Re: Strange delays when running remotely

Dear Jeff, At 06:11 PM 7/1/2002 -0400, Jeff Mincy wrote: >On 02 Jul 2002, ville.skytta(a)xemacs.org wrote: > > > On Mon, 2002-07-01 at 23:57, David E. Sigeti wrote: > > > >> Dear Bug Team! > >> > >> When running Xemacs version 21 remotely (displaying on my local X > >> server) I have found that any multicharacter kills (such as kill-line, > >> kill-word, kill-ring-save, etc.) are extremely slow (maybe 15-30 > >> seconds per kill). I only see this delay when I am running over a DSL > >> connection where the ping times to the remote system are on the order > >> of 100 milliseconds or more. When my local machine is on the same LAN > >> as the remote machine (much shorter ping times), I do not notice any > >> unusual delay in multicharacter kills. I also do not see this problem > >> when running Xemacs version 20. > >> > >> My local X server when using the DSL connection is Exceed 3D running > >> under Microsoft Windows 98SE. When using the LAN, my local X server > >> is also Exceed 3D but running under Microsoft Windows 2000. > > > > I have also received exactly the same complaints sometimes, the server > > side XEmacs being 21.4.6+ on Red Hat 7.x and the "remote X client" > > X-Win32 [1] on Windows XP. > > > > [1] <http://www.xwin32.com/> > >Try setting interprogram-cut-function and interprogram-paste-function >to nil. Thanks for a much for the suggestion. I will try it when I get home this evening. I assume that setting these hooks to nil will make it impossible to cut and paste between Xemacs and other applications. I don't cut and paste between Xemacs and other applications very often, but when I do I really want the capability. I suppose that I can restore the cut and paste capability (at the price of terrible performance when killing) by resetting the hooks to the default values. Does this sound reasonable? Version 20 of Xemacs allowed cutting and pasting without suffering from terrible performance when killing. Is this problem just a bug that we can expect to be fixed in a later version, or is there something fundamental about the way cutting and pasting are being done now that makes performance terrible on connections with high latencies? Yours, David --- Dr. David E. Sigeti Phone: 505-667-9239 E-mail: sigeti(a)lanl.gov Surface mail: MS-F645, Los Alamos National Laboratory, Los Alamos, NM 87545 USA