Re: cvs commit: httpd-2.0/server error_bucket.c

[Ben Hyde <bh...@pobox.com>]

I don't really do much coding these days, you've probably noticed.

But that commit pulls my cord just a little.

Have you tried gid, it's part of the gnu's id-utils package.

You install it, then you run mkid at the root of your source
tree.  Then you do 'gid foo'  and it will _instantly_ show
you ever occurance of foo in the sources.  If your an emacs
user M-x gid will create a buffer you can step thru like
compile errors to edit every occurance.

That makes it really easy to propogate a change to all the
'call sites' when your making a change like this.  and it
avoids leaving the repository in an inconsistent state.

But, it's reall benefit it is it changes completely how
you work on the code - because instantly is very different
from 'find . | grep '\.[hc]' | xargs grep ...' !

 - ben

jwoolley@apache.org wrote:
 > jwoolley    2002/06/02 13:19:54
 > 
 >   Modified:    server   error_bucket.c
 >   Log:
 >   Propogate apr-util change
 >   
 >   Revision  Changes    Path
 >   1.17      +1 -2      httpd-2.0/server/error_bucket.c
 >   
 >   Index: error_bucket.c
 >   ===================================================================
 >   RCS file: /home/cvs/httpd-2.0/server/error_bucket.c,v
 >   retrieving revision 1.16
 >   retrieving revision 1.17
 >   diff -u -d -u -r1.16 -r1.17
 >   --- error_bucket.c	31 May 2002 20:52:28 -0000	1.16
 >   +++ error_bucket.c	2 Jun 2002 20:19:54 -0000	1.17
 >   @@ -86,7 +86,6 @@
 >        h->data = (buf) ? apr_pstrdup(p, buf) : NULL;
 >    
 >        b = apr_bucket_shared_make(b, h, 0, 0);
 >   -    b->is_metadata = 1;
 >        b->type = &ap_bucket_type_error;
 >        return b;
 >    }
 >   @@ -104,7 +103,7 @@
 >    }
 >    
 >    AP_DECLARE_DATA const apr_bucket_type_t ap_bucket_type_error = {
 >   -    "ERROR", 5,
 >   +    "ERROR", 5, APR_BUCKET_METADATA,
 >        error_bucket_destroy,
 >        error_bucket_read,
 >        apr_bucket_setaside_notimpl,