Activity
From 05/27/2013 to 06/25/2013
06/25/2013
- 02:04 PM Enhancement #1551 (Resolved): Improve documentation of socket transport's "auto" mode
- For example, mention that @mode = auto@ and nothing else cannot work across multiple machines.
- 02:02 PM Enhancement #1550 (New): "Auto" mode of the socket transport should be smarter
- When configured as @server = auto, host = NOT-THE-LOCAL-HOST@, the process should not attempt to act as server.
No... - 01:59 PM Enhancement #1549 (Resolved): The documentation should mention that one Spread daemon is usually ...
- In many case, including distribution over multiple machines, a single Spread daemon is enough. In addition, configuri...
- 01:57 PM Enhancement #1548 (New): Transports returned by ParticipantConfig::getTransports() should be mutable
- @ParticipantConfig::getTransports()@ returns a @std::set@ of deep-copied @ParticipantConfig::Transport@ objects. As e...
- 11:50 AM Bug #1547 (Resolved): Unable to specify a manual converter selection strategy since cbd9e3c253
- Applied in changeset rsb-python|commit:63698ffcc9cd50be82c91dba77a1385368f1e1fe.
06/24/2013
- 04:56 PM Bug #1545 (Resolved): rsb-cpp-spread does not link on windows
- Applied in changeset rsb-spread-cpp|commit:d42f7e4c64b4704a12da435943982ca47c428a7f.
06/21/2013
- 02:34 PM Bug #1547: Unable to specify a manual converter selection strategy since cbd9e3c253
- I think the only way left now is by providing a complete configurator instance during the creation of participants, w...
- 02:32 PM Bug #1547 (Resolved): Unable to specify a manual converter selection strategy since cbd9e3c253
- With the changes introduced in commit:rsb-python|cbd9e3c253 my existing code broke because I was using something like...
- 08:52 AM Bug #1540 (Resolved): Documentation for installation of rsb-python from packages doesn`t work
- Applied in changeset rsb-manual|commit:ada183617495b666dbcf569aec33545e1f7250fe.
- 08:50 AM Bug #1546 (Resolved): Socket server dies because of incomplete message body
- NAO is sending images and a prorgram...
- 08:50 AM Bug #1173 (Resolved): Socket server dies because of incomplete message body
- From the comments I suspect this is resolved and I did not see such a crash again.
06/20/2013
- 11:00 AM Bug #1545 (Resolved): rsb-cpp-spread does not link on windows
- See wip-windowsfix and respective CI job
06/17/2013
- 07:16 PM Bug #1489: Selection of new fundamental converters breaks existing programs
- @Johannes: did you test this already? Please report results once you do.
- 11:28 AM Tasks #1541: Update Submodules and HEADs of these
- As long as we do not have gits or something running: should we add a jenkins job for automatically updating the submo...
06/14/2013
- 09:14 PM Tasks #1167 (Resolved): Provide Spread transport as plugin (C++)
- There has been at least some experimentation with this feature. Please report problems as separate issues.
06/12/2013
- 01:04 PM RSBag Tasks #1542: Document Python client API
- That's an old snapshot. The code lives in project:rsbag in the "rsbag-python repository":https://code.cor-lab.org/pro...
- 11:26 AM RSBag Tasks #1542 (Resolved): Document Python client API
- We should somewhere at least reference the existence of the Python binding, even if it is not officially supported ye...
- 02:46 AM Tasks #1538 (Resolved): Change "inter-transport" document to use socket transport
- Applied in changeset rsb-manual|commit:33406b7b853e0547a9ed42ebb7585180473b1bd9.
06/11/2013
- 01:05 AM Tasks #1541 (Resolved): Update Submodules and HEADs of these
- Applied in changeset rsb|commit:cab214794fbbf7187e74676c0c4c65b4059590b3.
- 08:59 PM Tasks #1541 (Resolved): Update Submodules and HEADs of these
- Not only the heads are outdated but also some new projects are missing, e.g. rsb-spread.
- 09:12 PM Tasks #1167: Provide Spread transport as plugin (C++)
- Works fine on Mac 10.7 built using the following new homebrew formula: https://github.com/swrede/homebrew-formulas/bl...
