Why are unified diffs only "grudgingly" accepted?

François Pinard pinard at iro.umontreal.ca
Sun Aug 6 12:42:18 EDT 2000


[Moshe Zadka -- about asking for `diff -c' instead of `diff -u']

> Yes -- it would be nice if there was interproject consistency -- but
> the most important thing is that there will be intraproject consistency.

You count imposing `-u' or `-c' diffs as "intraproject consistency"?  Wow!
I'd rather look at the patches after being applied, to measure consistency.

Unless you meant that it is an internal part of a project to consistently
impose debatable strict rules to users?

> [FP]
> > As I wrote a few times in the past, it progressively becomes more difficult
> > for users to report bugs than it used to be

> [MZ]
> I doubt it -- I think it has become easier.

OK, we disagree on that one.  Maybe easier for you, then.  It does not
necessarily means easier for everybody.

> [MZ]
> The maintainer time *is* precious [...]

Absolutely no doubt about that.  However, I hardly believe anyone's time is
_so_ precious that compromises are out of question, and nothing exists besides
what they prefer.

> [MZ]
> But a maintainer spends more time over user bug reports then the users,
> usually -- certainly if she's a good maintainer.  So the correct thing to
> do is to worry more about the time you spend in the maintainer hat.

I know a few users that submit very valuable bug reports to many maintainers,
and they spend a great deal of their time at this, being contributing users.
I really praise their time, and thank their contributions.

The correct thing to do is to be good citizens, and that implies give and
take, between people that are respectful of each other, and not between
classes that impose their methods over others.  I may decide to choose
methods to especially please someone, if they are affordable enough to me,
but I just do not like being forced into them.

> [MZ]
> I think that demanding that developers install another tool (which must be
> easy to use on MacOS, Windows and all Unices) is more of a hassle then
> demanding that users learn to pass the correct flag to diff.

We also quite disagree, here.  I'll better do a little or moderate effort on
my side, than ask thousands of people to make an effort all the time to save
me a few minutes once.  Maybe I consider my time precious, but surely not
*that* precious.  Or else, I would be one of those few despisable haughty
maintainers, who make me doubt about free software.  Note that Guido never
created me any real problem so far, do not depict him as worse than he is!

-- 
François Pinard   http://www.iro.umontreal.ca/~pinard




More information about the Python-list mailing list