Hi Fabrice,

On 08/26/2014 10:06 AM, Fabrice Jammes wrote:
> Build problem of Qserv master branch :
>
> I have next problem while building last Qserv version with g++ (GCC) 
> 4.8.3 20140624 (Red Hat 4.8.3-1)
> Next fix at line 23 of core/modules/xrdfs/SConscript.test solves the 
> problem :
> p = env.Program('testMySqlFs_1.cc', LIBS=["XrdXrootd", *"XrdUtils"*])
> This problem doesn't seems to happen with : g++ (GCC) 4.4.7 20120313 
> (Red Hat 4.4.7-4)
> Do you think we can apply this patch to Qserv master branch ?
I can't reproduce this easily, because I don't have a gcc 4.8 machine 
handy. Does the fix cause any problems for on RHEL6/Centos6 systems?

For now, perhaps it's better to comment it out rather than fix it--the 
xrdfs is likely to go away as soon as the dust settles on the xrootd 
port and the result protocol.

> Qserv dependencies versions management in eups :
>
> Furthermore, i see that Qserv master branch doesn't use a new xrootd 
> version.
> For now I don't know the right procedure to force a given Qserv 
> version to use a given version of its dependencies.
> It seems it is possible to do this with the rebuild command of lsstsw, 
> but how ?
> Please note that this problem seems to be the same than the one 
> described in qserv-l mail named : "Redundant(?) installation of doxygen"
I don't know how to solve this. I don't think it's a problem for new 
installations, because I think buildbot builds everything from scratch 
and it seems to build and test qserv without error. Perhaps your local 
eups stack doesn't know about the new xrootd package? Maybe there is 
something you can do with your installation to refresh package versions.

-Daniel

########################################################################
Use REPLY-ALL to reply to list

To unsubscribe from the QSERV-L list, click the following link:
https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=QSERV-L&A=1