10.1.3 bug fix candidates - DERBY-668 & 826?

[Myrna van Lunteren <m....@gmail.com>]

Hi,

I would like to get opinions on whether the fixes for the sysinfo issues
-  https://issues.apache.org/jira/browse/DERBY-668
   (sysinfo does not report the version of derby.jar if the
class[path] does not explicitly contain it (but it's loaded because of
the jar manifest))
and
- https://issues.apache.org/jira/browse/DERBY-826
  (sysinfo does not print the right information when Derby is not
loaded through the classpath (but through e.g. jvm ext dir loading))

are good candidates to add to the list for 10.1.3 fixes.

Is this too big a change for a minor release?
If not, I'll add them to the wiki & start working on it...

Thx,
Myrna

Re: 10.1.3 bug fix candidates - DERBY-668 & 826?

[Andrew McIntyre <mc...@gmail.com>]

On 6/6/06, Kathey Marsden <km...@sbcglobal.net> wrote:
> Myrna van Lunteren wrote:
>
> > So, in the end, I'll not be backporting these changes...
>
> I think so. These are great changes but it sounds  prudent to  wait for
> 10.2.  Thanks for looking at it.

Hi there!

Back from an extended vacation, and with an ear infection. Try flying
with one sometime. Not fun. :-/

I'd be comfortable porting the various sysinfo fixes to 10.1.3, but
there are currently two unresolved issues: I'd like confirmation from
Dan that DERBY-622 can be closed at this point, now that we have
test-specific policy files and the method calls that were causing
problems are wrapped in priv blocks. And I think the patch I posted
for DERBY-1273 should be committed, as it removes the need for the
getProtectionDomain permission, without losing any useful information.

I'm still working on a document describing the impact of the various
changes in different environments but its slow going - there are lots
of permutations of fixes / security manager enablement / jars in ext
dirs / jars vs. classes / etc..

Anyhow, don't give up on these changes for the 10.1.3 release just
yet. Getting the 10.1.3 release going is my main priority once I've
recovered and I'd like to see these fixes go into the branch if
there's agreement that it's acceptable.

andrew

(only 476 emails left to read... :-) )

Re: 10.1.3 bug fix candidates - DERBY-668 & 826?

[Kathey Marsden <km...@sbcglobal.net>]

Myrna van Lunteren wrote:

>
> So, in the end, I'll not be backporting these changes...
>
> Shall I take them off the wiki page too?

I think so. These are great changes but it sounds  prudent to  wait for 
10.2.  Thanks for looking at it.

Kathey

Re: 10.1.3 bug fix candidates - DERBY-668 & 826?

[Myrna van Lunteren <m....@gmail.com>]

On 5/30/06, Bryan Pendleton <bp...@amberpoint.com> wrote:
> > Is this too big a change for a minor release?
>
> Hi Myrna, I think it would be nice to port these fixes back to
> the maintenance release. I do not know of any remaining open
> issues with these changes. However, you'll want to have a look
> through the entire nest of changes that were done in this area
> because there are a bunch of interactions. In addition to the
> two bugs that you mentioned (668 and 826), I think you should
> have a look at 622, 1152, 1229, and 1273.
>
> I would be glad to help analyze the diffs and discuss the changes
> with you. Just let me know what I can do to be of help.
>
> thanks,
>
> bryan
>
>
>

I looked at the changes, and I feel it's too sensitive to backport these.
For one thing, like Bryan said, the fixes for DERBY-668 and DERBY-826
are interwoven/related with fixes for various other issues. Also,
DERBY-622 is still open, with a partial change in...(although maybe
that one can just be closed).

Another concern is that the fix for DERBY-668 requires additional
security privileges which may not be so relevant for running sysinfo
against embedded,  but may be more sensitive when running the network
server implementation. I don't want to cause any trouble with
upgrading existing apps.

So, in the end, I'll not be backporting these changes...

Shall I take them off the wiki page too?

Myrna

Re: 10.1.3 bug fix candidates - DERBY-668 & 826?

[Bryan Pendleton <bp...@amberpoint.com>]

> Is this too big a change for a minor release?

Hi Myrna, I think it would be nice to port these fixes back to
the maintenance release. I do not know of any remaining open
issues with these changes. However, you'll want to have a look
through the entire nest of changes that were done in this area
because there are a bunch of interactions. In addition to the
two bugs that you mentioned (668 and 826), I think you should
have a look at 622, 1152, 1229, and 1273.

I would be glad to help analyze the diffs and discuss the changes
with you. Just let me know what I can do to be of help.

thanks,

bryan

Re: 10.1.3 bug fix candidates - DERBY-668 & 826?

[David Van Couvering <Da...@Sun.COM>]

+1, I'd like to see these changes, if all the issues have been ironed 
out.  Thanks again to Bryan for working on this.

David

Kathey Marsden wrote:
> Myrna van Lunteren wrote:
> 
>>
>> Is this too big a change for a minor release?
>> If not, I'll add them to the wiki & start working on it...
> 
> Does anyone know if there are  any lingering issues regarding permission 
> errors or output changes when running under security manager that might 
> surprise users?  If those have all been resolved I think it would be 
> great to see these changes in 10.1.3.
> 
> Kathey
> 
> 
> 
> 
> 

Re: 10.1.3 bug fix candidates - DERBY-668 & 826?

[Kathey Marsden <km...@sbcglobal.net>]

Myrna van Lunteren wrote:

>
> Is this too big a change for a minor release?
> If not, I'll add them to the wiki & start working on it...

Does anyone know if there are  any lingering issues regarding permission 
errors or output changes when running under security manager that might 
surprise users?  If those have all been resolved I think it would be 
great to see these changes in 10.1.3.

Kathey