On Thu, Jun 26, 2003 at 10:42:35PM +1200, Martin Langhoff wrote: > jw schultz wrote: > >Stop thinking client and server. Once the session is > >initiated there is little difference. Think instead sender > >and receiver. > > Good advise. Have to rephrase myself, then ;) > > As I read the patch, it seems to be running an arbitrary command on the > receiver. > > I am thinking of running an arbitrary command on the server, when it is > preparing itself to open the file and start feeding data to the receiver > (to the socket, should I say?). > > Am I on the right track? Help/advise? Is it doable at all?
That sounds fairly doable. I'm thinking what you would do is to create a temporary file to match/send and delete it on close. Essentially wrap the applicable open and close ops. If i recall correctly (i've not gone back to review that patch) that patch required whole-file transfers two avoid doing the operation twice per file. If the filter is deterministic and particularly if changes have minimal cascade effect you may be able to still take advantage of the rsync algorithm. Actually given this and the fact that a compressed file would be smaller it makes a bit more sense to do the filter on the sender. To answer your earlier question regarding a possible merge it is very unlikely that this sort of patch would be accepted into mainline of rsync 2.x. If it is a clean patch that looks to have wide use it could be included in the patches directory but the maintainers could not be expected to keep it up to date. -- ________________________________________________________________ J.W. Schultz Pegasystems Technologies email address: [EMAIL PROTECTED] Remember Cernan and Schmitt -- To unsubscribe or change options: http://lists.samba.org/mailman/listinfo/rsync Before posting, read: http://www.catb.org/~esr/faqs/smart-questions.html