[postgis-users] Running createlang and postgis.sql

[chodgson@refractions.net]

I don't know off-hand where the begin _should_ go, but an extraneous end/commit 
will certainly not hurt anything.

I think the only reason for the transaction to be there is to cause either ALL 
or NONE of the postgis functionality to be loaded - otherwise it is rather 
painful to unload the half that got installed, and/or reload the rest, after 
you've fixed whatever problem caused it to fail in the first place.

I don't have time to look at the code right now, but I think there should be a 
begin at the VERY beginning, and an end at the VERY end - with all the separate 
files for different versions and components, it is not too hard to understand 
why there are unmatched begins/ends.

Chris


Quoting Mike Leahy <mgleahy@fes.uwaterloo.ca>:

> Robert,
> 
> You are correct about the commit.  I've found in the postgis.sql file
> that I'm using there is an extraneous "end transaction;" about two
> thirds of the way through the script (Line 761 of the attached file).  I
> think that despite the lack of a matching "begin transaction;",
> everything is working fine, but I don't want to find out later on that
> I'm missing a bunch of the functions that are defined in this file.  Can
> someone suggest whether I should add a begin transaction, and if so,
> where should it go?
> 
> Thanks for the help,
> Mike
> 
> -----Original Message-----
> From: postgis-users-bounces@postgis.refractions.net
> [mailto:postgis-users-bounces@postgis.refractions.net] On Behalf Of Paul
> Ramsey
> Sent: June 3, 2003 11:47 AM
> To: PostGIS Users Discussion
> Subject: Re: [postgis-users] Running createlang and postgis.sql
> 
> At a minimum we appear to have an extraneous 'commit' (or a missing
> 'begin') in the sql file which needs to be paired up. The other warnings
> 
> are normal and can be ignored -- when postgresql creates new types there
> 
> is a brief period of chicken-and-egg, when types have been created and
> their constructurs have not, which raises a warning.
> P.
> 
> Robert W. Burgholzer wrote:
> > Mike,
> > I just had the same question today, thanks for posting it. It seems
> that all
> > is right with the data import and I have successfully added data, and
> > retrieved it via avpgcon. So, I would say, no, you needn't be
> concerned,
> > however, I am not the expert, thus all I say is speculation.
> >
> > Perhaps this is a newer, more extensive error condition being reported
> by
> > postgresql, are you running postgresql 7.3.3?? (I am)
> >
> > Robert Burgholzer
> >
> >
> >>===== Original Message From PostGIS Users Discussion
> >
> > <postgis-users@postgis.refractions.net> =====
> >
> >>Just a couple questions.when I run the command "createlang plpgsql
> >>[DBNAME]" - I'm guessing that means the build that I got already
> >>includes this language?
> >>
> >>
> >>
> >>Next, when I run "psql -d independencia -f postgis.sql", I get the
> >>following notices:
> >>
> >>
> >>
> >>NOTICE:  CREATE TABLE / PRIMARY KEY will create implicit index
> >>'spatial_ref_sys_pkey' for table 'spatial_ref_sys'
> >>
> >>NOTICE:  CREATE TABLE / PRIMARY KEY will create implicit index 'gc_pk'
> >>for table 'geometry_columns'
> >>
> >>NOTICE:  ProcedureCreate: type box3d is not yet defined
> >>
> >>NOTICE:  ProcedureCreate: type spheroid is not yet defined
> >>
> >>NOTICE:  ProcedureCreate: type wkb is not yet defined
> >>
> >>NOTICE:  ProcedureCreate: type chip is not yet defined
> >>
> >>NOTICE:  ProcedureCreate: type geometry is not yet defined
> >>
> >>NOTICE:  COMMIT: no transaction in progress
> >>
> >>
> >>
> >>Is this an issue I should be concerned about?
> >>
> >>
> >>
> >>Kind regards,
> >>
> >>Mike
> >
> >
> >
> >
> > _______________________________________________
> > postgis-users mailing list
> > postgis-users@postgis.refractions.net
> > http://postgis.refractions.net/mailman/listinfo/postgis-users
> 
> 
> 
> _______________________________________________
> postgis-users mailing list
> postgis-users@postgis.refractions.net
> http://postgis.refractions.net/mailman/listinfo/postgis-users
>