Yong Sun wrote: > Hi, James, > > Here are the time consumed (in seconds) of gnome-session starting up > with different input method module selected, the first column is for > the 1st login, and the 2nd column is for re-logins > > scim: 25.7 4.6 > iiimf: 25.9 4.9 > simple: 21.0 3.2 > > It appears that iiimf/scim takes about 5 more seconds in the 1st > starting, and 2 more seconds for re-logins, than without any IM > selected. And my iiimf gtk-im-module is a debug version thought. And > the reason of why first login is slow, is probably that the system is > loading and dynamic linking the executables and libraries, and > starting some necessary services that only need to be launched in the > 1st login, like gconf. > > Regards, > > James Cornell wrote: >> W. Wayne Liauh wrote: >> >>> "The cool thing is that with iiim disabled (from gnome sessions as >>> well as via .profile) gnome starts considerably faster for me." >>> >>> http://www.opensolaris.org/jive/post!reply.jspa?messageID=308935 >>> >>> It's a shame that we have to include this sh*t in the OpenSolaris >>> LiveCD, designed primarily to give the public a first impression >>> running on desktops, w/o any deliberation. >>> >> Agreed. For British, American, Australian the non-global CD shouldn't >> even have it enabled. Heck, there's some languages other than those >> which don't need it either, but Sun sure is pushing it. I have no >> interest in it, and it slows load by about 1 minute, and that's on a >> very very fast reader. >> >> James >> _______________________________________________ >> indiana-discuss mailing list >> [email protected] >> http://mail.opensolaris.org/mailman/listinfo/indiana-discuss >> > It's still unacceptable you know. But I will forgive it, because I understand there are other GNOME dependencies like gconf. Perhaps a debug version is shipped by default with most builds of Indiana so far, and that's what the overhead comes from. It's more than possible, since I've seen other products on OpenSolaris in general shipped with debug libraries and executables. Hopefully the real release won't be so bad.
James _______________________________________________ indiana-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/indiana-discuss
