SA Bugzilla – Bug 1641
minor fixes in EvalTests
Last modified: 2003-03-18 06:37:24 UTC
2 buglets: - Razor was being loaded even with -L switch - AnyDBM_File module was not specified although we now use it in EvalTests
Created attachment 751 [details] fix to both bugs
Please open separate bugs even for trivial patches. I can OKAY only the first one (already committed), I don't know about the second. Is it really the right way to hardcode the {DB,GDBM,SDBM,NDBM}_File into the AnyDBM_File @ISA? It looks fishy to mee but I don't know anything about this code.
Created attachment 753 [details] fix for razor module-loading
Created attachment 754 [details] fix for other bug
Malte: that patch was not OKAY for 24 hours. OKAY: patch 754 (third attachment) is fine, this has been in HEAD for a while.
Actually, Malte -- first off, that's how AnyDBM_File works. But you're right, in a way ;) When I think about it, I have never seen an access-db that doesn't use the Berkeley DB format, so we don't need to use AnyDBM_File at all. Instead this new patch *just* uses DB_File, and protects against failing its load if that module is not available. I've tested that SA still works OK without DB_File being present, so this works, but I'd like an OKAY from Theo on it first...
Created attachment 755 [details] drops AnyDBM_File, uses DB_File directly instead
OKAY: the changes look ok to me for the accessdb change. however, that's not in 2.5, so you didn't need to wait for me anyway. ;) I'll apply the patch to my code to actually test it, but it looks like it ought to work. I used AnyDBM since I figured people may want the functionality without the standard tools, so they could use non DB_File modules. < shrug>
OK, the first patch (the is_razor2_available one) was applied quite a while ago. The second patch is applied to HEAD, which is the only place it's applicable. So this bug is closed.