[Bug 674008] Review Request: openrave - Open Robotics Automation Virtual Environment

bugzilla at redhat.com bugzilla at redhat.com
Wed Apr 6 10:34:41 UTC 2011


Please do not reply directly to this email. All additional
comments should be made in the comments box of this bug.


https://bugzilla.redhat.com/show_bug.cgi?id=674008

--- Comment #36 from Tim Niemueller <tim at niemueller.de> 2011-04-06 06:34:40 EDT ---
Hi Rosen. We've used OpenRAVE successfully during the RoboCup German Open 2011
last week. It has been integrated with Fawkes to drive a Katana arm. We're
preparing a demo video highlighting some of our work areas, I'll post a link
once it's ready.

I have to talk to my student, I think we've used
openrave-0.2.20-0.2.svn2227.fc14.x86_64 in the end. It still required a little
patch, due to bug #667294 (Boost was missing deprecated iostreams ctors). This
should have been resolved by now and I'm going to test it later today. Please
consider upgrading to non-deprecated calls in a future version.

Rosen, I have currently set to use single-precision floats to speed up OpenRave
a little. Do you think that has overly negative impact, or is that a good
default choice for a distro package? When compiling OpenRave at around 50% it
compiles openravepy_int.o, which takes more than 1.5GB of memory and is pretty
heavy on the system. Any chance of splitting this up in a future release?

We've been seeing some weird slowness on 32-bit systems with slow database
generation, while it ran fine with the identical (besides bitness) 64-bit
version. I'll discuss this with my team members tomorrow if this was an
intermediate problem and if so how it was resolved.

>From the packaging site I need to push updates for collada-dom (which is
basically ready and built, I just didn't have the time). I have to disable
assimp atm, until review bug #635511 is through. Once that is done, we get rid
of the ivcon integrated library, right?

After that, it leaves two libraries in there. We're clear that
convexdecomposition has no standalone upstream and it got patches, therefore
should stay. fparser has critical patches. Rosen, how is upstreaming going on
this? Maybe you can try again and tell them that'll bring fparser to the Fedora
universe as an incentive?

Peter, please have a look again at this review now, I think we've reached a
good state to continue, with the only things to resolve being assimp acceptance
and fparser patch upstreaming and eventually ripping it out of OpenRave. What
do you think?

SPEC has been changed in-place, the SRPM is at
http://fedorapeople.org/~timn/robotics/openrave-0.2.20-0.2.svn2227.fc14.src.rpm.
It contains the patch to work with old boost. It can build with or without
assimp, for the latter just install the review package. Once it has been
accpted I'll add it to the BRs.

-- 
Configure bugmail: https://bugzilla.redhat.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.



More information about the package-review mailing list