Uhm, having a program that can talk to 0.6 and 0.7 servers at the same time is not the hard problem, it took way less than five minutes to copy in both generated clients in the same project and rename the C# namespaces. Two apps and write to disk inbetween? Maven? That's crazy talk. :-D

What I was the most interested in is the fastest way to retrieve all rows through thrift. get_range_slices or get_multi_slices? And how can I be certain that I get all rows and the latest version of each row?

Say that I do get_range_slices, with some start token, a limit of 1000 and consistency level ALL, what will actually happen? Will I get the first 1000 rows from that token as they appear on the server I was talking to, or will the server fetch in additional rows that it doesn't have, but that fit the query?

If I do repeated calls to get_range_slices, and set the start_token to the last token of the previous slice, will I actually get all rows?


On Fri, May 6, 2011 at 09:07, Stephen Connolly <stephen.alan.connolly@gmail.com> wrote:

maven-shade-plugin could help with having two versions of thrift at the same time... but you'd need to build some stuff with maven, and some people don't like that idea

- Stephen

Sent from my Android phone, so random spelling mistakes, random nonsense words and other nonsense are a direct result of using swype to type on the screen

On 6 May 2011 01:11, "aaron morton" <aaron@thelastpickle.com> wrote: