tech-pkg archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: RHEL/Centos 7 builtin vs pkgsrc zlib clash for Makefiles that include devel/zlib/

On 2020-08-19 08:10, Greg Troxel wrote:
Jason Bacon <> writes:

On 2020-08-18 19:35, Brook Milligan wrote:
Are all the yes/no options lower case as in your patch?  Shouldn't they be case insensitive?  Or are you matching prior art?
Good question.  Manually set mk variables can certainly take
case-insensitive values, but I think the command-line flags should
agree with the usage message: "--prefer-pkgsrc <list|yes|no>".

If there's a consensus that the boostrap script should accept things
like "--prefer-pkgsrc YES" or "--prefer-pkgsrc Yes", I would be happy
to change it and update the usage message to match.

I also occurs to me now that we can never have a package named "yes"
or "no" without breaking these flags.
Given that the flags that are already there behave a certain way, it
seems clear this patch should not change that, and should have the same
yes/YES behavior as before.

(If people want to talk about changing yes/YES behavior in general, I
don't mean to say "that's prohibited".   Just "please keep separable
changes separable".  (But I also see no reason to change.))

Looking more closely, there are other variables that are compared only to lower case "yes" or "no" in places, e.g.

if [ "x$preserve_path" != "xyes" ]; then

case "$cwrappers" in
        echo "USE_CWRAPPERS=            $cwrappers" >> ${TARGET_MKCONF}

So supporting case-insensitive command-line arguments consistently would require some fairly extensive changes that are probably not worth the risk or clutter.  I don't recall any past complaints about LC-only bootstrap args, so I'm inclined to leave it alone.

Home | Main Index | Thread Index | Old Index