🚀 go-pugleaf

In article <1991Jun30.194112.13700@zia.aoc.nrao.edu> cflatter@zia.aoc.nrao.edu (Chris Flatters) writes:
>> Verifying environment ... finished
>> XView warning: invalid object (not a pointer), xv_set
>> XView warning: invalid object (not a pointer), xv_get
>> Error: CPS has not initialized or server connection failed

>The last of these error messages indicates that AnswerBook is using NeWS
>rendering in the viewer window.

A friend looked into this, he said he was _fairly_ sure this was for output
only, so something like ghostscript, or ralpage (a newer version of xps), or
DECWindow's DPS extention could do the rendering.  That is, as long as you
make something listen on the NeWS port and "do the right thing".  Too bad
Sun didn't make an X extention to find out where to talk to for your NeWS
server (it would be even better if it could tell you if the server did
PostScript output only, Display PostScript, or NeWS...).

>                                 Your users are going to have to run the
>X/News server.  Changing from the MIT Xsun server to xnews is not a
>significant change for your users: unless you have programs that rely
>on server extensions other than MIT-SHM all X11R4 clients (including
>window managers) will run as they did with Xnews.

Modulo X/NeWS bugs (some events in wrong order, XQueryBestSize returning
bogus results, and who knows what else).  This is not to say that X/NeWS
has more bugs then MIT's X (it might, it might not), but it is one of the
few servers that isn't too much like MIT's X inside; it's bugs are different.
Time to learn a new set.  Also last time I looked there was no shape extention
(which can be used for more then just round clocks...), however I don't
know about OW2.0 (I have it, I have checked a few bugs against it, but I
didn't look at the output from xdpyinfo, sorry).

>                                                   The only perceptible
>difference will be a decrease in speed on machines without GX boards
>and an increase in speed on machines with GX boards.

Runs my application slower.  GX, or not.  I only have a 16M box and get
_way_ more pageing with X/NeWS then X11R4.  I think it gets poorer results
on some xbench catagorys.  Not that xbench is a very good server benchmark
(it is a good graphic operation benchARK).  It does lots of the same
operation in a row, so smart servers with graphic accel's can re-use alot
of the same accel request.  And big servers will not see the effect of
pageing, because the best of 3 runs is taken and that will leave out the
time it pages, and it shouldn't page during the run (same code is executed
again and again on data that should go into more or less the same
pages...).

-- 
           stripes@eng.umd.edu          "Security for Unix is like
      Josh_Osborne@Real_World,The          Multitasking for MS-DOS"
      "The dyslexic porgramer"                  - Kevin Lockwood
"CNN is the only nuclear capable news network..."
    - lbruck@eng.umd.edu (Lewis Bruck)