- 04:14 PM Tasks #1167 (Feedback): Provide Spread transport as plugin (C++)
- I think this is done. Please test and report problems.
- 02:42 PM Tasks #1167 (In Progress): Provide Spread transport as plugin (C++)
- Manual has been updated.
- 09:52 AM Tasks #1167 (Feedback): Provide Spread transport as plugin (C++)
- Done:
* The plugin is in the @rsb-spread-cpp@ repository
* I removed Spread from @rsb-cpp@ in a feature branch
* B... - 04:36 PM Bug #1537 (Closed): debian packaged rsb0.9 links against libspread, but doesn`t have spread as de...
- @librsb@ does not directly use Spread anymore. Instead, the Spread transport is implemented as a plugin (@rsb-spread-...
- 02:40 PM Bug #1540 (Resolved): Documentation for installation of rsb-python from packages doesn`t work
- The current installation docs for installation of rsb-python from packages are not working.
http://docs.cor-lab.de//... - 11:41 AM Tasks #1538 (Resolved): Change "inter-transport" document to use socket transport
- The example program will be changed and would no longer match the description.
06/09/2013
- 08:17 PM RSBag Feature #1534 (Resolved): Allow hard-wired foreign library paths
- Applied in changeset rsbag-tools-cl|commit:ba8f0d793c1e76fe2d21c5a11da034d29d810c38.
06/07/2013
- 07:41 PM Feature #1535 (Resolved): Allow hard-wired foreign library paths
- Applied in changeset rsb-tools-cl|commit:f363b72f4d3bedd5f6781dbeec7a36aeb528f779.
- 01:25 PM Feature #1535 (Resolved): Allow hard-wired foreign library paths
- This is necessary for enabling installations without LD_LIBRARY_PATH hacks in some cases.
- 04:20 PM Bug #1537: debian packaged rsb0.9 links against libspread, but doesn`t have spread as debian depe...
- We will turn the Spread transport into a plugin for the 0.9 release.
The best option therefore seems to be having
... - 04:13 PM Bug #1537 (Closed): debian packaged rsb0.9 links against libspread, but doesn`t have spread as de...
- @librsb.so@ from the @librsb0.9@ debian package links against the system @libspread.so@...
- 01:23 PM RSBag Feature #1534 (Resolved): Allow hard-wired foreign library paths
- This is necessary for enabling installations without LD_LIBRARY_PATH hacks in some cases.
06/06/2013
- 03:37 AM Feature #1442: RPC should support argumentless methods properly
- @Sebastian: I looked at this once and it seemed fine (if I remember correctly). Would you rather first finish everyth...
06/05/2013
- 08:26 PM Bug #1489 (Feedback): Selection of new fundamental converters breaks existing programs
- Should be fixed in rsb-python and rsb-cl.
rsb-python now registers only a subset of the available builtin converte... - 03:08 PM Bug #1489 (In Progress): Selection of new fundamental converters breaks existing programs
05/31/2013
- 01:47 PM Tasks #1167 (In Progress): Provide Spread transport as plugin (C++)
- 01:45 PM Feature #1521 (Resolved): LocalServer lacks a Callback<void,void> specialization
- 01:45 PM Feature #1521 (In Progress): LocalServer lacks a Callback<void,void> specialization
- An example can be found in the @examples/server@ directory.
- 01:45 PM Feature #1521 (Resolved): LocalServer lacks a Callback<void,void> specialization
- Applied in changeset rsb-cpp|commit:2eefc46c2880bea316bb9e6711d47173514a7bec.
- 12:40 PM Feature #1521 (In Progress): LocalServer lacks a Callback<void,void> specialization
- 01:11 PM RSBag Feature #1481 (In Progress): Support changing timestamp used for indexing in bag-transform
- @bag-merge@ already has the @--transform-timestamp@ option. I have to investigate whether anything needs to be done.
05/30/2013
- 10:04 AM Feature #1521: LocalServer lacks a Callback<void,void> specialization
- I already did this once but it did not quite work for some reason I can't remember. I will try again.
05/29/2013
- 04:00 PM Bug #1497 (Resolved): Receive and Deliver-Timestamps not set correctly
- Looks correct to me now.
