[MITgcm-devel] useMNC not in PARAMS.h ?

[Patrick Heimbach]

I am guessing Ed's intention was to switch on MNC if
it's set in packages.conf _and_ available on the platform, 
and off otherwise.
For the verification experiments / genmake2 that works well
if useMNC doesn't need to be set/unset in _data.pkg_ ,
but it will be less clear otherwise 
(i.e. if #undef ALLOW_MNC and useMNC=T)

I guess it's presently benign since all MNC stuff is 
bracketed by ALLOW_MNC in addition to useMNC
so even if useMNC was set to TRUE in data.pkg it
would just be ignored (assuming we break the rule,
which we did, and don't have a check for consistency 
in packages_check ;o)

Bottom line: 
including useMNC in data.pkg will preserve
the "automatic" sitch/unswitch of ALLOW_MNC by genmake2.

-Patrick



  > -----Original Message-----
  > From: mitgcm-devel-bounces at mitgcm.org 
  > [mailto:mitgcm-devel-bounces at mitgcm.org] On Behalf Of Chris Hill
  > Sent: Saturday, October 09, 2004 7:23 PM
  > To: MITgcm-devel at mitgcm.org
  > Subject: RE: [MITgcm-devel] useMNC not in PARAMS.h ?
  > 
  > 
  > Hi Jean-Michel,
  > 
  >  You are right useXXX flags should _not_ be in a package 
  > specific header
  > file. It looks like it should work if it is changed and 
  > that the current
  > setup is just an oversight - Ed?
  >  
  >  In the doc session this week we should include reviewing 
  > and updating
  > "package" guidelines to make sure the instructions and 
  > rules are clear -
  > some of the stuff is still buried in the initial text notes 
  > on packages I
  > wrote years ago and probably some stuff may either be 
  > confusing or missing.
  > 
  > Chris
  > > -----Original Message-----
  > > From: mitgcm-devel-bounces at mitgcm.org 
  > > [mailto:mitgcm-devel-bounces at mitgcm.org] On Behalf Of 
  > > Jean-Michel Campin
  > > Sent: Saturday, October 09, 2004 7:11 PM
  > > To: MITgcm-devel at mitgcm.org
  > > Subject: [MITgcm-devel] useMNC not in PARAMS.h ?
  > > 
  > > Hi,
  > > 
  > > I don't understand why useMNC is not like all the other pkg 
  > > On/Off flag, sitting in PARAMS.h
  > > 
  > > I found this annoying, for 2 reasons:
  > > 
  > > 1) I want to compile mnc pkg, and decide not to use it 
  > > (useMNC=F), but get stuck in land_readparams because of 
  > > land_mnc_init ...
  > > > PID.TID 0000.0001) *** ERROR *** 
  > MNC_GET_NEXT_EMPTY_IND: array size
  > > >    1000 exceeded--try increasing MNC_MAX_ID
  > > I cannot simply put a IF (useMNC) in front of the call, in 
  > > land_readparams.F, because useMNC is not where it should be 
  > > (in PARAMS.h), but in MNC_PARAMS.h (in pkg/mnc) and I don't 
  > > want to include this file in land_readparams.F with those 
  > ugly #ifdef.
  > > 
  > > 2) I spend sometime to make the different packages working in 
  > > a a similar manner (I remember this packages_write_pickup 
  > > stuff), and now have the impression that this is completely 
  > > useless if the goal is to make things the less standard 
  > as possible. 
  > > 
  > > Jean-Michel
  > > _______________________________________________
  > > MITgcm-devel mailing list
  > > MITgcm-devel at mitgcm.org
  > > http://dev.mitgcm.org/mailman/listinfo/mitgcm-devel
  > > 
  > 
  > _______________________________________________
  > MITgcm-devel mailing list
  > MITgcm-devel at mitgcm.org
  > http://dev.mitgcm.org/mailman/listinfo/mitgcm-devel
  >