defert@cernvax.UUCP (defert) (02/28/89)
Is there a reliable and efficient version of X11R2 or X11R3 running under SR9.7 and/or SR10.1 ? Here at CERN (Geneva) we have got ADUS X11R2 that runs not so quick but bearable and crashes often over 9.7. On 10.1 it is so slow that it is unusable. Do you know why it is so unreliable or so slow? Is it related to X itself or TCP/IP or ... ? Does somebody know when Apollo will support X ? Thanks for your help. Philippe. --------------------------------+-------------------------------------- Philippe Defert | SPS Application Software Group --------------------------------+-------------------------------------- Earth mail: | E-mail: CERN CH-1211 Geneve 23 | defert@cernvax.bitnet Switzerland. | defert@cernvax.UUCP Voice: | + 41 22 835579 | psi%02284681140551::defert
lnz@LUCID.COM (Leonard N. Zubkoff) (03/02/89)
Here is a likely reason the X11R2 server is slow on SR10.1. In gpr_tmgr.c you'll
find the code:
trait_$mgr_dcl (gpr_io_$uid, io_$trait, trait_$kind_local, &gpr_io_$epv, status);
Change it to
trait_$mgr_dcl(gpr_io_$uid,io_$trait,
(trait_$kind_t)(1<<(int)trait_$kind_local),
&gpr_io_$epv,status);
The problem is that the original code is bogus, but was never
checked until SR10.1. I had this same problem in a GPR type manager
I use in GNU Emacs.
Leonard