Hi, ACE_LIB_TEXT should only be used in the ACE library itself. Please upgrade to ACE+TAO x.4.6, which you can download from ...
Johnny Willemsen
jwillemsen@...
Aug 1, 2005 8:27 am
19536
Hi, ... Thanks for using the PRF form. This version is already really ancient ... There is nothing in the RTEC to explicitly guarantee any order. If your ...
Johnny Willemsen
jwillemsen@...
Aug 1, 2005 8:29 am
19537
... So what's the "proper" way to destroy the ORB, so that it doesn't cause a segmentation fault? Even if I destroy nothing else, just the ORB, it's generating...
Frederick Heckel
fwph@...
Aug 1, 2005 9:36 pm
19538
Hi, ... That is possible. I'm currently looking into another, but potentially related, Any/TypeCode demarshaling problem. -Ossama -- Ossama Othman...
Ossama Othman
ossama@...
Aug 1, 2005 10:21 pm
19539
Hello Frederick, ... If (as the stack trace you sent in an earlier message seems to show) the problem is that the ORB is left with a dangling pointer to a...
EMalenfant@...
Aug 1, 2005 11:15 pm
19540
Hi, ... The problem was more that the naming service couldn't be found and then things crash because of not nice cleanup. The easiest is to start the naming ...
Johnny Willemsen
jwillemsen@...
Aug 2, 2005 2:57 pm
19541
... The naming service isn't a problem-- the programs I've been writing would've failed to run at all if I had not started with the -m 1 commandline. What I...
Frederick Heckel
fwph@...
Aug 2, 2005 5:25 pm
19542
Hi, ... Ok, that was not clear, you are talking about your app. ... Do you explicitly destroy the root poa? Johnny...
Johnny Willemsen
jwillemsen@...
Aug 2, 2005 7:59 pm
19543
... Frederick, I think the problem is related to servant reference counting; which apparently has changed over the years. Here is the recipe I use: Make sure...
Ken Sedgwick
ken@...
Aug 2, 2005 8:11 pm
19544
... Sorry about that. :) ... No, because I found that explicitly destroying the local/root poas or the orb would cause a segmentation fault immediately, as...
Frederick Heckel
fwph@...
Aug 2, 2005 10:12 pm
19545
... I'm presuming you mean TAO_ServantBase_var here? ... No dice; still getting the same segfault. I'll hack at this a bit more and try to see if I can dig up...
Frederick Heckel
fwph@...
Aug 2, 2005 10:57 pm
19546
Hi, sorry it took me so long. ... I packaged my sample program, which is adapted from tests from TAO. You may download it from...
Daniel Krueger
dani_tao-users@...
Aug 3, 2005 12:09 am
19547
Hi, ... No. PortableServer::ServantBase_var. TAO_ServantBase_var is the TAO-specific name which is not portable. -Ossama...
ossama@...
Aug 3, 2005 1:45 am
19548
Hi Frederick, ... I recommend that you run valgrind/purify on your code to see if something's awry with memory management. Thanks, Doug -- Dr. Douglas C....
Douglas C. Schmidt
schmidt@...
Aug 3, 2005 3:16 am
19549
Guys, what's the trick which tells the name server to not put ip names in the IOR and just the ip numbers? I'm having a tough time with a setup between XP and...
Stephen Adler
adler@...
Aug 3, 2005 10:58 pm
19550
Hi Stephen, ... Add "-ORBDottedDecimalAddresses 1" to the command line of your servers to force the ORB to encode IP addresses instead of host names in the...
Steve Totten
totten_s@...
Aug 3, 2005 10:59 pm
19551
Thats it!!! thanks. Is it the same for sun's java corba orb?...
Stephen Adler
adler@...
Aug 3, 2005 11:19 pm
19552
... This unfortunately didn't really tell me anything new. I've put up the output of valgrind http://www.cse.wustl.edu/~fwph/code/valgrind.output and the code...
Frederick Heckel
fwph@...
Aug 3, 2005 11:52 pm
19553
... No. The -ORBDottedDecimalAddresses option is specific to TAO. Sorry, I don't know if the Sun Java ORB has an equivalent option. Steve -- ... Steve Totten,...
Steve Totten
totten_s@...
Aug 4, 2005 12:52 am
19554
Hi, ... In imageserver.cpp: myHelper->registerName(buffer,myServer_var.in()); The CORBAHelper::registerName() method accepts a CORBA::Object_var instead of...
Ossama Othman
ossama@...
Aug 4, 2005 5:20 am
19555
Hello, TAO VERSION: 1.4.4 ACE VERSION: 5.4.4 HOST MACHINE and OPERATING SYSTEM: SuSE Linux 9.1 TARGET MACHINE and OPERATING SYSTEM, if different from HOST: ...
Charles de QUAY
charles.dequay@...
Aug 4, 2005 10:08 am
19556
There are more problems down this alley. In the initial IIOP handshake the server seems to put a an IP address on the return packer (or a DNS if...
Danny Loeb
danl@...
Aug 4, 2005 10:34 am
19557
CIAO VERSION: 0.4.6 TAO VERSION : 1.4.6 ACE VERSION : 5.4.6 HOST MACHINE and OPERATING SYSTEM: Win XP Professional, Winsock 3.10.0.103 CrossCompilation for...
Friedhelm Wolf
Friedhelm.wolf@...
Aug 4, 2005 11:58 am
19558
CIAO VERSION: 0.4.6 TAO VERSION : 1.4.6 ACE VERSION : 5.4.6 HOST MACHINE and OPERATING SYSTEM: Win XP Professional, Winsock 3.10.0.103 CrossCompilation for...
Friedhelm Wolf
Friedhelm.wolf@...
Aug 4, 2005 12:16 pm
19559
Schmidt
schmidt@...
Aug 4, 2005 1:09 pm
19560
Hello, my first email sent on this mailing list :) ... AFAICS, the problem is that your server is encoding the default IP address into its object references,...
Frank Rehberger
Frank.Rehberger@...
Aug 4, 2005 1:36 pm
19561
... The configuration I sent works fine and opens both the localhost address and an external address for listen on the same port for that ORB. In any case, if...
Danny Loeb
danl@...
Aug 4, 2005 4:17 pm
19562
Please forgive this intrusion, but I think that there is a better solution to your problem. Given that you have distinct roles to offer clients depending on...
Phil Mesnier
mesnier_p@...
Aug 4, 2005 4:24 pm
19563
Hi Danny, ... Using the endpoints you specified: arg[1]=-ORBListenEndpoints arg[2]=iiop://127.0.0.1:5010 arg[3]=-ORBListenEndpoints ...
Steve Totten
totten_s@...
Aug 4, 2005 4:46 pm
19564
... I agree ... My client connects to the server using "corbaloc::iiop:1.2@...:5010/<interface-name>". Note that the client knows at this point the...
