Subject: git-man: apply.ignorewhitespace should be apply.ignore-whitespace
Date: Thu, 23 Jun 2011 19:54:30 +0200
Package: git-man
Version: 1:1.7.5.4-1
Severity: normal
As using apply.ignorewhitespace gives an error and git apply --help shows that
a dash is needed, the man page of "git-config" should be fixed to use apply
..ignore-whitespace
-- System Information:
Debian Release: wheezy/sid
APT prefers testing
APT policy: (990, 'testing'), (500, 'unstable'), (500, 'stable')
Architecture: amd64 (x86_64)
Kernel: Linux 2.6.39-2-amd64 (SMP w/4 CPU cores)
Locale: LANG=de_AT.UTF-8, LC_CTYPE=de_AT.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash
-- no debconf information
Acknowledgement sent
to Jonathan Nieder <[email protected]>:
Extra info received and forwarded to list. Copy sent to Gerrit Pape <[email protected]>.
(Thu, 23 Jun 2011 18:06:03 GMT) (full text, mbox, link).
Subject: Re: git-man: apply.ignorewhitespace should be apply.ignore-whitespace
Date: Thu, 23 Jun 2011 13:03:47 -0500
Hi Philipp,
Ph. Marek wrote:
> As using apply.ignorewhitespace gives an error and git apply --help shows that
> a dash is needed, the man page of "git-config" should be fixed to use apply
> ..ignore-whitespace
Could you elaborate (for example, what error message do you get)? I
tried putting
[apply]
ignorewhitespace = change
in ~/.gitconfig, then ran "git format-patch -1" to get a patch to test
with, changed some whitespace, and ran "git apply -R 0001-*" to use
the patch. It seemed to work.
That said, I think that the git-config(1) manual should document that
[apply]
ignore-whitespace = change
is accepted as a backward-compatibility synonym, in the same spirit as
the documentation for add.ignoreerrors. Thanks for reporting it.
Hope that helps,
Jonathan
Acknowledgement sent
to Philipp Marek <[email protected]>:
Extra info received and forwarded to list. Copy sent to Gerrit Pape <[email protected]>.
(Fri, 24 Jun 2011 07:00:03 GMT) (full text, mbox, link).
Subject: Re: git-man: apply.ignorewhitespace should be apply.ignore-whitespace
Date: Fri, 24 Jun 2011 08:47:36 +0200
Hello Jonathan!
> > As using apply.ignorewhitespace gives an error and git apply --help
> > shows that a dash is needed, the man page of "git-config" should be
> > fixed to use apply ..ignore-whitespace
>
> Could you elaborate (for example, what error message do you get)?
Starting "git gui" and choosing a chunk to commit gave an error:
"fatal: unrecognized whitespace ignore option 'yes'".
The keyword with dash was accepted, and "git apply" had it this way, too.
So, which is correct now?
Thank you!
Regards,
Phil
Acknowledgement sent
to Jonathan Nieder <[email protected]>:
Extra info received and forwarded to list. Copy sent to Gerrit Pape <[email protected]>.
(Fri, 24 Jun 2011 14:24:07 GMT) (full text, mbox, link).
Subject: Re: git-man: apply.ignorewhitespace should be apply.ignore-whitespace
Date: Fri, 24 Jun 2011 09:21:57 -0500
tags 631421 + upstream
quit
Hi again,
Philipp Marek wrote:
> Hello Jonathan!
>> Could you elaborate (for example, what error message do you get)?
>
> Starting "git gui" and choosing a chunk to commit gave an error:
> "fatal: unrecognized whitespace ignore option 'yes'".
>
> The keyword with dash was accepted, and "git apply" had it this way, too.
Ah, I see now that I was confused. There *is* no
apply.ignore-whitespace option, and the documentation is correct on
that point. I had confused it for another option when saying the
dashed form is accepted for compatibility; sorry for the sloppiness.
For apply.ignorewhitespace, "yes" is not a recognized value. The
typical way to stop caring about whitespace distinctions is to set the
value to "change". That is what the above message is about. It would
be nice to introduce "all", too, and to make "yes" a synonym for
"change".
Since apply.ignore-whitespace (like apply.jfkdasfd) is not a
recognized option, git does not take it to mean anything and is happy
with it having any value.
> So, which is correct now?
I see a few bugs here:
- the "[apply] IgnoreWhitespace" option does not accept "true" as a
value, while it would be reasonable to expect it to
- the error message for unrecognized values could be clearer
- the documentation should emphasize somewhere that configuration
item names are case-insensitive and the usual convention is for
them never to contain a dash or underscore
> Thank you!
No, thank you. :) If people never reported these things, the program
would never get any better.
Regards,
Jonathan
Debbugs is free software and licensed under the terms of the GNU General
Public License version 2. The current version can be obtained
from https://bugs.debian.org/debbugs-source/.