Re: [g-a-devel]at-spi event crack smoking ...

[Bill Haneman <bill haneman sun com>]

On Fri, 2002-08-09 at 17:34, Michael Meeks wrote:
> Hi Bill,
> 
> 	I was somewhat appalled to look at the at-spi/cspi/spi_main.c code and
> see what has been done to it.

I appreciate constructive comments about at-spi, but this sort of public
commentary is unnecessary and inappropriate.

Marc and I are the maintainers of this module.  As such it's perfectly
normal for us to commit changes without public review in some
situations, though I agree that public posting of patches is ideally a
better way to go, when other considerations such as urgency and time do
not mitigate against it.

-Bill

> 	This:
> 
> 	/*
> 	 * must ref before doing "direct pointer" identity comparisons,
> 	 * e.g. "accessible == a".
> 	 * Alternatively, one can use cspi_object_equal (a, b)
> 	 */
> 	Accessible_ref (event->source); 
> 		
> 	Is just rubbish.

I beg to differ.  Direct pointer comparisons may work for ORBit2 but
they do not work for the general case where object references may come
from other ORBs.  

> 	The way cspi_object_new is used is extraordinarly nasty; there are far
> easier, more efficient and transparent ways to do all of this, how did
> this get done this way with no review ? how about this:
> 
>       ref = malloc (sizeof (Accessible));
> //    ref->objref = CORBA_Object_duplicate (corba_object, cspi_ev ());
> 
> 	A C++ style comment ?

I am surprised that slipped by, sorry.

I do not think that the tone of your post is warranted by the source
issues you raise.  

-Bill
 
> 	Glancing at cspi-bonobo-listener.c (cspi_event), out of interest, how
> is the source, allocated by cspi_object_new ever freed ? I could go on
> and on ...
> 
> 	Basically, I look back at this code after a month or so away, and I
> find it littered with half finished bodges, much like it was before I
> cleaned it up. This code got to a point where it was clean, and fairly
> lucid and functional, it's falling back into the mire - why ?
> 
> 	This is extremely depressing; it is unacceptable for you not to agree
> to maintain at-spi co-operatively, and submit your patches to review,
> just incredible.
> 
> 	Marc - I'll re-write this code, to work if I get some time - and try to
> revert some of the last few months' breakage.
> 
> 	Regards,
> 
> 		Michael.
> 
> -- 
>  mmeeks gnu org  <><, Pseudo Engineer, itinerant idiot
> 
> _______________________________________________
> Gnome-accessibility-devel mailing list
> Gnome-accessibility-devel gnome org
> http://mail.gnome.org/mailman/listinfo/gnome-accessibility-devel