Re: [Geotools-devel] JDBC1Datastore vs AutoCommit

2006-10-18 Thread Cory Horner
Cory Horner wrote: >We're changing the way uDig uses transactions -- so a transaction is >AUTO_COMMIT until a modification occurs, and it remains open until the >user commits the transaction: > >http://jira.codehaus.org/browse/UDIG-1051 > > On second though, this does seem to be needed for Pos

Re: [Geotools-devel] JDBC1Datastore vs AutoCommit

2006-10-17 Thread Jody Garnett
Cory Horner wrote: > Howdy, > > We're changing the way uDig uses transactions -- so a transaction is > AUTO_COMMIT until a modification occurs, and it remains open until the > user commits the transaction: > Note: Some feature source / store methods make use of their own internal transaction

Re: [Geotools-devel] JDBC1Datastore vs AutoCommit

2006-10-17 Thread David Adler
I'm in favor. Is this related to GEOT-894? At 07:14 PM 10/17/2006, Cory Horner wrote: >Howdy, > >We're changing the way uDig uses transactions -- so a transaction is >AUTO_COMMIT until a modification occurs, and it remains open until the >user commits the transaction: > >http://jira.codehaus.org/

[Geotools-devel] JDBC1Datastore vs AutoCommit

2006-10-17 Thread Cory Horner
Howdy, We're changing the way uDig uses transactions -- so a transaction is AUTO_COMMIT until a modification occurs, and it remains open until the user commits the transaction: http://jira.codehaus.org/browse/UDIG-1051 Here is a patch for module/main which shows the line which needs to be rem