PostgreSQL not working in CREATE TABLE node_<field> code

Thanks. I don't know a thing to PostgreSQL, and I have no config to test.
Is this a bug fix or 'just' an improvement ?
Plus : can you provide the update function as well ? :-)

Is this a bug fix or 'just' an improvement ?
Sorry, the title says 'not working' so I guess I have the answer...

Update function still accepted, though :-)

I'm not sure we need an update function actually :
If cck is currently broken for prostgre, maybe there are no tables to update...
Plus cck 5 is beta.

Karen, dopry, what do you think ?

This seems to be in line with core, but needs to be expanded. The content.install file has some changes that need to be made and there is another place in the content_db_add_column function that needs a change -- the place where the mappings are done. Currently the mapping script is going to create 'integer unsigned' instead of 'int_unsigned', so some thought needs to go into how that script should be changed to produce the right results. There may be other places that need patching, those were two I saw.

Changing status...

Bumping this issue. This is needs more work and needs to get finished up.

the current patch seems to fix the issue for module content_admin.inc. the module content_crud.inc needs the same corrections.
also see the issue http://drupal.org/node/106892 i opened for this postgres problem.

I'm marking http://drupal.org/node/109060 and http://drupal.org/node/106892 as duplicates of this issue. We'll consolidate work on a patch to fix this in this issue.

Here's my proposed patch against HEAD for this problem. I need to know if it is correct and fixes the problem, if it fixes all the places that need to be fixed, and confirm that it does not break version 7 compatibility or whatever our standard is.

The patch fixes content.install, content_crud.inc, and content_admin.inc, including the $mappings.

Need to change the status...

Guess I should also change who's assigned to this.

I never got any confirming reviews for this patch, but it's consistant with the way core is handling it so I assume it is OK and I just committed it to HEAD and 5.x. It looks like there is nothing to change in 4.7 since 4.7 doesn't use 'integer unsigned' for postgres anywhere.