Opened 14 years ago
Closed 10 years ago
#1564 closed defect (worksforme)
Do some sanity checks on the omc executable before copy to build/bin (or rename the old one)
Reported by: | Adrian Pop | Owned by: | Adrian Pop |
---|---|---|---|
Priority: | high | Milestone: | Bootstrapping |
Component: | Build Environment | Version: | |
Keywords: | Cc: | Adrian Pop |
Description
If omc for some reason (you made some mistakes) exits with error the GenerateOMCHeader.mos script will fail and you cannot compile omc anymore.
So, you're stuck with an omc that is not working and you need a working omc in order to fix it.
We should maybe do a couple of basic tests on omc (or actually run a test with GenerateOMCHeader.mos) before we overwrite build/bin/omc.
I got bitten by this several times already :)
Lucky for me I had a working omc available someplace else.
Change History (7)
comment:1 by , 14 years ago
comment:3 by , 14 years ago
If my omc is broken, I just remove it and recompile... omc-diff is already built by the build command ;)
comment:4 by , 14 years ago
Damn, haven't thought of that :)
You're right but people might get confused like me, so I still think is better to check a bit.
comment:5 by , 14 years ago
omc-diff is not build by the build command, at least not in Windows:
If i do a "make clean all", then after the omc is build and I try to
run a test from the command line, I have no omc-diff. I'll check the
MinGW makefiles.
comment:6 by , 11 years ago
Cc: | adrpo, → adrpo |
---|---|
Milestone: | → 2.0.0 |
On Linux/OSX you can now configure --with-omc=/usr/bin/omc for example. This will use another omc so the build environment will never be messed up. It would be good if Windows could use this approach as well.
comment:7 by , 10 years ago
Milestone: | 2.0.0 → Bootstrapping |
---|---|
Resolution: | → worksforme |
Status: | new → closed |
Maybe we should have a test with the prerequisites for the omc compilation:
Anything more we need?