minshall@violet.berkeley.edu (11/25/87)
I have brought up (or at least tried to) X11 on my RT PC running ACIS 4.3 (whatever it is called) with an APA16 display. I have fixes 1-60 applied (in theory at least). When I run uwm (and maybe when I don't; but there isn't much functionality in NOT, so I haven't run that way much) I often get into a situation in which pieces of the screen will not get updated. In other words, these pieces remain constant, even if uwm wants to put, say, a menu there. These areas seem to be defined by "y in [ytop,ybottom]". Sometimes more than one "swath" will be so affected. Running "xrefresh" normally makes things happy again. I've never looked at X code (and would just as soon not at the present). Has anyone seen this bug, and if so, is there a fix? (I have fixes 1-60 installed on my machine.) Thanks, Greg Minshall
papa@pollux.usc.edu (Marco Papa) (11/25/87)
We have been running X11 on the IBM RT/PC with apa16 monitors since it was distributed, and have not experienced any problems. We did install all the fixes, EXCEPT #68. -- Marco Papa
grogers@uiucdcsm.cs.uiuc.edu (11/30/87)
The RT's hc compiler has a few bugs, one of which you are experiencing. What you need to do is find every place where a pointer to a structure is being incremented or decremented and dereferenced all at once and break these into separate steps. So, look for a++->x, a--->x, ++a->x, --a->x and change these to a++; a->x, etc. When I first installed X11 I had the same problem you are having and this fixed it up. Also, don't forget to run server/ddx/ibm/RTSETUP.
bilbo.geoff@SEAS.UCLA.EDU (Geoff Kuenning) (12/01/87)
> in which pieces of the screen will not get updated. In other words, these > pieces remain constant, even if uwm wants to put, say, a menu there. > These areas seem to be defined by "y in [ytop,ybottom]". Sometimes > more than one "swath" will be so affected. Running "xrefresh" normally > makes things happy again. > Has anyone seen this bug, and if so, is there a fix? (I have fixes 1-60 > installed on my machine.) I've seen it, but assumed it was a bug in my conversion of the "mfb" code to our system. I'm interested to learn it's universal. I haven't investigated it, though. Geoff Kuenning geoff@lcc.ucla.edu geoff@ITcorp.com