Re: [evolution-patches] (groupwise provider) change in modification time

[Not Zed <notzed ximian com>]

I'd suggest one minor change:
if the stat fails, then use time(0), otherwise dont default the value to time(0) since you always overwrite it with the stat call's response anyway.

e.g.
              time_t mod_time;

       cache_file_name = g_strdup (folder->summary->summary_path);
       if (stat(cache_file_name, &buf) == 0)
           mod_time = buf.st_mtime;
              else
                     mod_time = time(0);

       g_free(cache_file_name);

also, get out of the habit of putting spaces before the final ';' on every line, it is a totally inconcistent style with anything else in evolution, or anything i've ever seen anywhere else for that matter.


On Thu, 2005-02-17 at 10:38 +0530, Parthasarathi Susarla wrote:
> Attached is the revised patch.
> The modification time now would be got from the summary file itself, so
> 'stat'ing of the summary file is required each time. 
> No more static variable
> Thanks 
> -partha
> On Thu, 2005-02-17 at 08:06 +0800, Not Zed wrote:
> >
> >Please use time(0) rather than g_get_current_time().
> >
> >Also, rather than using a static variable which wont be thread-safe,
> >just call time(0) before calling gmtime().
> >
> >This code then seems redundant  (its stating of the summary doesn't
> >seem that useful anyway).
> >
> >if (!mod_time) {
> >cache_file_name = g_strdup (folder->summary->summary_path) ;
> >stat (cache_file_name, &buf) ;
> >mod_time = buf.st_mtime ;
> >g_free (cache_file_name) ;
> >}
> >
> >tm = gmtime (&mod_time) ;
> >
> >On Wed, 2005-02-16 at 15:55 +0530, Parthasarathi Susarla wrote: 
> >>Hi,
> >>this patch updates the modification time that is sent with each
> >>getQuickMessage request.
> >>
> >>thanks,
> >>partha