkey: | darkblue_B: thanks |
| 09:49:08 | TheSteveMonkey: | then perhaps I should make a layer that is pre-unioned bc it is taking way tooo long |
| 09:50:12 | darkblue_B: | why do you need a Union( ) ? |
| 09:51:09 | TheSteveMonkey: | I can't use a subquery bc within( ) does not accept a geometry collection |
| 09:51:56 | darkblue_B: | I just used an AND in the WHERE |
| 09:51:58 | darkblue_B: | z.comb_zn like 'A%' and ST_Within( p.the_geom, z.the_geom ) |
| 09:52:22 | TheSteveMonkey: | that is what I am doing and it complains about collection |
| 09:52:58 | darkblue_B: | I dont get a collection.. I used two poly tables.. |
| 09:53:16 | TheSteveMonkey: | ahh I see |
| 09:53:22 | TheSteveMonkey: | I am doing the same and I tried that |
| 09:53:40 | TheSteveMonkey: | but I will go back and try again |
| 10:13:01 | TheSteveMonkey: | thanks darkblue_B - it works |
| 11:05:48 | CIA-6: | kneufeld * r2887 /trunk/doc/ ( reference.xml reference_new.xml ): fixed some minor xrefences in the documentation so the autobuild doesn't throw errors. |
| 13:01:59 | darkblue_B: | TheSteveMonkey: I created an account at the deCarta Dev Zone |
| 13:02:10 | darkblue_B: | I am "unapproved" !!? |
| 13:02:25 | darkblue_B: | I feel so left out :p |
| 13:02:57 | darkblue_B: | so where is the "live PostGIS hacking" going to happen? |
| 13:03:10 | darkblue_B: | in my experience, it will take the two weeks to do it :- ) |
| 13:04:25 | darkblue_B: | apparently no underscores allowed.. so I am darkblueb |
| 20:43:24 | CIA-6: | robe * r2888 /trunk/doc/reference_new.xml: minor formatting changes, note about ST_NumPOints being an alias for ST_NPoints moving forward. |
| 21:11:39 | CIA-6: | kneufeld * r2889 /trunk/doc/ ( 11 files in 3 dirs ): moved over st_centroid and provided examples in both svg and png format. Eventually, ImageMagick will be used to generate the png files, but for now they are committed into svn. |
| 21:44:24 | CIA-6: | kneufeld * r2890 /trunk/doc/ ( reference.xml reference_new.xml ): moved over st_pointonsurface and provided examples. |