Dirk wrote:
Toby Johnson schrieb:
Since it seems that the time I can put in with this project will
remain sporadic for the foreseeable future, I would like to alter the
criteria I'm using for 0.10-blocker tickets. My proposal is to hold
up the 0.10 release only for issues which cause an invalid dumpfile
to be created (and by "invalid" I mean one which causes "svnadmin
load" to fail).
This is ok with me. I think, that the next version is in the queue
already, since my PinBranch comes up very nicely. Yesterday evening it
failed on my broken archive file, but the logic up to this point is
tested, and I don't expect to many problems from there. But still this
branch could need more testing. I have to look into your sanity checker.
Glad to hear that part is coming along well. After 0.10 I will probably
move the sanity checker to its own class; it ended up being much bigger
(in terms of code size) than I first planned.
I also consider it a "last line of defense" against creating a bad
dumpfile, for those special types of VSS errors which we may never be
able to fully account for. Of course the preferred course of action
would be to correctly handle all special cases *before* that point, but
as I've come to realize, that just may not be possible.
_______________________________________________
vss2svn-users mailing list
Project homepage: http://www.pumacode.org/projects/vss2svn/
Subscribe/Unsubscribe/Admin:
http://lists.pumacode.org/mailman/listinfo/vss2svn-users-lists.pumacode.org
