leech@alanine.cs.unc.edu (04/30/89)
When the DEC 'dxsession' program runs my .X11Startup file, any error messages produced don't show up in the messages window. This makes debugging the shell script very difficult. Is there some way to get output from .X11Startup captured? DEC's version of xterm claims to recognize (e.g. xterm -help) the '-C' option to capture console output, but when the option is actually given, it is not recognized. Is there any other way to get console output in an xterm window? Finally, we've seen some very strange behavior from the 3100 X server. After killing the DEC window manager 'dxwm' and replacing it with uwm or twm (as required for the sanity of most users), moving a window sometimes leaves behind a residue which overlaps all other contents of the original window location. This residue cannot be disposed of by having the window manager refresh the screen, or changing the root bitmap - only logging off will get rid of it. This is a severe problem, apparently with the server. Any fixes/workarounds? Thanks! -- Jon Leech (leech@cs.unc.edu) __@/ UNDERWHELMING OFFER OF THE MONTH: "Please feel free to skip the payment on this month's statement. Normal finance charges will apply." - NCNB VISA
steve@umiacs.umd.EDU (05/01/89)
The -C is definitely missing, so far as I can tell. From the poking around I've been able to do (mostly by means of enlightened guessing), the only way to get console stuff to show up someplace other than on the console is to have something constantly read /dev/xcons. There is a 'xcons' program, but the man entry I found is obtuse, and my fiddlings with it didn't seem to produce much result. (I'd love to hear from someone else who could tell me just how DEC does pseudo-console stuff. From what I know so far, I like the Sun scheme better...) I don't have a clue on how to make .X11Startup problems show up in dxsession, or as to why you're seeing display droppings. I haven't yet seen any display droppings, though I haven't yet pounded on my DS3100's X stuff all that hard. I'm using uwm. Perhaps it's something that twm is doing? If you do get an answer, I would of course be interested in hearing it. -Steve Spoken: Steve Miller Domain: steve@mimsy.umd.edu UUCP: uunet!mimsy!steve Phone: +1-301-454-1808 USPS: UMIACS, Univ. of Maryland, College Park, MD 20742
jg@jumbo.dec.com (Jim Gettys) (05/01/89)
In article <7977@thorin.cs.unc.edu> leech@alanine.cs.unc.edu () writes: > Finally, we've seen some very strange behavior from the 3100 X >server. After killing the DEC window manager 'dxwm' and replacing it >with uwm or twm (as required for the sanity of most users), moving a >window sometimes leaves behind a residue which overlaps all other >contents of the original window location. This residue cannot be >disposed of by having the window manager refresh the screen, or >changing the root bitmap - only logging off will get rid of it. This >is a severe problem, apparently with the server. Any fixes/workarounds? > Try using the "Xcfb2" server on the unsupported kit (presuming you are on a color display). It did not see as much testing as Xcfb, but is later and has a region problem fixed (and does stipple fill much faster as well). - Jim Gettys
joel@gilroy.pa.dec.com (Joel McCormack) (05/19/89)
The first release of the DECStation 3100 server (god, can't we just call it a pmax, regardless of what the marketing people did to the name) had a very nasty region bug in it. Region and window computations use much faster code than that in R3 (su