< Previous by Date Date Index Next by Date >
< Previous in Thread Thread Index  

Re: [reSIProcate] Makefile woes


Remove the use of gperf - it's trivial to generate a hash that uses an
irrelevantly small amount of memory that is just as good.

make rule to build the file and if the build fails just touch the file. You
can do this with something like
   gperf foo && touch foo
 a more advanced version could check if gperf existed with something
conditions and `which gperf`


On 3/2/05 8:09 AM, "Fischl jason" <jason.fischl@xxxxxxxxx> wrote:

> Scott, 
> 
> Do you have a suggestion for how to resolve this issue in the Makefile itself?
> 
> Jason
> 
> 
> 
> On Wed, 02 Mar 2005 07:45:17 -0500, Scott Lawrence
> <slawrence@xxxxxxxxxxx> wrote:
>> On Tue, 2005-03-01 at 12:42 -0800, Fischl jason wrote:
>> 
>>> I'm not sure why this is happening since the MethodHash.cxx file was
>>> checked in after MethodHash.gperf. When svn retrieves the files the
>>> timestamp on the .cxx file should be later than the .gperf file so
>>> gperf should not be run. What seems to happen is that the timestamp on
>>> the files are made to be the current time rather than the timestamp
>>> when the file was checked in.
>> 
>> Subversion does not preserve timestamps.  The timestamp of the file is
>> the time it was checked out, not the time it was checked in.  This is a
>> fairly frequent topic of discussion on the subversion mailing lists, and
>> the developers make (I think) a very good case for why this is correct.
>> 
>> --
>> Scott Lawrence
>> Consulting Engineer
>> Pingtel Corp.
>> http://www.pingtel.com/
>> +1.781.938.5306 x162
>> 
>> 
> _______________________________________________
> resiprocate-devel mailing list
> resiprocate-devel@xxxxxxxxxxxxxxxxxxx
> https://list.sipfoundry.org/mailman/listinfo/resiprocate-devel