[companion-integrate] RTI - Minor fix to xmms build

Steven M. Christensen sunfreeware at gmail.com
Tue Mar 25 01:36:02 PDT 2008


Mike -

I have redone the patch and Makefile.sfw as you suggest.  See

http://companion.sunfreeware.com/downloads/xmmswebrev/

As far as I can see, here is where fts.h appeared



4887 	17-Aug-2007  	chin 	6437624 RFE: Add ksh93 (as /usr/bin/ksh93) and 
libshell.so to OS/Net
6505835 AST tools and library (libpp) required for creating l10n 
messages for ksh93
PSARC/2006/550 Korn Shell 93 Integration
PSARC/2006/587 /etc/ksh.kshrc for ksh93
PSARC/2007/035 ksh93 Amendments
Contributed by Roland Mainz <roland.mainz at nrubsig.org



fts.h is in /usr/include/ast and elsewhere.  It is in 81
and perhaps in slightly earlier versions.

If the above fix is OK I will do the commit.  If this is appropriate for 
a bug report, I can do that as well if there is a good place for it.

Steve C.



Mike Sullivan wrote:
> Steven M. Christensen wrote:
>> This is a RTI for a minor fix to the build of the xmms package.
>> The configure step is now finding a fts.h file in recent versions of 
>> Nevada and this breaks the build.   The patch here removes the problem 
>> and a nightly in NV 84 for SPARC and x86 now works with no problem.
>>
>> http://companion.sunfreeware.com/downloads/xmmswebrev/
>>
>> This has been reviewed by Paul Cunningham.  See companion-discuss.
>> His suggested alternatives to this fix would not work and so
>> this original change has been kept.   I am still open to
>> some other method if it can be made to work.
> 
> So I have two concerns:
> 
> 1. I don't think I like this being called 'patch'. That's very
>    generic, it would seem better to have a more descriptive name
>    (or put it in the other patch, but I think I like separating patches
>    for different things). So maybe patch.removeftw_h?
> 
> 2. It almost feels like there should be a comment describing what
>    we are doing and why here (since we have no bug database yet
>    that could be referred to). Though I'm open to the argument that
>    it's obvious - but it still might be nice to identify the build
>    ftw.h appeared in maybe.
> 
> I'm flexible on #2, but if you at least fix #1 then then I approve :)
> 
>     Mike
> 


More information about the companion-integrate mailing list