[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

19990708: Garp2.02

Steve et al.,

I have built Garp 2.02 on all our platforms- here are the basic
picky things different compilers choke on:


1) In guiutil.c, function SetDefaultOptionButton is defined to
return an int, but does not set a return value if succeeds.
Added return 0; to end of function.

2) In stringtools.c,ToUpper and ToLower defined as returning int,
do not set return value if string == NULL. Changed to:
   if ( string == NULL ) return 0;

3) In renderimg.c line 334, OSF flags the following line as undefined
    behavior

   *imgptr++ =  *(imgptr - xwdth);

   presumably, this is because imgptr may be incremented before its use
   in the calculation. Everything appears to work using no optimization.


In time_cb.c, 
void GetClockString ( char* timeformat, char *string )

timeformat is passed but never used. Presumably this will be used
in the call to 
strftime ( string, 256, "%a %b %d %H:%M:%S UTC %Y", timeptr);

The calling routine:
 void UpdateStatusBar( WindowObjectType *wo )
Never initializes time_format leading to uninitialized pointer messages. 
I added to the declaration:
char            *time_format=NULL;


I can build on HPUX 11.0 but get a X_polyfillrect error upon launching
that is triggered by the xw.a routine xgsdat.c in the Atom interning.
I can run using a HPUX 10.20 binary...seems to be something HP might
have broken in the upgrade to Motif 2.1 and X11R6.3.

------------------------------------------------------------------------
This may not be of concern to you since its a Gempak problem...

I ran into a problem displaying the GINI composite GMS/Meteosat/GOES
composite images on Channel 4 of NOAAport.

I found the following problems-

1) in $GEMPAKHOME/source/gemlib/im/imgi2gm.f:

The composite product is being sent with a satellite number=6
in Octet #2. There is no option for this...and it is not a documented
id in the interface control document I have. Don't know if its
supposed to be GMS, or just a composite flag. Anyhow, I added:
C
C*      Set the satellite number. (Octet 2)
C*
C*         Value       Sat
C*      -----------------------
C*           6       GMS, METEOSAT, GOES composite
C*           9       METEOSAT
C*          10       GOES-7
C*          11       GOES-8
C*          12       GOES-9
C*          13       GOES-10
C*          14       GOES-11
C*          15       GOES-12

and
           ELSE IF ( isat .eq. 6 ) THEN
            imsorc = 70

I just used the source for Goes-E to use those defined color tables.


2) The lov value being transmitted in the header is 2492048
   (octets 28-30), which corresponds to a clon = -110.7952.

   The above clon results in an incorrect mapping. Looking at the
   NESDIS documentation, this is supposed to be AWIPS 201 projection
   which specifies clon=-105.000. I will ask Mary Glackin about the
   discrepancy. I hardcoded the lov value 2550000 for this projection and
   sector id #10 (Octet 3) temporarily.

Otherwise, all of the GINI images we have been pulling off of NOAAport
are fine.


Steve Chiswell
Unidata User Support