Randy,
This is a good place for problems, it allows for more visibility and also for
more give-and-take. If you've followed this thread, you can see I've tried to
replicate your trouble with no success on any platform including aix (but
4.3.2 in my case, I do not have a 4.2 machine ). I have checked in two
changes to _im_image.c which debuted in 4.0.9 and might conceivably address
some of the problems but they are not involved in the native-dx supported
formats. I have run your scenarios and more (no extension, redundant
extension, (un)specific format, through purify on an sgi and dxexec comes
through it clean).
Please if you can, do a clean (-g debuggable) build, and perhaps do some
troubleshooting, briefly:
setenv DXROOT /usr/local/dx (you may need to set DXDATA, DXMACROS
too)
dx -edit -uionly &
cd /usr/local/src/dx/src/exec/dxexec
(gdb | dbx | xldb) dxexec
(the mechanisms for passing -r -B arguments to dxexec vary. for xldb they
go on the command line after dxexec, for gdb and dbx it is run -r -B ).
in ui, connect to running server, do your tests, report stack traces (e.g.
where) etc.
Have you considered upgrading to 4.3.2?
Regards,
Pete
Randy Polson wrote:
> Hello all.
> I wanted to check something so I went to my copy of dx4.0.6 on aix4.2.
> I'm working from the Save image dialog that comes from the file menu of an
> image. If I chose RGB format and enter a file name "water1" then it saves
> fine. If I enter the name "water2.rgb" it saves and then gives a
> segmentation fault and disconnect from server. The same behavior happens
> with the greyscale
> postcript, miff and tiff file types.
>
> -Randy
>
> p.s. is this the proper place for bugs, or should i report it elsewhere?
