Opened 17 years ago
Closed 16 years ago
#242 closed defect (fixed)
auto validate, test fit and mem failing.
| Reported by: | duncan | Owned by: | ole |
|---|---|---|---|
| Priority: | normal | Milestone: | |
| Component: | Architecture and API | Version: | |
| Severity: | normal | Keywords: | |
| Cc: |
Description
the fit is using more memory than previously. Could be related to ticket#241. Indicates a memory leak.
Change History (8)
comment:1 Changed 17 years ago by
| Resolution: | → wontfix |
|---|---|
| Status: | new → closed |
comment:2 Changed 16 years ago by
| Resolution: | wontfix |
|---|---|
| Status: | closed → reopened |
When caching problem in ticket:302 was fixed, this problem reappeared since the fitting algorithm again has to rebuild the mesh.
Now that caching can handle objects the the best solution would be to equip domain with a mesh instance instead of inheritance and pass that object into the cached fitting call.
Passing all of the domain does not work as quantity numbers change between calls.
comment:3 Changed 16 years ago by
| Owner: | changed from Duncan to ole |
|---|---|
| Status: | reopened → new |
comment:4 Changed 16 years ago by
Equipping domain with a Mesh object would imply a composition pattern rather than an inheritance pattern. This feels right in addition to solving the memory problem above.
comment:5 Changed 16 years ago by
| Component: | Appearance and visualisation → Architecture and API |
|---|
comment:6 Changed 16 years ago by
See changeset:5855 for last changes in quantity.py discussing this issue.
comment:7 Changed 16 years ago by
The transition from inheritance to composition of the Mesh-Domain relationship was implemented in changeset:6191
comment:8 Changed 16 years ago by
| Resolution: | → fixed |
|---|---|
| Status: | new → closed |
This was completed in changeset:6196 and changeset:6197
It is OK. Since generate mesh doesn't ask for as much memory as it used to, fitting has to ask for more.