Skip to contentSkip to conversation

get_maintainers.pl is rude, was Re: [PATCH 05/19] USB: inode.c: move assignment out of if () block

DISCUSSION Christoph Hellwig 3 participants 5 messages Apr 30, 2015 – May 5, 2015 4 messages not in the archive GET_MAINTAINER script

lore
Share & tools
lore

A message that is not in the archive (sent off-list or not yet received): [email protected]

A message that is not in the archive (sent off-list or not yet received): [email protected]

A message that is not in the archive (sent off-list or not yet received): [email protected]

A message that is not in the archive (sent off-list or not yet received): [email protected]

#4 lore raw
Quoted 2 lines · Greg Kroah-Hartman

scripts/get_maintainer.pl is bringing up your name for this file as you have modified it in the past:

I've probably modified a large part of files in the kernel in the past.

And I'm annoyed about this stupid behavior in get_maintainers.pl.

Joe, can you please fix your bloody script to not report me for every goddamnt file I touched once? Authorship of changes to a file is a completely wrong metric. I can see the point to guess it by non-maintainer signoffs, but authorship is plain wrong and highly annoying.

If you're unwilling to fix this please at least add a get_maintainers.ignore file and add me as the first entry. Thanks!

#5 lore raw
Quoted 5 lines · Christoph Hellwig
Quoted 2 lines · Greg Kroah-Hartman

scripts/get_maintainer.pl is bringing up your name for this file as you have modified it in the past:

I've probably modified a large part of files in the kernel in the past.

[...]
Does it make sense(Is it enough to filter the changes to only include changes from the last 6 months or so (if there are newer changes;-)?

People probably don't care if they didn't touch it for ages anyway.

Or ignore changes which touch only a few lines completely (to not include typo fixes or trivial fixups)?

Kind regards,
Bernd

Signature · 4 lines
"I dislike type abstraction if it has no real reason. And saving
on typing is not a good reason - if your typing speed is the main
issue when you're coding, you're doing something seriously wrong."
    - Linus Torvalds
maintainer
#6 lore raw
In reply to Christoph Hellwig
Quoted 2 lines · Christoph Hellwig

Joe, can you please fix your bloody script to not report me for every goddamnt file I touched once?

For files that have no specific maintainer, the default period for reporting commit sign-offs and authors as possible interested parties in a patch is 1 year.

And given how little my sweet wife likes me when I swear around her, I'm not inclined to do that specifically for you, no.

Quoted 2 lines

Authorship of changes to a file is
a completely wrong metric..

That was Andrew Morton's idea.
https://lkml.org/lkml/2013/10/17/460

Quoted 3 lines

I can see the point to guess it by
non-maintainer signoffs, but authorship is plain wrong and highly annoying.

That's an assertion, but you've neglected to give a rationale for it. I think authorship is quite a good reason to be cc'd on something as given that you've spent the effort to originate code, you're also quite likely to be interested in patches for that code.

Quoted 2 lines

If you're unwilling to fix this please at least add a get_maintainers.ignore file and add me as the first entry. Thanks!

If a get_maintainers.ignore file is created, (which seems like a reasonable idea, thanks), I'm not maintaining it.

Maybe this: (if you create a .get_maintainer.ignore file with your name in it like)

$ cat .get_maintainer.ignore
Christoph Hellwig <[email protected]>
$
---
 scripts/get_maintainer.pl | 32 ++++++++++++++++++++++++++++++++
 1 file changed, 32 insertions(+)
scripts/get_maintainer.pl +32 -0
index d701627..fc169fd 100755
@@ -186,6 +186,27 @@ if (-f $conf) {
     unshift(@ARGV, @conf_args) if @conf_args;
 }
 
+my @ignore_emails = ();
+my $ignore_file = which_conf(".get_maintainer.ignore");
+if (-f $ignore_file) {
+    open(my $ignore, '<', "$ignore_file")
+	or warn "$P: Can't find a readable .get_maintainer.ignore file $!\n";
+    while (<$ignore>) {
+	my $line = $_;
+
+	$line =~ s/\s*\n?$//;
+	$line =~ s/^\s*//;
+	$line =~ s/\s+$//;
+	$line =~ s/#.*$//;
+
+	next if ($line =~ m/^\s*$/);
+	if (rfc822_valid($line)) {
+	    push(@ignore_emails, $line);
+	}
+    }
+    close($ignore);
+}
+
 if (!GetOptions(
 		'email!' => \$email,
 		'git!' => \$email_git,
@@ -513,6 +534,16 @@ if ($web) {
 
 exit($exit);
 
+sub ignore_email_address {
+    my ($address) = @_;
+
+    foreach my $ignore (@ignore_emails) {
+	return 1 if ($ignore eq $address);
+    }
+
+    return 0;
+}
+
 sub range_is_maintained {
     my ($start, $end) = @_;
 
@@ -1868,6 +1899,7 @@ sub vcs_assign {
 	my $percent = $sign_offs * 100 / $divisor;
 
 	$percent = 100 if ($percent > 100);
+	next if (ignore_email_address($line));
 	$count++;
 	last if ($sign_offs < $email_git_min_signatures ||
 		 $count > $email_git_max_maintainers ||
#7 lore raw
Quoted 10 lines · Joe Perches
Quoted 3 lines

I can see the point to guess it by
non-maintainer signoffs, but authorship is plain wrong and highly annoying.

That's an assertion, but you've neglected to give a rationale for it. I think authorship is quite a good reason to be cc'd on something as given that you've spent the effort to originate code, you're also quite likely to be interested in patches for that code.

It's not. Patch authorship != driver authorship. There are people like me or Al that keep fixing interface all over the tree. That doesn't really mean I need patches touching those same files again in my inbox. I generally couldn't care less, and if I do I will pick it up through the mailinglists.

Quoted 9 lines
Quoted 2 lines

If you're unwilling to fix this please at least add a get_maintainers.ignore file and add me as the first entry. Thanks!

If a get_maintainers.ignore file is created, (which seems like a reasonable idea, thanks), I'm not maintaining it.

Maybe this: (if you create a .get_maintainer.ignore file with your name in it like)

If we can't get the bullshit heuristics fixed properly please add this support, and feel free to use me as the initial seed for it.

maintainer
#8 lore raw
Quoted 4 lines · Christoph Hellwig
Quoted 3 lines · Joe Perches
Quoted 3 lines

I can see the point to guess it by
non-maintainer signoffs, but authorship is plain wrong and highly annoying.

The commit signer/author lookup is only done for files that do not have a specified maintainer.

Quoted 7 lines
Quoted 6 lines

That's an assertion, but you've neglected to give a rationale for it. I think authorship is quite a good reason to be cc'd on something as given that you've spent the effort to originate code, you're also quite likely to be interested in patches for that code.

Quoted 5 lines

It's not. Patch authorship != driver authorship. There are people like me or Al that keep fixing interface all over the tree. That doesn't really mean I need patches touching those same files again in my inbox. I generally couldn't care less, and if I do I will pick it up through the mailinglists.

There are also people that author patches that are interested in follow-on changes too.

I also touch many files across the kernel, and I get the odd random patches I'm not uninterested in. I have no issue with the 3 seconds it takes to ignore them.

Quoted 2 lines

If we can't get the bullshit heuristics fixed properly please add this support, and feel free to use me as the initial seed for it.

As I said, I'm not interested in doing anything to any .ignore file.

I'll submit the patch to the script, you can submit the patch to add your name to the .ignore file.

cheers, Joe