- 02:06 AM Bug #1497 (Feedback): Receive and Deliver-Timestamps not set correctly
- @Johannes: Please check whether these changes fix the problem for you.
- 02:51 PM Feature #1521 (Resolved): LocalServer lacks a Callback<void,void> specialization
- Currently it is not possible to register a method with no argument and no return value to a rsb server.
Solution: ...
05/28/2013
- 12:10 AM Bug #1497 (In Progress): Receive and Deliver-Timestamps not set correctly
- 03:27 PM Bug #1519 (Resolved): Option for controlling TCP_NODELAY has wrong name in CL implementation
- Applied in changeset rsb-cl|commit:dcdd182c41b757b0c845ebf3744ebbac97454732.
- 01:18 PM Bug #1519 (Resolved): Option for controlling TCP_NODELAY has wrong name in CL implementation
- The option is called @nodelay@ instead of @tcpnodelay@.
- 03:08 PM Bug #1408 (Resolved): Converter registration should be possible without rewriting default partici...
- Applied in changeset rsb-python|commit:8ec3b1e64674144c7f1165338d82ce35dbf6d447.
- 11:23 AM Bug #1408: Converter registration should be possible without rewriting default participant config...
- I have pushed respective changes to the implementation and the manual. Can someone please review them and verify the ...
- 01:40 PM Bug #1517 (Resolved): TCPNODELAY is off by default in C++
- Applied in changeset rsb-cpp|commit:7672063f86aada6f1056c05599803bd11e30e37f.
- 11:35 AM Bug #1517: TCPNODELAY is off by default in C++
- I think so as well. I just didn't get to it yesterday.
- 11:34 AM Bug #1517 (Resolved): TCPNODELAY is off by default in C++
- TCPNODELAY is off by default in C++, although it should be on, right?
- 01:04 PM Bug #1518 (Resolved): RSB_TRANSPORT_SOCKET_NODELAY documentation seems erroneous
- Applied in changeset rsb-manual|commit:c0f78250c62c74c96a7271043b0b617760af65b6.
- 12:47 PM Bug #1518: RSB_TRANSPORT_SOCKET_NODELAY documentation seems erroneous
- Forgot to paste the link:
http://docs.cor-lab.org/rsb-manual/trunk/html/common.html - 11:51 AM Bug #1518 (Resolved): RSB_TRANSPORT_SOCKET_NODELAY documentation seems erroneous
- In the RSB manual, @tcp_nodelay@ for the socket transport is documented to be set through the environment variable @R...
- 11:31 AM Feature #1149: Allow checking if handlers are registered in Listener
- Martin, Robert, are you planing on implementing the proposed changes or should we do this?
05/27/2013
- 09:40 PM Bug #1408 (In Progress): Converter registration should be possible without rewriting default part...
- 06:31 PM Enhancement #1516 (New): Separate Converter selection strategies for sending and receiving
- Both directions differ significantly. This was already the reason why the connector infrastructure is split into two ...
- 06:29 PM Enhancement #1515 (New): Add data type to all participants, especially listeners
- This would automatically allows to select the correct converter for deserialization. Also, right now the situation is...
- 06:26 PM Bug #1489: Selection of new fundamental converters breaks existing programs
- * For python this means: only register the converters for these types
* For C++: Clients are obliged to use these cl... - 06:21 PM Bug #1489: Selection of new fundamental converters breaks existing programs
- Verdict: We want the same defaults for all languages. This means there is one int and one float type that is consiste...
- 05:21 PM Bug #1497: Receive and Deliver-Timestamps not set correctly
- Generally the problem does not exist in the core but is merely a display problem of the cl logger. What happens is th...
- 04:55 PM Bug #1230 (Closed): Unintuitive error message upon misconfiguration
- The current behavior is OK in all cases Johannes and I could test on MacOS and Linux (But we can still document this,...
- 04:55 PM Tasks #1513 (Resolved): Document exception handling in main function of RSB-using C++ programs
- Each RSB client program should handle exceptions at least in its main method because on certain OSes the exception ty...
- 01:09 PM Tasks #1167: Provide Spread transport as plugin (C++)
- As a separate repository for the single language implementation.
Also available in: Atom