On Thu, Jul 17, 2014 at 12:45 PM, Magnus Hagander <magnus@hagander.net> wrote:
> On Wed, Jul 16, 2014 at 2:01 PM, MauMau <maumau307@gmail.com> wrote:
>> From: "Amit Kapila" <amit.kapila16@gmail.com>
>>
>> So as a conclusion, the left over items to be handled for patch are:
>>>
>>> 1. Remove the new usage related to use of same event source name
>>> for registration from pgevent.
>>> 2. Document the information to prevent loss of messages in some
>>> scenarios as explained above.
>>
>>
>> I noticed the continued discussion after I had prepared and submitted the
>> revised patch. OK, how about the patch in the previous mail, Magnus-san?
>
> I mean that the -e option is valid for *all* commands, not just
> register/unregister. If you include it in register, pg_ctl will write
> it to the commandline used for start -- so clearly it is valid for the
> start command as well. And it is certainly possible for a completely
> different service to run pg_ctl start, in which case it will also be
> used.
>
> I have updated the patch with this change, so please verify that it
> still works. I also rewrote the documentation slightly.
>
> With that, applied. Thanks!
Did anyone actually test this patch? :)
I admit I did not build it on Windows specifically because I assumed
that was done as part of the development and review. And the changes
to pg_event.c can never have built, since the file does not include
the required header.
I have reverted that part of the patch for now, hopefully that'll
unbreak the buildfarm.
For the time being at least I left the other changes to DEFAULT_EVENT_SOURCE in.
-- Magnus HaganderMe: http://www.hagander.net/Work: http://www.redpill-linpro.com/
