how do you compile NEMO when you have put new subroutines in?

[bs]

I have added a biogeochemistry model into NEMO (subdirectory MEDUSA in TOP_SRC)
and committed the changes to fcm (revision 2242) as part of the iMARNET project. My question is how will the umui know to compile the new routines? - do we have to change the fcm file for instance? Best wishes, Bablu Sinha

[ros]

Hi Bablu,

Sorry for the delay, I was away last week.

Because the MEDUSA directory is not in the trunk FCM doesn't know to extract it. Now I'm not entirely sure of the correct way of forcing it, but since you are only including one NEMO branch I think the easiest way to get this to work is to modify the nemo_IBM_3.2_base_medusa.cfg FCM configuration file to make your VN3.2_iMARNET_27_06_2012_1540 branch the trunk. This is what I had to do in the past.

So I suggest changing repos::nemo::base and revision::nemo::base to be

repos::nemo::base       fcm:nemo_br/dev/bs/VN3.2_iMARNET_27_06_2012_1540/NEMO

revision::nemo::base    2243

In the meantime, I will check with the NEMO guys that this is still the method required.

Regards,
Ros.

[ros]

Updating ticket with official solution:

Having spoken with the Met Office this is a bit of an FCM problem / UMUI problem.
The new mechanism in FCM which the UM/UMUI hasn't migrated to yet handles new directories much better than the current one. Until then we're stuck with modifying the nemo_IBM_3.2_base_medusa.cfg file.

So we need to have

repos::nemo::base      fcm:nemo_tr/NEMO
revision::nemo::base   vn3.2

repos::nemo::medusa      fcm:nemo_br/dev/bs/VN3.2_iMARNET_27_06_2012_1540/NEMO
revision::nemo::medusa   2243

src::nemo::base            LIM_SRC_2
src::nemo::base            LIM_SRC_3
src::nemo::base            C1D_SRC
expsrc::nemo::base         OPA_SRC
expsrc::nemo::base         TOP_SRC
expsrc::nemo::medusa      OPA_SRC
expsrc::nemo::medusa      TOP_SRC

Basically we need to add an expsrc statement for each branch to match any expsrc statements used with the base. For the UM this is done automatically, so it looks like an oversight on the NEMO/CICE front. A fix for this has been requested and will hopefully be sorted out by the Met Office for UM8.3