Re: [Fwd: glib::GList:g_list_append(), g_list_last() performance problem]

[otaylor redhat com]

Andrew Bartle / 93011 <a_bartle stest ch> writes:

> Hi Gnomes,
> 
> As supplementary info, m'colleague indicates that this performance problem
> also exists in glib-1.2.8
> 
> Am I sending this to the right place now ? Still no answer from "the other
> place"

As Havoc said, these are inherently O(n) operations. GTK+-2.0 will
have the GQueue datatype which is double-ended queue, and has fast
add/remove at both ends of the list.

Regards,
                                        owen