Opened 13 years ago
Last modified 7 years ago
#1770 new patch
long transaction
Reported by: | cruizch | Owned by: | cruizch |
---|---|---|---|
Priority: | medium | Milestone: | PostGIS Fund Me |
Component: | postgis | Version: | 2.0.x |
Keywords: | long transaction | Cc: | strk |
Description
The long transaction support requires to be cleaned up and updated, this process involves the actualization of the WFS_locks and long_xact SQL code, the function signatures change (to begin with ST_) and the adding of some regression tests.
These changes will only affect the WFS_locks.sql.in and the long_xact.sql.in.c files.
Attachments (1)
Change History (13)
comment:1 by , 13 years ago
comment:2 by , 13 years ago
OK. Just the long_xact.sql.in.c file will be affected. The extras/WFS_locks will be erased by the trunk manager.
comment:3 by , 13 years ago
I'm thinking that ST_ prefix would be inappropriate because the module has nothing to do with geometries. It provides locking support for _any_ kind of row. A postgis_ one would still make sense to identify the functions as being installed by postgis…
comment:5 by , 13 years ago
The long_xact.sql.in.c file has been modified with the new function signatures, code cleansing, some function documentation added, and credits from the WFS_locks.sql.in added. Also, three regression tests were created.
comment:6 by , 13 years ago
Milestone: | PostGIS 2.0.1 → PostGIS 2.1.0 |
---|
comment:7 by , 12 years ago
Do we have issue with accepting this patch. I don't have need for long transaction support so not a good judge for this.
comment:8 by , 12 years ago
Milestone: | PostGIS 2.1.0 → PostGIS 2.2.0 |
---|
comment:9 by , 9 years ago
Milestone: | PostGIS 2.2.0 → PostGIS 2.3.0 |
---|---|
Priority: | low → medium |
Type: | task → patch |
Pity this hasn't made it into 2.2.0, it has a patch attached !
comment:10 by , 9 years ago
Yah how did we let it slide. I think part of the reason is that long transaction support used to only be used by geoserver and unless one of us is using geoserver, it's hard to verify the patch doesn't break something. I'm not even sure if geoserver uses that anymore. Not sure if QGIS uses it and that whole long transaction stuff probably needs a closer revisiting so the patch may already be obsolete.
comment:11 by , 8 years ago
Milestone: | PostGIS 2.3.0 → PostGIS Future |
---|
I'm tempted to just mark this as won't fix as I suspect since long transaction support hasn't been touched in ages, it's probably not used anymore otherwise more people would have complained.
It seems to me that the long_xact.sql.in.c is a verbatim copy of WFS_locks.sql.in in that case the whole extras/WFS_locks directory should probably die…