[M3devel] Darwin upgrade/fingerprint error

[Jay]

Starting with an older release on PPC_DARWIN and running upgrade.sh, I keep getting:

=== package /dev2/cm3/m3-sys/m3middle ===
 +++ cm3 -build -DROOT='/dev2/cm3' -DCM3_VERSION_TEXT='d5.5.1' -DCM3_VERSION_NUMBER='050501' -DCM3_LAST_CHANGED='2007-12-30'  && cm3 -ship -DROOT='/dev2/cm3' -DCM3_VERSION_TEXT='d5.5.1' -DCM3_VERSION_NUMBER='050501' -DCM3_LAST_CHANGED='2007-12-30'  +++
--- building in PPC_DARWIN ---

ignoring ../src/m3overrides

stale imports -> compiling CoffTime.m3
stale imports -> compiling M3Process.i3
stale imports -> compiling M3Process.m3
stale imports -> compiling TFloat.m3

Fatal Error: bad version stamps: TFloat.m3

version stamp mismatch: TargetMap.Float_types
  <31f4d45a27120488> => TFloat.m3
  <663449bb704e9c85> => TargetMap.i3  
 *** execution of  failed ***


I assume this will be fixed by:

ijayk:/dev2/cm3/scripts jay$ cvs diff upgrade.sh
Index: upgrade.sh
===================================================================
RCS file: /usr/cvs/cm3/scripts/upgrade.sh,v
retrieving revision 1.4
diff -r1.4 upgrade.sh
52a53,55
> echo "$ROOT/scripts/do-pkg.sh" "$@" "realclean ${P}"
> "$ROOT/scripts/do-pkg.sh" "$@" "realclean" ${P} || exit 1
> 
58a62,64
> echo "$ROOT/scripts/do-cm3-core.sh" "$@" "realclean"
> "$ROOT/scripts/do-cm3-core.sh" "$@" "realclean" || exit 1
> 

Will that really harm/slow anyone's workflow?

In fact, notice how the older upgrade, while broken for other reasons, did this
dance with upgrading and restoring the pkg store.
win\upgrade.cmd still does.
Something to consider..
(And I realize that backup/restore of install/pkg is different than realclean of root/*, so maybe
that diff is not the fix.)

Though it is a problem to manage the versions, in doing the backup/restore.

Perhaps this all falls under the heading of -- not worth automating, people know what they are doing, else they can pickup a current binary distribution??

I'll get something working locally no doubt, I don't need any further help. :)
(Though past help has helped shed light here, the local/global/ship thing is still just a tad vague to me and I'm just slightly unconvinced it has to be this way, but probably it does, what with necessary circularities between compiler and runtime, building with an older compiler and all that...)

 - Jay

_________________________________________________________________
Share life as it happens with the new Windows Live.
http://www.windowslive.com/share.html?ocid=TXT_TAGHM_Wave2_sharelife_012008
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://m3lists.elegosoft.com/pipermail/m3devel/attachments/20080104/6462cc68/attachment-0002.html>