Thanks Paul. Best to have consistent formatting. Webrev updated in place.
On the beginning quote observation, I wasn't too such myself. Presumably
a typo. I've removed for now.
Regards,
Sean.
On 31/05/2016 19:48, Paul Benedict wrote:
Hi Sean,
I just have a few minor comments.
Nearly all the new messages follow the message/colon/space/details
format. There are a few missing the space between the colon and details:
*) ImageHeader:
"jimage header not the correct size:"
*) JrtPath
throw new ProviderMismatchException("path class:"
*) ConstructorFinder
all new messages
Some other observations:
*) ImageLocation: why the beginning quote?
throw new InternalError("\"Missing jimage attribute data");
*) JrtFileSystem: extra space before colon
"option class : " + option.getClass().getName());
Cheers,
Paul
On Tue, May 31, 2016 at 12:57 PM, Seán Coffey <sean.cof...@oracle.com
<mailto:sean.cof...@oracle.com>> wrote:
I've gone ahead with a trimmed down webrev as per Alan's request.
new webrev :
http://cr.openjdk.java.net/~coffeys/webrev.8151832.v2/webrev/
<http://cr.openjdk.java.net/%7Ecoffeys/webrev.8151832.v2/webrev/>
Regards,
Sean.
On 16/05/2016 15:10, Alan Bateman wrote:
On 16/05/2016 14:45, Seán Coffey wrote:
On 16/05/16 13:51, Alan Bateman wrote:
On 16/05/2016 13:44, Seán Coffey wrote:
Some extra capturing of context in exception
handling. I've re-based the original suggested
patch and added some minor edits.
https://bugs.openjdk.java.net/browse/JDK-8151832
webrev :
http://cr.openjdk.java.net/~coffeys/webrev.8151832/webrev/index.html
<http://cr.openjdk.java.net/%7Ecoffeys/webrev.8151832/webrev/index.html>
Would it be possible to leave out the changes to the
source files in the module and loader directories for
now? We have many changes to this code coming that
replace parts of this code and I think would be better
to do a pass over the exceptions in a few months once
the code is more stable.
Yes, I can hold off. I figured such improvements might
help people while they adopt and set up modules but let's
cancel until the code stabilizes some more!
Thanks. The rest mostly look okay although I think several of
these exceptions in the jimage code need to re-examined - for
example IOOBE is thrown in several places where the root cause
must be a corrupt or truncated jimage file.
-Alan
