Re: [PATCH 06/11] commit: add tests of commit races

[Michael Haggerty <mhagger@alum.mit.edu>]

On 02/10/2015 08:12 PM, Junio C Hamano wrote:
> Stefan Beller <sbeller@google.com> writes:
> 
>> On Sun, Feb 8, 2015 at 8:14 AM, Michael Haggerty <mhagger@alum.mit.edu> wrote:
>>
>>> +# Copyright (c) 2014 Michael Haggerty <mhagger@alum.mit.edu>
>>
>> What is the projects stance on copyright lines?
> 
> I do not think we have a strong one.
> 
>> I've seen files (most of them from the beginning) having some copyright lines,
>> other files (often introduced way later) not having them, "because
>> we're git and have
>> history, so we know who did it".
> 
> I personally agree with that statement.  Also, a copyright notice
> per file is often added when a new file is added, but that ends up
> giving false sense of "ownership" to everybody else down the line
> even after the file has been extensively modified.  It's not like
> Michael solely owns all lines in this file in later versions.  And
> even if people added their name at the top every time they make any
> change, their names tend to stay even when their contributions are
> later completely rewritten or removed.
> 
> In a sense, my agreement with your statement is stronger than "Yes,
> Git can tell us who did what anyway".  What we can find in the
> history is the sole source of truth, and in-file copyright notice is
> misleading.  You do not even have to have one in the Berne signatory
> nations anyway.

I only put a copyright notice there because I thought it was standard
practice. I think it is ugly and would rather do without it, even aside
from the practical problems that Junio mentioned.

On the other hand, there's this [1] and this [2] from the FSF, which
recommend a copyright blurb at the beginning of every source file.
Though actually the recommendation is to include a GPL blurb too, not
just a naked copyright line like I used. But I get the feeling that the
FSF's recommendation is more for ideological than for legal reasons.

If I don't hear anything else, I'll delete the copyright line in the reroll.

>> The tests themselves look fine.
>>
>> Is there a reason you did not append the tests in 7509 ?
> 
> Hmph.

I don't know what "Hmph" means in this context.

The description for t7509 is "git commit --reset-author", which doesn't
seem to describe the new tests.

There are also

    t7500 "git commit / Tests for selected commit options"
    t7501 "git commit"
    t7502 "git commit porcelain-ish"

I suppose the new tests could go in any of these. But since the tests
are thematically a bit unusual (dealing with races rather than testing
command-line options) and they start with an orphan commit, I thought it
would be just as easy to put them in their own file to make it clear
that they are independent.

I really don't care either way.

Michael

[1] http://www.gnu.org/licenses/gpl-howto.html
[2] http://www.gnu.org/licenses/gpl-faq.html#NoticeInSourceFile

-- 
Michael Haggerty
mhagger@alum.mit.edu