Date: Thu, 06 Apr 1995 07:34:43 -0700 From: "Justin T. Gibbs" <gibbs@estienne.cs.berkeley.edu> To: Kai.Vorma@hut.fi Cc: "Jordan K. Hubbard" <jkh@freefall.cdrom.com>, "Rodney W. Grimes" <rgrimes@gndrsh.aac.dev.com>, jkh@freebsd.org (Jordan K. Hubbard), CVS-commiters@time.cdrom.com, cvs-ports@time.cdrom.com Subject: Re: cvs commit: ports/net/sup/patches patch-ae Message-ID: <199504061434.HAA23314@estienne.cs.berkeley.edu> In-Reply-To: Your message of "Thu, 06 Apr 1995 10:41:07 %2B0300." <199504060741.KAA00423@vinkku.hut.fi>
next in thread | previous in thread | raw e-mail | index | archive | help
>Justin T. Gibbs writes: > > > It also doesn't work when the "delete" option is turned on so it gives > > you a false sense of security. SUP will go ahead and delete the file > > you're working on. > >No. Look at the patches. Every unlink uses rname as which is the one >with #sup suffix. I did look at the patch. Very closely actually. It does not work for files that no longer exist on the sup server because the name in the last file is the one without the suffix, and we only stat files that the server says we need. Since the file that needs to be deleted is never mentioned by the server, we use rname which contains the wrong value (remeber, this is a separate traversal down the lastT, so rname cannot work because it is not updated again for each node, only during recvone). > > Please, just back this out for now. The real way to implement this is to > > piggy back it ontop of my "rename" option that i've had working and tested > > for months at TCS. I was just cleaning up the comments so I could submit > > it when this stuff allstarted to happen. The rename option takes care of > > all of the delete problems and you can simply write your feature as an > > extension to it (ie pretend that we are renaming the destination file to > > filename.#sup and the rest of the book keeping is handled automatically). > >That's fine. I wanted a simple but important feature for myself and >thought someone else would like it. If you have something better I I've never said that the feature wasn't valid or appropriate, its just not implemented well and to make it the default so quickly was a mistake. >will gladly use it. But before that happens I will use sup and my >patch because it works now. > >..vode > -- Justin T. Gibbs ============================================== TCS Instructional Group - Programmer/Analyst 1 Cory | Po | Danube | Volga | Parker | Torus ==============================================
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199504061434.HAA23314>