On behalf of the PMC I am delighted to announce Jess Balint as a new Calcite
committer. Jess has been making a steady stream of improvements to RelBuilder
and the RelToSqlConverter, making them more usable and higher quality. Please
give Jess a warm welcome to the project!
Jess, If you would
Thank you Julian for the intro, and also thanks to the PMC for inviting me.
I currently work at Dremio, and these past weeks/months I have been
focusing on improving our JDBC/ODBC experience, but I also have some fun
poking to the other parts of Calcite and Avatica, and hopefully I'll expand
Zhiqiang He created CALCITE-1641:
Summary: MATCH_RECOGNIZE syntax support for calcite
On Thu, Feb 16, 2017 at 2:34 PM, Julian Hyde wrote:
> On behalf of the PMC I am delighted to announce Laurent Goujon as a new
> Calcite committer. Laurent has been rounding out the product in several
> areas, including support for the standard set of built-in
Zhiqiang He created CALCITE-1644:
Summary: RelToSql for MATCH_RECOGNIZE
Issue Type: Sub-task
Zhiqiang He created CALCITE-1643:
Summary: SqlToRel for MATCH_RECOGNIZE
Issue Type: Sub-task
Calcite differs from Catalyst in many ways. First of all, Catalyst is
essentially a heuristic optimizer, while Calcite optimizers often combine
heuristics and cost-based optimization. Catalyst pushes down predicates and
projections to most data sources, while Calcite can often push down full
Playing with Postgres a bit it looks like its array types can behave in
Druid-y ways. The three queries below are what Druid would do with "select"
and "groupBy" queries against a multi-value column. I'm not sure how
standard this behavior is or if Calcite does the same thing, but this seems
I've been using Spark recently and have stumbled across a couple surprising
bugs/feature gaps. It got me curious about how Calcite would handle the
same scenarios. Basically, I'm wondering if Calcite would handle these
scenarios directly or if it would defer to the underlying runtime. I.E.,
on postgres I tried timestamp + interval, because it seems timetampadd is
JIRA title is updated
On Fri, Feb 17, 2017 at 12:51 AM, Julian Hyde wrote:
> On Postgres did you try
hongbin ma created CALCITE-1639:
Summary: issue for timestampadd month
Issue Type: Bug
Thanks for you reply.
I have confirmed on postgresql, it behaves exactly as mysql.
For ansi SQL 92 I could find http://www.contrib.andrew.cmu.edu/~shadow/sql/
sql1992.txt, however for SQL 2011 I can't find a counterpart source. Can
you please kindly show me a link? just to make sure
thanks a lot for your thoughts and advice!
I'd like to give a bit more background on why we were thinking about system
attributes and why we did not want to allow changing the time attribute.
In Flink, the event-time attribute is handled as meta data of each record.
Users do not have
On Postgres did you try TIMESTAMPADD or did you try timestamp + interval? I ask
because if we fix this, we are going to have to change the behavior of
timestamp + interval.
Also please change the JIRA case subject to something more descriptive.
For the SQL standard, search for "ISO/IEC 9075-2”
Yes I have, although it's disabled in Druid for now because of a bug:
https://issues.apache.org/jira/browse/CALCITE-1601. I haven't had time yet
to look into why that's happening.
On Thu, Feb 16, 2017 at 1:59 PM, Julian Hyde wrote:
> Yes, exactly.
> FLOOR( .. TO
On behalf of the PMC I am delighted to announce Gian Merlino as a new Calcite
committer. Gian is one of the main developers of Druid, and has been adding SQL
support to Druid over the last few months. The SQL support is based on
Calcite’s Druid adapter, and he has been contributing back
On behalf of the PMC I am delighted to announce Laurent Goujon as a new Calcite
committer. Laurent has been rounding out the product in several areas,
including support for the standard set of built-in functions, and improving
JDBC compliance in Avatica. Please give Laurent a warm welcome to
Mail list logo