Many appologies - I just haven't had the time to clean up the code on my
laptop with all my changes and then get them submitted for a commit by
somebody else. I have a few outstanding items to resolve before
submitting the code to check in. I know that you have been busy working
towards CTS
On Mar 14, 2005, at 7:07 AM, Mark wrote:
Many appologies - I just haven't had the time to clean up the code on
my laptop with all my changes and then get them submitted for a commit
by somebody else. I have a few outstanding items to resolve before
submitting the code to check in. I know that
I started looking through this also -- it looks very nice. However, I
have an ulterior motive :-) If I understand the stub/skeleton
generation code correctly, it does not currently handle overloaded
methods properly, it just assumes the IDL operation name is the same as
the java method
Finally got some free time to respond again :)
On Jan 24, 2005, at 7:28 PM, Mark wrote:
Thanks for the review. I will go through your comments one-by-one
over the next few days.
A few things to keep in mind:
1. The rmi/iiop code is fairly solid (with my fingers crossed). I
hope to get in a
Is it safe to say the we should use CGLIB instead of the
java.lang.reflect package?
Thanks
Mark
Again there is double check locking. Also, if this is called in
the critical path, I suggest we use CGLIB instead of reflection
for construction as it is way faster.
MD I can look into CGLIB.
On Jan 26, 2005, at 7:55 AM, Mark wrote:
Is it safe to say the we should use CGLIB instead of the
java.lang.reflect package?
For deployment/startup code it doesn't matter much. For code executed
during a request it is a good idea to use CGLIB.
I'm in favor of enhancing the classes during
Also does anyone have a URL to the spec? I personally would find it
convent to have a link in the files and in the Notice file.
-dain
On Jan 26, 2005, at 9:17 AM, Geir Magnusson Jr wrote:
On Jan 24, 2005, at 10:05 PM, Mark wrote:
The GIOP, IOP, IIOP were taken from the CORBA 3.x spec. I should
Corba 3.0.2 Spec is here:
http://www.omg.org/technology/documents/formal/corba_iiop.htm
Dain Sundstrom wrote:
Also does anyone have a URL to the spec? I personally would find it
convent to have a link in the files and in the Notice file.
-dain
On Jan 26, 2005, at 9:17 AM, Geir Magnusson Jr
Thanks for the link Jeff.
If you download the IDF file pack on that page, it contains just idl
files without any notice or license, but if you pull down the actual
CORBA spec it has a two page license right at the top. After a quick
read I think the most important part for us is (remember
Thanks for the review. I will go through your comments one-by-one over
the next few days.
A few things to keep in mind:
1. The rmi/iiop code is fairly solid (with my fingers crossed). I hope
to get in a bunch of test cases that following the Geronimo design.
2. The stub/skeleton generation is
Mark wrote:
The GIOP, IOP, IIOP were taken from the CORBA 3.x spec. I should
probably add a readme.txt to the interop/src/idl directory with the
locations of where the files came from.
I would suggest the NOTICE file, both in the iiop module and in the
project root, as that is the traditional
Mark,
Wow! This is very impressive work. I spent some time reading over this
(well the first half... the thing is huge). Most of the comments, I
have are just my curiosity and don't really need to be fixed. Other
are notes on differences between this code an geronimo, and I these
cases
12 matches
Mail list logo
