If this is related to a problem with DAT_PUT (as confirmed by Alan) it
seems to be that the error message is useless for exactly the reason I
reported back in March - none of the code in dat_put1x.f (or
dat_put0x.f) associates an error message with the error status.
e.g in dat_put1x.f:
call dat_size(loc, dims(1), status)
if (nval .ne. dims(1)) then
status = DAT__BOUND
else
call dat_putC(loc, 1, dims(1), values, status)
endif
This style of code leads to the error message below. Is there a reason why
error messages are not being attached? If not, can we include error
messages (I've been bitten by this before - you end up having to track
down the status code, looking up it's translation and then grepping
through likely looking code for the reason). Additionally, it would help
the first bit if the "STATUS set with no error" actually mentioned the
status value in the message.
Tim
On Wed, 3 Dec 2003, Starlink Software Librarian wrote:
> Tim, Alan,
>
> The specx test fails on the command,
>
> >> $rm l483core.sdf
> >> CONVERT-VAX-FILE l483core.dat
> !! ERR_/EMS_BEGIN: STATUS set with no error report (improper use of EMS).
> ! DAT_PUT: Error writing value(s) to an HDS primitive.
> -- SPECX#010 -W- Error opening file --
>
> Steve.
>
>
> On Fri, 28 Nov 2003, Tim Jenness wrote:
>
> >
> > Steve, Alan,
> >
> > Apologies but specx 6.8-1 is essentially broken in that you can't create
> > new maps. I've attached the files that I needed to fix to make it work.
> > Can you please put them in to your repository and put a specx 6.8-2 on the
> > software store and errata? (I fear this bug report is going to become
> > more common)
> >
> > Thanks
> >
> > --
> > Tim Jenness
> > JAC software
> > http://www.jach.hawaii.edu/~timj
> >
> >
>
>
--
Tim Jenness
JAC software
http://www.jach.hawaii.edu/~timj
|