[Dev] unable to configure glib in chandler 0.1

Morgen Sagen morgen at osafoundation.org
Fri Apr 25 14:42:45 PDT 2003


I called David today to try and work out what is going on, and here is a 
summary:

When wxWindows is built on top of GTK, it asks gtk-config what include and 
link flags it should use when compiling wxWindows.
The version of gtk-config on David's FreeBSD machine (I don't know if this 
applies to all FreeBSD installs) says that wxWindows
should link to libintl, whereas the gtk-configs I have seen on RedHat do not. 
His libintl also links to his libiconv, which is apparently incompatible with 
one that we build.  When running Chandler, the dynamic loader was picking up 
the libiconv that we build and that was causing the error he was seeing. 
Removing our libintl and libiconv from the debug/lib directory allowed him to 
run Chandler.  That can suffice for the moment, but I am working on the 
long-term fix.

~morgen


--On Friday, April 25, 2003 10:11 AM -0700 David Paigen <paigen at heathen.com> 
wrote:

>
> Morgan wrote:
>
>> Come to think of it -- you already have enough code compiled to run Chandler
>> on your FreeBSD machine.  The glib that you're getting stuck on is not yet
>> being used by the existing Chandler codebase -- it's in preparation for the
>> upcoming Repository Access Protocol library that Lou Montulli is working on.
>> So in fact, assuming that you previously ran "hardhat.py -dB" from your 
osaf/
>> chandler/Chandler/ directory and it got as far as you indicated (up to 
glib),
>> if you now cd to your osaf/chandler/Chandler/ directory and run "hardhat.py
>> -db" (little b means just build in the current module, and not the module's
>> dependencies) you should then have enough built to run Chandler (which you 
can
>> do by running "hardhat.py -dx" from the osaf/chandler/Chandler/ directory).
>
> Didn't work.  I still have a problem with iconv.  First I tried the build:
>
> forte% time ../../../hardhat/hardhat.py -db
> Project path:  /opt2/src/chandler/Chandler_src_0.1/osaf/chandler
> Module path:   Chandler
> Module name:   Chandler
> HardHat log:   /opt2/src/chandler/Chandler_src_0.1/osaf/chandler/hardhat.log
>
> [ Chandler ]: Copying RunDebug to debug
> 0.186u 0.103s 0:01.68 16.6%     776+1357k 114+16io 28pf+0w
> forte%
>
> 1.6 seconds didn't seem like long enough.  But hey, no errors.  So I tried:
>
> forte% time ../../../hardhat/hardhat.py -dx &
> [1] 17683
> forte% Project path:  /opt2/src/chandler/Chandler_src_0.1/osaf/chandler
> Module path:   Chandler
> Module name:   Chandler
> HardHat log:   /opt2/src/chandler/Chandler_src_0.1/osaf/chandler/hardhat.log
>
> [ Chandler ]: Executing
> [ Chandler ]: Running Chandler
> [ Chandler ]: 
'/opt2/src/chandler/Chandler_src_0.1/osaf/chandler/debug/bin/pyth
> on_d','Chandler.py'
>
> Execution output - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
-
>
> Traceback (most recent call last):
>   File "Chandler.py", line 7, in ?
>     from application.Application import wxApplication
>   File 
"/opt2/src/chandler/Chandler_src_0.1/osaf/chandler/Chandler/application/
> Application.py", line 8, in ?
>     from wxPython.wx import *
>   File 
"/opt2/src/chandler/Chandler_src_0.1/osaf/chandler/debug/lib/python2.3/s
> ite-packages/wxPython/__init__.py", line 20, in ?
>     import wxc
> ImportError: /usr/local/lib/libintl.so.4: Undefined symbol "iconv_open"
> [5805 refs]
>
> Execution complete - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
-
>
> [ Chandler ] ***ERROR*** : Command exited with code = 1
>
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>  A program that HardHat spawned exited with a non-zero exit code.
>           Please view the file 'hardhat.log' for details.
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>
> [1]    Done                          ../../../hardhat/hardhat.py -dx
> 0.890u 0.188s 0:03.64 29.3%     1151+2509k 104+21io 91pf+0w
> forte%
>
>
> _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
>
> Open Source Applications Foundation "Dev" mailing list
> http://lists.osafoundation.org/mailman/listinfo/dev







More information about the Dev mailing list