i hope you and your family be ok really
and this too for all the derby world ;-)
i will not do that question about the backup system ;-) :-p
because i'm the first person who dont use it and by that mistake , i los 6 months of work
because i was trusted on hardware backup
but with storm there is no way to trust in anything
jajaajaja :-/ i luagh to no cry :-s
by the way i was read the lines that are in your blog
really , thanks to you and to every body that has the amiability to do this thinks
i mean contributions to all us with some comment , some project
and is true you are right , the english is not my language
i speak and write it like a parrot (tucan for example)
but i;m looky i read it very well ;-)
thats enought starting in the way that i read 24 languages
other not so good by i defend me in reading :-p
and if i cant read something there is my good friend Google :-p with the translator ;-)
thanks again
On 3/22/06, David W. Van Couvering <[EMAIL PROTECTED]> wrote:
Well that really, really sucks. It appears that my version of
DerbyLoader.java on disk is corrupt, and that is my *only* copy. Don't
get me started on why I don't yet have a working backup strategy.
I know what I need to do and can do it again, but that was a lot of work
down the drain. Sigh...
Basically what I had written was a "child-first" classloader that loads
from its search path *before* it loads from the parent. I don't think
you really need that, Deepa, your classloader works fine. The usage of
interfaces that you see in the rest of the code would work just as well
for a standard URLClassLoader.
David
Deepa Remesh (JIRA) wrote:
> [ http://issues.apache.org/jira/browse/DERBY-1075?page=all ]
>
> Deepa Remesh updated DERBY-1075:
> --------------------------------
>
> Attachment: derby-1075-v2.diff
> derby-1075-v2.status
>
> Thanks Andreas for pointing out the new case added to phaseTester. I had not noticed the change as it was committed after I uploaded my first patch. I have added this test to UpgradeTester and also updated the master. The new patch is uploaded as 'derby-1075-v2.diff'. With this patch, I have built and run the new upgrade test on WinXP using Sun JDK 1.4.2. Once this patch gets reviewed and committed, it will replace phaseTester.java and runPhases.ksh and we can remove these files.
>
> David, thanks for uploading your classloader work. I extracted the jar file using jar command but I find DerbyLoader.java in unreadable format. I am curious to understand what you have done but do not plan to work on it immediately. So whenever you get time, please check the jar file to see if you are able to read DerbyLoader.java.
>
>
>>Combine the script and the test available for upgrade testing into a java test
>>------------------------------------------------------------------------------
>>
>> Key: DERBY-1075
>> URL: http://issues.apache.org/jira/browse/DERBY-1075
>> Project: Derby
>> Type: Sub-task
>> Reporter: Deepa Remesh
>> Assignee: Deepa Remesh
>> Priority: Minor
>> Fix For: 10.2.0.0
>> Attachments: classloader.jar, derby-1075-patch1-draft.diff, derby-1075-patch1-draft.status, derby-1075-v1.diff, derby-1075-v1.status, derby-1075-v2.diff, derby-1075-v2.status, loader.java, loader_v2.java
>>
>>Separating a part of DERBY-514 into a sub-task. This sub-task is to combine the script 'runphases.ksh' and the java file 'phaseTester.java' into a single java test. This will help to later add the upgrade tests into the regression suite. Also, this sub-task can be done independent of DERBY-1049 which is blocking DERBY-514.
>
>
