John Crispin wrote:
> also i am not sure if we want the script to abort if one of the feeds has
failed
The typical scenario for somone building regularly private builds is that
updating fails due to conflicting local modifications. The user might have a
"daily" build script that updates the sources and then continues with the
build. Having the build script to stop after a feed update failure is
practical. You generally want to fix the conflict instead of just charging
blindly ahead and building from partially stale sources.
For example, my build environment contains this proposed patch and the daily
build script includes check for update failure:
./scripts/feeds update -a
[ "$?" -ne 0 ] && echo "Updating the feeds failed." && exit 1
If you feel that the "break after one feed fails to update" is to strict, one
solution could be to continue updating other feeds despite the failure, but
still return the error code in any case just to indicate that one feed
failed. That would enable the user to select, if he wants to check for that
error code.
Currently the error code is lost. There is no feedback of a possible failure.
_______________________________________________
openwrt-devel mailing list
openwrt-devel@lists.openwrt.org
https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
