postgres-xc-developers

[Postgres-xc-developers] Status of the current master

[Koichi S. <koi...@gm...>]

The master was updated with master_pg93_merge and it was found that we
still need some more work to stabilize it, some conflict and errors
found in DBT-1 benchmark run.

Please allow a bit more for stabilization.

My apology for the inconvenience.
---
Koichi Suzuki

Re: [Postgres-xc-developers] Status of the current master

[Michael P. <mic...@gm...>]

On Fri, Jan 31, 2014 at 10:29 AM, Koichi Suzuki <koi...@gm...> wrote:
> The master was updated with master_pg93_merge and it was found that we
> still need some more work to stabilize it, some conflict and errors
> found in DBT-1 benchmark run.
>
> Please allow a bit more for stabilization.
That's a development branch. It is made to be unstable :). You could
actually have done the merge of 9.3 directly on master IMO without
using the intermediate branch master_pg93_merge.
-- 
Michael

Re: [Postgres-xc-developers] Status of the current master

[Koichi S. <koi...@gm...>]

Yes, this is the mistake we've made in this merge.    The merge took
very long due to internal API change in the planner.    I planned to
make master branch as stable as possible and the merge can be seen as
a single commit.   This made all the process and communication almost
invisible.

I'm going to do the following after I stabilize the current master:

1) Merge with PG 9.3,
2) Merge current master with the latest PG master.

They may take a bit long and I'm planning to make intermediate status
visible to public so that I can have many more help.

Ideas or thoughts?
---
Koichi Suzuki


2014-01-31 Michael Paquier <mic...@gm...>:
> On Fri, Jan 31, 2014 at 10:29 AM, Koichi Suzuki <koi...@gm...> wrote:
>> The master was updated with master_pg93_merge and it was found that we
>> still need some more work to stabilize it, some conflict and errors
>> found in DBT-1 benchmark run.
>>
>> Please allow a bit more for stabilization.
> That's a development branch. It is made to be unstable :). You could
> actually have done the merge of 9.3 directly on master IMO without
> using the intermediate branch master_pg93_merge.
> --
> Michael

Re: [Postgres-xc-developers] Status of the current master

[Michael P. <mic...@gm...>]

On Fri, Jan 31, 2014 at 10:47 AM, Koichi Suzuki <koi...@gm...> wrote:
> This made all the process and communication almost invisible.
And this is a problem. It was hard to guess for long months what was
going on, with who doing what.

> I'm going to do the following after I stabilize the current master:
> 1) Merge with PG 9.3,
Looks good. You should not have much issues here, I don't recall that
*much* API changes since 9.3beta1.

> 2) Merge current master with the latest PG master.
You should wait for the moment when REL9_4_STABLE is out. Better to do
that once a year by experience.

> They may take a bit long and I'm planning to make intermediate status
> visible to public so that I can have many more help.
Each commit done should actually be justified on publicly visible MLs.
Regards,
-- 
Michael

Re: [Postgres-xc-developers] Status of the current master

[Koichi S. <koi...@gm...>]

2014-01-31 Michael Paquier <mic...@gm...>:
> On Fri, Jan 31, 2014 at 10:47 AM, Koichi Suzuki <koi...@gm...> wrote:
>> This made all the process and communication almost invisible.
> And this is a problem. It was hard to guess for long months what was
> going on, with who doing what.
>
>> I'm going to do the following after I stabilize the current master:
>> 1) Merge with PG 9.3,
> Looks good. You should not have much issues here, I don't recall that
> *much* API changes since 9.3beta1.
>
>> 2) Merge current master with the latest PG master.
> You should wait for the moment when REL9_4_STABLE is out. Better to do
> that once a year by experience.

I'd like to merge PG master into XC master more frequently to make
each work simpler and robust.    This is the background we changed the
release cycle from June to January(this time it is taking longer
though).

>
>> They may take a bit long and I'm planning to make intermediate status
>> visible to public so that I can have many more help.
> Each commit done should actually be justified on publicly visible MLs.
> Regards,
> --
> Michael

Re: [Postgres-xc-developers] Status of the current master

[Ahsan H. <ahs...@en...>]

On 31 Jan 2014 06:56, "Michael Paquier" <mic...@gm...> wrote:
>
> On Fri, Jan 31, 2014 at 10:47 AM, Koichi Suzuki <koi...@gm...>
wrote:
> > This made all the process and communication almost invisible.
> And this is a problem. It was hard to guess for long months what was
> going on, with who doing what.
>

True.  Going forward the merge with PG will be done incrementally so people
can get visibility along the way.

> > I'm going to do the following after I stabilize the current master:
> > 1) Merge with PG 9.3,
> Looks good. You should not have much issues here, I don't recall that
> *much* API changes since 9.3beta1.
>
> > 2) Merge current master with the latest PG master.
> You should wait for the moment when REL9_4_STABLE is out. Better to do
> that once a year by experience.
>
> > They may take a bit long and I'm planning to make intermediate status
> > visible to public so that I can have many more help.
> Each commit done should actually be justified on publicly visible MLs.
> Regards,
> --
> Michael
>
>
------------------------------------------------------------------------------
> WatchGuard Dimension instantly turns raw network data into actionable
> security intelligence. It gives you real-time visual feedback on key
> security issues and trends.  Skip the complicated setup - simply import
> a virtual appliance and go from zero to informed in seconds.
>
http://pubads.g.doubleclick.net/gampad/clk?id=123612991&iu=/4140/ostg.clktrk
> _______________________________________________
> Postgres-xc-developers mailing list
> Pos...@li...
> https://lists.sourceforge.net/lists/listinfo/postgres-xc-developers