One of our engineers has had some good results with QuickLZ. It probably wouldn't be hard to make a TQuickLzTransport in the style of TZlibTransport, though I'm not sure if we would be able to include it in the primary Thrift distribution for licensing reasons. We could probably post it on the Wiki, though.
--David Todd Lipcon wrote: > I don't think that's the same thing. The compression algorithm I'm > referring to is called "Zippy" - pretty sure it's referenced in one of > their papers on another topic - maybe the Bigtable paper? I know I've seen > it in a public document somewhere. Here's a blog post about it: > > http://www.c10n.info/archives/218 > > As far as I know, the details are secret. > > -Todd > > On Mon, 7 Jul 2008, Jeff Hammerbacher wrote: > >> are you talking about this paper? >> > http://portal.acm.org/citation.cfm?id=1325851.1325984&coll=&dl=&type=series&idx=SERIES11272∂=series&WantType=Proceedings&title=VLDB > <http://portal.acm.org/citation.cfm?id=1325851.1325984&coll=&dl=&type=series&idx=SERIES11272&part=series&WantType=Proceedings&title=VLDB> >> . >> >> On Mon, Jul 7, 2008 at 6:35 PM, Todd Lipcon <[EMAIL PROTECTED]> wrote: >> >>> Also no mention of their proprietary super-fast LZO-like compression, > which >>> is the most interesting thing IMHO that stubby/proto buffers have over >>> Thrift. >>> >>> -Todd >>> >>> >>> On Mon, 7 Jul 2008, Jeff Hammerbacher wrote: >>> >>> pretty interesting: >>>> >>>> > http://google-code-updates.blogspot.com/2008/07/protocol-buffers-our-serialized.html >>>> . >>>> no mention of stubby... >>>> >>>> >> >
