[comp.windows.news] Running NeWS on SG Iris workstations

zevpeter@aerospace.aero.org (q) (10/21/89)

I'm getting a strange error from the C side when I drive a
NeWS program on the Silicon Graphics Iris workstation with
cps.  The program runs perfect on a Sun-4 workstation.  Worse, 
the * example * program given in the Iris manual (a copy of 
the Sun example) bombs in the same exact way on two different
Iris workstations.

Basically, the loader dies with the error message "inet_addr
undefined"--as if it were looking for an Internet address!
The NeWS code itself works * fine * on both the Sun and the
Iris.  It's * just * when I use the C driver that it tries to
look for "inet_addr".  Does anyone have any ideas?  

(Also, some NeWS programs on Suns can't be ported too well to
Iris's...what are the important things one needs to watch out for
when writing NeWS code for use on the Iris?

ADVThanksANCE Very Much!!

pel@unhd.UUCP (Paul E. Leclerc) (10/22/89)

In article <59852@aerospace.AERO.ORG> zevpeter@aerospace.aero.org (q) writes:

>Path: unhd!uunet!snorkelwacker!usc!aero!zevpeter
>From: zevpeter@aerospace.aero.org (q)
>Newsgroups: comp.windows.news
>Keywords: NeWs inet_addr problems
>Date: 20 Oct 89 22:12:15 GMT
>Organization: back in civilization
>Lines: 19


>I'm getting a strange error from the C side when I drive a
>NeWS program on the Silicon Graphics Iris workstation with
>cps.  The program runs perfect on a Sun-4 workstation.  Worse, 
>the * example * program given in the Iris manual (a copy of 
>the Sun example) bombs in the same exact way on two different
>Iris workstations.
I tried the example in the Iris manual.  At first it didn't even
compile correctly.  I called SGI.  They (after doing some checking)
said they got their stuff from Sun and that it is wrong at the
source.  Summary:  Don't use the example in the manual.

>Basically, the loader dies with the error message "inet_addr
>undefined"--as if it were looking for an Internet address!
>The NeWS code itself works * fine * on both the Sun and the
>Iris.  It's * just * when I use the C driver that it tries to
>look for "inet_addr".  Does anyone have any ideas?  
Make sure to also link in the BSD libraries.   ie  -lbsd
and it should work.


>(Also, some NeWS programs on Suns can't be ported too well to
>Iris's...what are the important things one needs to watch out for
>when writing NeWS code for use on the Iris?
I am still struggling to learn NeWS and postscript (on any machine).
I wish I had a few simple examples of interfacing C and NeWS but
so far I have had no luck.  I understand that the Sun NeWS manuals
are good but alas we do not have access to them.

There is a NeWS archive at Sun.  Try mailing to
   news-archive@sun.com
with a subject of help


>ADVThanksANCE Very Much!!
-- 
===========================================================================
Paul Leclerc,  Computer Specialist III |Bitnet:    P_LECLERC@UNHH.BITNET
University of NH                       |UUCP:      ...uunet!unh!pel

ciemo@bananapc.wpd.sgi.com (Dave Ciemiewicz) (10/24/89)

In article <PEL.89Oct21143809@unhd.UUCP>, pel@unhd.UUCP (Paul E. Leclerc) writes:
> In article <59852@aerospace.AERO.ORG> zevpeter@aerospace.aero.org (q) writes:
> 
> >Path: unhd!uunet!snorkelwacker!usc!aero!zevpeter
> >From: zevpeter@aerospace.aero.org (q)
> >Newsgroups: comp.windows.news
> >Keywords: NeWs inet_addr problems
> >Date: 20 Oct 89 22:12:15 GMT
> >Organization: back in civilization
> >Lines: 19
> 
> 
> >I'm getting a strange error from the C side when I drive a
> >NeWS program on the Silicon Graphics Iris workstation with
> >cps.  The program runs perfect on a Sun-4 workstation.  Worse, 
> >the * example * program given in the Iris manual (a copy of 
> >the Sun example) bombs in the same exact way on two different
> >Iris workstations.
> I tried the example in the Iris manual.  At first it didn't even
> compile correctly.  I called SGI.  They (after doing some checking)
> said they got their stuff from Sun and that it is wrong at the
> source.  Summary:  Don't use the example in the manual.
> 
> >Basically, the loader dies with the error message "inet_addr
> >undefined"--as if it were looking for an Internet address!
> >The NeWS code itself works * fine * on both the Sun and the
> >Iris.  It's * just * when I use the C driver that it tries to
> >look for "inet_addr".  Does anyone have any ideas?  
> Make sure to also link in the BSD libraries.   ie  -lbsd
> and it should work.
> 

I can't comment on whether the examples work or not.  You don't even state
which example it is that you are using.

It does seem, however, that there may be the problem of not loading
executable with the proper libraries.  Try adding -lbsd to the
load command line (ld or cc).

> 
> >(Also, some NeWS programs on Suns can't be ported too well to
> >Iris's...what are the important things one needs to watch out for
> >when writing NeWS code for use on the Iris?
> I am still struggling to learn NeWS and postscript (on any machine).
> I wish I had a few simple examples of interfacing C and NeWS but
> so far I have had no luck.  I understand that the Sun NeWS manuals
> are good but alas we do not have access to them.
> 

Some problem areas could be in the area of the window manager code
which is derived from the litewin package with many mods.  As Sun
has claimed in their documentation that litewin was unsupported.

Also, there have been some assumptions made about the behaviour of
NeWS which are not necessarily supported in all implementations.
For instance, 4Sight (SGI's implementation of NeWS) does not support
retained canvases (I hope I'm right on this one).

If you are looking for a decent example, try the looking in
/usr/NeWS/clientsrc/glnews.*.  It is an example of a NeWS and GL
application communicating together.

There are also 4Sight/NeWS customization examples in
/usr/people/4Dgifts/.4sight/* which might be of interest to you.

Some of these things are in as early as 3.1D.  Some new features were
added in 3.2.

> There is a NeWS archive at Sun.  Try mailing to
>    news-archive@sun.com
> with a subject of help
> 
> 
> >ADVThanksANCE Very Much!!
> -- 
> ===========================================================================
> Paul Leclerc,  Computer Specialist III |Bitnet:    P_LECLERC@UNHH.BITNET
> University of NH                       |UUCP:      ...uunet!unh!pel