Re: [evolution-patches] debugging for GW messages

[Sarfraaz Ahmed <asarfraaz novell com>]

Hi,

Allowing to build e-d-s with debug flags and enabling an env variable
for debugging info at runtime, i guess are 2 different things. I dont
see how they overlap eachother. 

Developers should still be given an option to either enable or disable
debugging at compile time. I would prefer keeping the --enable-debug
flag in the configure.in, so that it can be used as the developer wishes
[ i.e., use either --enable-debug or --disable-debug ]

Rather, getting the debug info by setting the env variable should be
independent of the --enable-debug option. I presume that the
--enable-debug option just sets the '-g' compiler flag [ which would
help in attaching the debugger ] and, by setting the env variable, the
code will spew some debugging info [ and would need code hacking to spew
any more info that i need at runtime for any new/experiemental code i
write ]

Just my 2 cents on this.

On Tue, 2004-03-30 at 20:25 -0700, Gary Ekker wrote:
> Speaking from a support perspective. It would be nice to have it the way
> it currently is. That is to not have it a compile time option. This
> would be like enabling the camel debug spew which is very handy when you
> are trying to figure out whether the problem is on the client or server,
> and what the problem is, etc. This is especially true given the fact
> that both the client and server are new implementations.
> 
> Just a thought.
> 
> -Gary
> 
> On Mon, 2004-03-29 at 11:29 -0500, Jeffrey Stedfast wrote:
> > On Mon, 2004-03-29 at 17:08 +0200, Rodrigo Moya wrote:
> > 
> > > On Mon, 2004-03-29 at 23:50 +0800, Not Zed wrote:
> > > 
> > > > ahaaah, yeah, it sounds beneficial, although i did misread it the first
> > > > time.  But its really up to you, since i wont be debugging it :)
> > > > 
> > > well, if we do what I said, you'll have to always use --enable-debug
> > > yourself (and everyone of course :-) to get basic debugging info for any
> > > evolution-d-s code. I am not talking about the GW stuff, which is
> > > already set to allow debugging via an evironment variable, but about
> > > disabling all debugging info (related to g_return*, g_assert, etc) by
> > > default and allow all of it when --enable-debug is specified.
> > 
> > disabling the g_return_if_fail() stuff sounds kinda bad, considering
> > how often I see spewage from those from various evolution components.
> > 
> > Jeff
> > 
> > > 
> > > cheers
> > > 
> > > _______________________________________________
> > > Evolution-patches mailing list
> > > Evolution-patches lists ximian com
> > > http://lists.ximian.com/mailman/listinfo/evolution-patches
> 
> _______________________________________________
> Evolution-patches mailing list
> Evolution-patches lists ximian com
> http://lists.ximian.com/mailman/listinfo/evolution-patches
-- 
Cheers
	Sarfraaz Ahmed <asarfraaz novell com>