DO NOT MERGE make mac make nicer

[edsantiago]

...and on all other platforms, too.

BINSFX has grown beyond its original intention, and is now
misleading and unmaintainable. Replace it with more natural
target names DEFAULTBIN and REMOTEBIN.

Signed-off-by: Ed Santiago [email protected]

None

[edsantiago]

The other part of this morning's conversation was something about make validate on Mac. This has been a wild-goose chase.

• It's possible to run GOOS=darwin make validate on Linux. Also =freebsd. This produces 500+ and 64,000+ lines of output respectively. I can't tell if this is a problem with missing libraries/packages on Linux, i.e., I have no way of knowing if this is what happens on the respective platforms.
• I surmise that it's possible to run GOOS=linux make validate on Mac but
  • I have no idea whether it will work or not, and
  • Is that really desirable in the first place?

If make validate, when run on a Mac, is as broken as GOOS=darwin make validate is on Linux, and if GOOS=linux make validate yields useful results on a Mac (someone please test this), then maybe it's a good idea to hardcode GOOS=linux into the Makefile for make validate. This is something I can't test.

[openshift-ci]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: baude, edsantiago

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [baude,edsantiago]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[baude]

will test now ...

[baude]

for failures, there seems to be a precedent instead of echo in the Makefile for:

$(warning test goes here ...

should we follow suit or make the previous usage echos?

[baude]

the update here is that make podman is not being "trapped" with this patch ... it is not obvious to me why not, but i can try to keep poking at this

[edsantiago]

Re: $(warning -- yes, totally, for make docs.

For bin/podman, though, I'm leaning more and more toward just silently redirecting it to $(MAKE) bin/podman-remote or something that quietly works. No warnings, no echo, just make it happen. It just seems kinder in every sense.

[edsantiago]

the update here is that make podman is not being "trapped" with this patch ...

Can you paste or link to the full output?

[edsantiago]

OH! I think I got it. Gimme a while to confirm and look for a fix...

[edsantiago]

Try this. (Relevant change involves recognizing $(SRCBINDIR). I think this will work on Linux as well as Mac, but this is a trickier change that I need to test a lot more. ITM, please see if it helps on Mac.

[edsantiago]

@baude here's a much longer, much trickier delta that I think may be overall more maintainable in the future. Could you test it at your convenience please? I'm a little concerned that we might see make: duplicate target warnings and if we do I have no idea how to fix them. So let's hope we don't.

[edsantiago]

sys failure is a flake, #10927, one of our oldest most beloved ones. No point in rerunning it though, my goal here was to confirm that make still works with these changes.

[edsantiago]

This never got any traction. I don't know if the original problem is still a problem, and without a Mac I don't even have any way to test this. I'll just assume the problem has been fixed.