db-torque-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Brendan Miller" <bmil...@dotster.com>
Subject Re: Any reason not to move to 3.3 final yet?
Date Fri, 08 Feb 2008 17:19:08 GMT
>>
>> Will there be a do-all-the-stuff-that-doesn't-justify-another-version
>> version soon?  Can we have a 3.4 plan that does some of these "leftovers"
>> without waiting another year?  I've written lots of workaround code for
>> TORQUE-107, but we maintain our own copy of Village for TORQUE-8.
>> Getting rid of these types of issues would make development with Torque
>> a lot easier.
>>
> While I commend you on raising these issues and I know you are working on 
> providing test cases for them, the basic problem here is that these issues 
> have been raised very late in the 3.3 development cycle (which for many 
> reasons we do not want to extend any further) and they have not been 
> confirmed to exist via test cases (yes, you have provided one now and are 
> working on the other) or reports from other users.

That totally makes sense.  I wasn't necessarily looking to extend the
3.3 release cycle either.  I guess I should rephrase what I said earlier
as "What are the post-3.3 plans?"  I'm not looking for my "pet" release
(okay, maybe a little), but I guess I was looking for hope that the things 
I can fix or contribute toward can make it into some subsequent release.

I have used Torque for years, but never followed development as closely
as I do now.

> I have taken the following specific actions:
> 1. TORQUE-8: I have resurrected the Village test case, added a test case 
> for this issue and the problem does not occur.  I have also noted that the 
> documentation indicates that it the ResultSet objects are automatically 
> closed.

Yes, I read of this.  We earlier thought we were experiencing the
issues described in TORQUE-8, which is why I brought it to life with
the patch that I previously provided.  We were only running with a
patched Village on a small project that was query heavy and running
out of cursors.  At this point, I'm not sure if that was the problem
or not.  I'll eventually make it back to retesting (thanks for the test
case!) and see if I can re-create in a Village-only environment, and
re-open if necessary.  I have read the ResultSet documentation as well
and noted their "auto-close" behavior, so I'm not sure from where our 
cursor problems stemmed.

> 2. TORQUE-107: I have had a fairly close look and have been unable to spot 
> a code path that will lead to this problem.  If this is indeed a problem 
> then it has been there for a long time without being reported so I think 
> users will survive until the next point release.

I have had numerous cases that will exhibit the problem in our
application code; I simply need to sit down and write something
standalone or in the framework of the Torque test project (thanks
for the link, BTW).

> 3. TORQUE-108: Now that you have supplied a test case I am reasonably 
> comfortable with this going in - the decision is not mine to make but I can 
> make it an option on the release vote (i.e. there will be a secondary vote 
> as to whether or not to include this in 3.3 final).

Like I said, it doesn't have to be in 3.3 final if the project voters
decide it shouldn't.  My only hope is that it doesn't languish for a
long time before the next release.  I basically want to minimize the
length of time I have to maintain my own patchset.

Thanks for your feedback and direction.

Brendan

---------------------------------------------------------------------
To unsubscribe, e-mail: torque-dev-unsubscribe@db.apache.org
For additional commands, e-mail: torque-dev-help@db.apache.org


Mime
View raw message