I've had a look at the outstanding pull request queue again, and have either commented on, or merged them, into my private 'merge' branch, fixing up the merge conflicts that were inevitable given a long backlog of changes (currently about 100 changesets newer than upstream/master). I've done cursory testing with those that I've merged, but not extensive QA.
https://github.com/mdomsch/s3cmd/tree/merge The signurl patch needs a little work to keep it from breaking other commands. I don't understand why object_copy() had commented out the use of extra_headers. This lack of understanding gates my willingness to accept either of the two pull requests that want to "fix" that. I do think we need a new command that only changes metadata about an object, self-cp is not the right answer. Think chattr/chmod/chown. This pull request: support multibyte key name on getting object operation (https://github.com/s3tools/s3cmd/pull/92) scares me because I don't understand where normal vs unicode strings are used throughout the tool, or how they're stored and retrieved from S3. A seemingly simple 1-line patch feels to me like it would have deeper implications. Or maybe it's just a stupid 1-line bug. Does anyone else have better insight? Ongoing, I don't have time to do QA on all of these features and pull requests myself. I don't know how QA was done before, though I'm sure mludvig did a lot of it himself. I'm happy to see the number of proposed patches increase, and in general the patches have been of high quality (I've commented if I've thought otherwise). Most have been very reasonable requests to solve particular needs - all well and good. But lacking QA, I don't feel comfortable doing a formal versioned release. Who can help overcome this obstacle? I would also like to see s3cmd be able to perform the same kinds of tasks for other object storage systems, not just S3. My professional interest lies in having s3cmd work against the OpenStack Swift API as well as S3 (not just the Swift-exposed S3 API). This means moving all the code in s3cmd (the app) that reads or writes Amazon headers (only about 6 lines right now) into S3/S3.py or similar, at a start. It could mean renaming the files and objects from S3* to something else - but I don't want to go there yet. Food for thought. Thanks, Matt -- Matt Domsch Technology Strategist Dell | Office of the CTO ------------------------------------------------------------------------------ LogMeIn Rescue: Anywhere, Anytime Remote support for IT. Free Trial Remotely access PCs and mobile devices and provide instant support Improve your efficiency, and focus on delivering more value-add services Discover what IT Professionals Know. Rescue delivers http://p.sf.net/sfu/logmein_12329d2d _______________________________________________ S3tools-general mailing list S3tools-general@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/s3tools-general
