On 10/19/2004 7:33 PM, Ed L. wrote:

> On Tuesday October 19 2004 11:20, Jan Wieck wrote:
>> On 10/19/2004 10:10 AM, Andrew Sullivan wrote:
>> > On Tue, Oct 19, 2004 at 07:49:09AM -0400, Jan Wieck wrote:
>> >> transaction. A dummy update like "set relname=relname" is enough.
>> >
>> > It hasn't always been enough for me, is what I'm saying.  Also, Tom
>> > Lane once warned me that he thought there were still some lingering
>> > bits of code where the tg_enabled flag wasn't checked.
>> >
>> > A
>>
>> Which would be because nobody ever used that flag. With 1.0.3 available
>> and those changes only intended for 1.1, why not make Slony the "pusher"
>> to find all those spots, squeeze them out of the code, have new dot
>> releases for 7.3.x and newer and force users to upgrade to the latest
>> patchlevel in their release branch, if they want to use Slony 1.1 ?
> 
> I have some certain needs for using slony to replicate 7.3.4 into both 7.3.4 
> and into 8.0.  Am I hearing you correctly that my options for 7.3.4->7.3.4 
> and 7.3.4->8.0 replication are 1.0.3 ONLY?  (Unless I want to be the guinea 
> pig for finding the overlooked tg_enabled spots in the Pgsql code?)

1.0.3 will be your preferred version for that, as that is STABLE. The 
1.1 branch isn't even near BETA yet. If it ever get's there, it'd be a 
good thing (and a great contribution) to run it through your test lab to 
help us finding all bugs we possibly stirr up.


Jan

-- 
#======================================================================#
# It's easier to get forgiveness for being wrong than for being right. #
# Let's break this rule - forgive me.                                  #
#================================================== JanWieck at Yahoo.com #