On Mon, 4 Dec 1995, Chris Rode wrote:
>
> chuck_hines_at_vnet.ibm.com writes:
>
> } >>>>> "igb" == igb <igb_at_axnd01.cern.ch> writes:
> }
> } igb> Since I have had no response to my original problem, maybe
> } igb> someone can tell me where to look to try and understand what is
> } igb> going on. As it stands I cannot use fvwm2 (pl39) at all.
> }
> } I didn't respond initially because I hoped that another DEC Alpha user
> } would speak up. I've never heard or seen anything like this myself
> } (on any platform). I know that there are other Alpha users out there
> } that have been using the 2.xx series without seeing this.
>
> i'm using 2.0.39 on DEC alphas running OSF/1 2.1 and OSF/1 3.2, and have
> had no problems like this...
>
> } igb> Original mail was:
> } >>
> } >> The system is DEC alpha OSF1/3.2a , compiler was "cc -migrate" but
> } >> I have tried also with "cc" both with and without the optimiser -
> } >> same results.
> } >>
> } >> Anyone have any ideas??
>
> i compiled with gcc, but i doubt if that's the source of your problem...
DEC Unix 3.2x added a subtle dependancy on libots.so. This is a math library.
We've built many apps that link fine, start fine, but incure a run time symbol
not resolved long after the app has started. It does spit out a nasty message
to stderr, but depending on how you started fvwm, you may not see this message.
Furture more, gcc substitutes it's own math lib and we've also seen apps that
where compiled with gcc did not groak.
So add: -lots to the link line for fvwm... ;-(
curt smith
--
To unsubscribe from the list, send "unsubscribe fvwm" in the body of a
message to majordomo_at_hpc.uh.edu.
To report problems, send mail to fvwm-owner_at_hpc.uh.edu.
Received on Mon Dec 04 1995 - 14:15:31 GMT