On Mon, Mar 23, 2009 at 7:23 AM, Brian Candler <[email protected]> wrote:
> Scroll down to "Transactional Semantics with Bulk Updates" (either in > normal > view or raw view) > > After that section I see: > > ---- /!\ '''Edit conflict - other version:''' ---- > > Followed by a different version of the same para, then > > ---- /!\ '''Edit conflict - your version:''' ---- > > ---- /!\ '''End of edit conflict''' ---- > > I believe the second version is more accurate, but I look forward to seeing > some definitive information here. > OK, thanks, I didn't read down far enough. I fixed the conflict and made some more updates to make the overall section more clean and consistent, and added some more meat to the examples. > > Also, I have a particular question. The wiki says that the old atomic > _bulk_docs was removed because it could not be implemented on a sharded > cluster without resorting to two-phase commits. However, I cannot see how > the new semantics of _bulk_docs?all_or_nothing=true can be implemented > across such a cluster without two-phase commit either. > That's a good point. I am pretty sure what you actually get all_or_nothing_per_shard :). Since CouchDB doesn't do its own sharding right now, I am not sure if we need to document this? David > Regards, > > Brian. > -- David W. Van Couvering I am looking for a senior position working on server-side Java systems. Feel free to contact me if you know of any opportunities. http://www.linkedin.com/in/davidvc http://davidvancouvering.blogspot.com http://twitter.com/dcouvering
