Opened 16 years ago
Closed 13 years ago
#2454 closed defect (wontfix)
Reset error list before (after?) GC map object
Reported by: | project10 | Owned by: | dmorissette |
---|---|---|---|
Priority: | low | Milestone: | 6.0 release |
Component: | MapServer C Library | Version: | svn-trunk (development) |
Severity: | trivial | Keywords: | |
Cc: |
Description
I noticed that, under WxS wrapper (mapscript within long-running php process), that the error list was not reset between requests and would continue to accumulate.
Adding msResetErrorList() to msFreeMap() in mapobject.c fixes the issue for me. Perhaps there is a more elegant place to put this call, though.
Change History (6)
comment:1 by , 16 years ago
Milestone: | 5.0.1 release → 5.2 release |
---|---|
Status: | new → assigned |
comment:2 by , 16 years ago
Priority: | normal → low |
---|---|
Severity: | normal → trivial |
comment:4 by , 15 years ago
Milestone: | 5.2.1 release → 5.6 release |
---|
Do we want to add a call to msResetErrorList() in msFreeMap() for 5.6?
comment:5 by , 14 years ago
Milestone: | 5.6 release → 6.0 release |
---|---|
Owner: | changed from | to
Status: | assigned → new |
comment:6 by , 13 years ago
Resolution: | → wontfix |
---|---|
Status: | new → closed |
I'll close WONTFIX. I'm worried that resetting the error list in msFreeMap() could cause side-effects to scripts using multiple map objects. Let's leave it up to the script itself to decide when is the right time to reset the error list.
I'm not sure it is prudent to address this in 5.0.1.
I think two issues arise. One is whether we should automatically clear the error stack in some strategic locations to avoid accumulating lots of old messages. The other is whether msResetErrorList() ought to be exposed to mapscript.
I'll drop a line to -dev to see if there is other input before I proceed.