Thus said Kelly Dean on Tue, 06 Jan 2015 21:03:03 +0000: > That's horrible. You have to re-download the entire repository if one > packet is lost at the end?
If you're suggesting that Fossil should treat network errors differently during cloning and complete with whatever artifacts have been transferred, I think that would make sense in some cases. Fossil certainly can handle missing artifacts (they should show up as phantoms and then be requested the next time sync happens). But other things may need to be considered (like the config options that are transferred during cloning). Also, the user would need to be notified that a partial clone happened so they don't open the clone and start hacking on trunk, only to discover that it was trunk from 2 years ago. What I'm interested in discovering is why the clone completed and there was no project-id... Any chance you can use --httptrace when cloning with whatever is causing this problem? --httptrace will leave behind a handful of .txt files beginning with http- which have the requests and replies Fossil makes. > -rw-r--r-- 1 root root 2337792 Jan 6 19:48 fossil-src-1.fossil Can one of these can be made available? > # fossil pull -R fossil-src-1.fossil https://www.fossil-scm.org/ > Round-trips: 1 Artifacts sent: 0 received: 0 > Error: bad command: pull 9718e6fa34b5fa4860a40da1976fb087bc78d180 > > Round-trips: 1 Artifacts sent: 0 received: 0 > Pull finished with 2770 bytes sent, 251 bytes received Odd that you would get a bad command like that. I should have asked to have --httptrace added to one of these. Also, what OS is this running on? From the prompt it appears to be some *nix variant, but it could be Cygwin on Windows, or some such. Does there happen to be a proxy between your client and https://www.fossil-scm.org? Thanks, Andy -- TAI64 timestamp: 4000000054ac7366 _______________________________________________ fossil-users mailing list fossil-users@lists.fossil-scm.org http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
