Opened 11 years ago
Closed 11 years ago
#498 closed enhancement (fixed)
Resurrect the MD5 checksums
| Reported by: | Juergen Reuter | Owned by: | kilian, jr_reuter |
|---|---|---|---|
| Priority: | P3 | Milestone: | v2.2.0 |
| Component: | core | Version: | 2.1.1 |
| Severity: | normal | Keywords: | |
| Cc: |
Description
clear ...
Change History (9)
comment:1 Changed 11 years ago by
| Priority: | P3 → P1 |
|---|
comment:2 Changed 11 years ago by
comment:3 Changed 11 years ago by
2013/09/09: done for the phase space setup (default, i.e. "wood" method). See r4598.
To be done: reusing grids, when possible
comment:4 Changed 11 years ago by
What about events? They were reused without MD5 checksums? (i.e. demanding x events when y < x events are already present)
comment:5 Changed 11 years ago by
Events: yes, of course. Right now: reused without checking. Which is wrong.
comment:8 Changed 11 years ago by
| Priority: | P1 → P3 |
|---|---|
| Type: | defect → enhancement |
Events/simulate done in r4609.
Possible bugs aside, one minor issue remains: the 'rescan' command should check a separate MD5 sum which just covers the process configuration, no parameters. Currently, it doesn't check at all, which is good enough.
Keeping the ticket but ranking down, thus.
comment:9 Changed 11 years ago by
| Resolution: | → fixed |
|---|---|
| Status: | new → closed |
The separate MD5 sum for processes is done in r4630. Rescanning now works as long as the process definition is compatible, all other changes are ignored. This should settle the issue, hopefully.
2013/09/04: done for process components, processes, and process libraries. This is rather complex, since there are many different possible process/library states. However, all relevant cases are hopefully covered by r4578.
The next steps are checks for phase-space and grid files, which reuse the process checksums and introduce additional ones. Should be more straightforward, fortunately.