Wunderbar!!
Thank you, Pierre, for spearheading this effort.
-Original Message-
From: Anuradha Hegde [mailto:anuradha.he...@esgyn.com]
Sent: Thursday, December 21, 2017 10:01 AM
To: dev@trafodion.apache.org; d...@trafodion.incubator.apache.org
Subject: RE: Congratulations!!
Excellent ne
nition of Trafodion as
> > one of the top-tier open-source SQL engines.
> >
> >
> > Many thanks, Pierre!
> >
> >
> >
> >
> > From: Hans Zeller
> > Sent: Thursday, December 21, 2017 12:09:21 PM
> > To: dev@trafodion.apache.org;
Hi Dave,
The simple reason is that the person who implemented the [first n] feature is
not a compiler developer.
Ideally, we would be aware of the [first n] throughout the compilation and have
a new required property in the optimizer that says "optimize for first N rows",
so that we could favo
ce.
This should catch existing [first n] + ORDER BY views since view composition
happens in the binder. Now the FirstN node will be generated there and the
existing Normalizer check will catch it and flag it not updatable.
What do you think?
Dave
-Original Message-
From: Hans Zeller [mai
quire an order?
Dave
-Original Message-
From: Hans Zeller [mailto:hans.zel...@esgyn.com]
Sent: Monday, January 8, 2018 3:28 PM
To: dev@trafodion.apache.org
Subject: RE: Anomaly with [first n] and ORDER BY
Hi Dave,
Overall, I like the idea of moving some of this logic into the optimizer.
.apache.org
Subject: RE: Anomaly with [first n] and ORDER BY
Hi Hans,
Thanks. I was looking just now at RelExpr::createAContextForAChild. It seems to
pass its own required property down to its left-most child (at least for
sorting). I'm confused why this isn't good enough.
Dave
-Origin
+1 (non-binding)
Based on testing the previous RC on an Amazon VM.
-Original Message-
From: Liu, Ming (Ming) [mailto:ming@esgyn.cn]
Sent: Thursday, January 25, 2018 7:22 AM
To: dev@trafodion.apache.org
Subject: [VOTE] Apache Trafodion release 2.2.0 RC 2
Hi to everyone in the Trafodi
Hi Steve, does this really justify an entire new round? All the code (ok, maybe
only 99.98 %) for this release was written in 2017. The fact that we voted in
2018, does that really justify an update to the copyright year?
-Original Message-
From: Steve Varnau [mailto:steve.var...@esgyn.c
Hi,
Transact-SQL (Microsoft SQL Server) is another SQL dialect that allows SELECT
without FROM.
I assume you would treat these three queries as equivalent and produce the same
plan for them?
select 1, 'a', ?, uuid();
select 1, 'a', ?, uuid() from dual;
values (1, 'a', ?, uuid());
T
Hi, just installed libuuid-devel on edev08. This should bring it in line with
the other edev machines. The others do not have uuid-c++ and uuid-c++-devel
installed, so I didn't do that on edev08, either.
Can you try again?
Hans
-Original Message-
From: Dave Birdsall [mailto:dave.birds.
Hi Ming,
You mention "Trafodion never did semantics checking in the parser?" Actually, I
think it is ok to raise errors in the parser when it is clear from the syntax
that there is an error. You should not look up any metadata information in the
parser, though.
There are many places in the par
Thanks for volunteering, Sean!
As the release manager, you'll be able to set the schedule. I'm sure the
community will understand that you also have a day job.
-Original Message-
From: Dave Birdsall
Sent: Thursday, April 19, 2018 1:48 PM
To: dev@trafodion.apache.org
Subject: RE: 2.3 Re
Hi,
My two cents: +1 on Dave's suggestion to clean up the tests. I think some of
these leftover tables happen when people comment out the cleanup code for
debugging and then accidentally check that change into git.
About speeding up regressions: I really like Ming's idea (
https://issues.apach
+1 on Venkat's proposal.
The Google style seems reasonably compatible with what we have already, both in
Java and in C++.
Here are a few guidelines that should help avoiding such large differences,
whether we adopt the Google style or not:
- Don't format entire files, especially not if there i
If you want to install on a single node test environment, you can take a look
at this script:
https://github.com/apache/trafodion/blob/master/core/sqf/sql/scripts/install_local_hadoop.
I hope someone else can comment on how to do this on a multi-node cluster.
Hans
-Original Message-
Fr
Hi Yuan,
No, unfortunately there are no join hints in Trafodion. Maybe you can try
cardinality hints to influence the optimizer to choose a hash join:
select * from t1 <<+ cardinality 1e9>> join t2 <<+ cardinality 1e6>> on
t1.a=t2.b;
Hans
-Original Message-
From: Liu, Yuan (Yuan)
Se
+1
-Original Message-
From: Xu, Kai-Hua (Kevin)
Sent: Thursday, July 26, 2018 1:12 AM
To: dev@trafodion.apache.org
Subject: Re: [VOTE] New Trafodion Reease
+1
Regards,
徐恺华(Kevin)
易鲸捷信息技术有限公司
+86 136 4197 5902
在 2018/7/26 10:50,“Jin, Jian (Seth)” 写入:
+1
Br,
+1
What we do right now is to run some critical tests for the first time during
the voting phase, which makes that process very inefficient.
I wonder how we could make sure those tests are run before the vote starts.
Maybe we should find volunteers to run them before the release and then just d
+1
That would be a very good idea, IMHO.
For code changes, that should make things really easy. For documentation and
web site changes, I hope it would not make the process any more complicated
than it already is.
Hans
-Original Message-
From: Pierre Smits
Sent: Monday, September 10
19 matches
Mail list logo