Few things I ran into after doing a merge of the upstream/master branch
into my local branch:
1. Had to manually remove the 'ref' directory since mvn clean did not
remove it
2. sqlline jdbc connection:
sqlline -u jdbc:drill:zk=local -n admin -p admin
(previously, I was using jdbc:schema=parquet-local)
3. file path qualification (note the `backquote`.. this used to be "double
quotes" and the prefix 'dfs' for the file system.
select N_NATIONKEY from dfs.`/tmp/parquet/nation.parquet`;
Aman
On Tue, Mar 4, 2014 at 7:35 AM, Timothy Chen <[email protected]> wrote:
> Why not just use jira patch available list to track them?
>
> Tim
>
> Sent from my iPhone
>
> > On Mar 4, 2014, at 12:21 AM, Jacques Nadeau <[email protected]> wrote:
> >
> > Hey Guys,
> >
> > I've pushed a bunch of pending items into master. These changes include:
> >
> > - Hive metastore and serde read support
> > - Thinning out the SQL layer and moving SQL parsing out of the JDBC
> driver
> > into the server
> > - Moving the JDBC driver to work on top of avatica rather than optiq
> > - Renaming the StorageEngines to StoragePlugins.
> > - Updating/separating out storage plugins from file format plugins
> > - Fixes to avoid cast and _MAP[] syntax in queries
> > - Parquet and JSON field selection support
> > - Support for basic INFORMATION_SCHEMA capabilities
> > - Various outstanding bug fixes
> >
> > Please rebase your outstanding patches. Additionally, can you please
> > update this spreadsheet<
> https://docs.google.com/spreadsheet/ccc?key=0AvUC_YMxQ9UkdEdlYmswZ3ZXV19yZVEwWi0xcTRZZ3c#gid=0
> >so
> > we can collect what patches are outstanding and make sure someone gets
> > to them. We'll just use this to get caught up then move back to a JIRA
> > workflow (or PR request).
> >
> > Note, when you rebase, you'll likely need to update the storage parts of
> > your plans if you were working with file systems given the file/format
> > separation.
> >
> > Thanks,
> > Jacques
>
