Why does make run git?

classic Classic list List threaded Threaded
29 messages Options
12
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Why does make run git?

Richard Stallman
[[[ To any NSA and FBI agents reading my email: please consider    ]]]
[[[ whether defending the US Constitution against all enemies,     ]]]
[[[ foreign or domestic, requires you to follow Snowden's example. ]]]

I was surprised to see this output from `make'.  What does this do?
It seems fundamentally unmodular for building Emacs from sources
to concern itself in any way with how you got those sources.

Configuring local git repository...
‘.git/config’ -> ‘.git/config.~1~’
git config transfer.fsckObjects 'true'
git config diff.elisp.xfuncname '^\(def[^[:space:]]+[[:space:]]+([^()[:space:]]+)'
git config diff.m4.xfuncname '^((m4_)?define|A._DEFUN(_ONCE)?)\([^),]*'
git config diff.make.xfuncname '^([$.[:alnum:]_].*:|[[:alnum:]_]+[[:space:]]*([*:+]?[:?]?|!?)=|define .*)'
git config diff.shell.xfuncname '^([[:space:]]*[[:alpha:]_][[:alnum:]_]*[[:space:]]*\(\)|[[:alpha:]_][[:alnum:]_]*=)'
git config diff.texinfo.xfuncname '^@node[[:space:]]+([^,[:space:]][^,]+)'
Installing git hooks...
‘build-aux/git-hooks/commit-msg’ -> ‘.git/hooks/commit-msg’
‘build-aux/git-hooks/pre-commit’ -> ‘.git/hooks/pre-commit’
‘.git/hooks/applypatch-msg.sample’ -> ‘.git/hooks/applypatch-msg’
‘.git/hooks/pre-applypatch.sample’ -> ‘.git/hooks/pre-applypatch’

--
Dr Richard Stallman
President, Free Software Foundation (gnu.org, fsf.org)
Internet Hall-of-Famer (internethalloffame.org)
Skype: No way! See stallman.org/skype.html.


Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Why does make run git?

Eli Zaretskii
> From: Richard Stallman <[hidden email]>
> Date: Mon, 31 Jul 2017 21:19:51 -0400
>
> I was surprised to see this output from `make'.  What does this do?

Configure Git commands for Emacs, mostly.

> It seems fundamentally unmodular for building Emacs from sources
> to concern itself in any way with how you got those sources.

It is indeed somewhat controversial.

> git config transfer.fsckObjects 'true'

This makes Git abort if it sees a bad object being pulled or pushed.

> git config diff.elisp.xfuncname '^\(def[^[:space:]]+[[:space:]]+([^()[:space:]]+)'
> git config diff.m4.xfuncname '^((m4_)?define|A._DEFUN(_ONCE)?)\([^),]*'
> git config diff.make.xfuncname '^([$.[:alnum:]_].*:|[[:alnum:]_]+[[:space:]]*([*:+]?[:?]?|!?)=|define .*)'
> git config diff.shell.xfuncname '^([[:space:]]*[[:alpha:]_][[:alnum:]_]*[[:space:]]*\(\)|[[:alpha:]_][[:alnum:]_]*=)'
> git config diff.texinfo.xfuncname '^@node[[:space:]]+([^,[:space:]][^,]+)'

These tailor the Git's notion of a "function" for the various
languages.  It's handy in several commands that need to know where a
function begins and ends.

> Installing git hooks...
> ‘build-aux/git-hooks/commit-msg’ -> ‘.git/hooks/commit-msg’
> ‘build-aux/git-hooks/pre-commit’ -> ‘.git/hooks/pre-commit’
> ‘.git/hooks/applypatch-msg.sample’ -> ‘.git/hooks/applypatch-msg’
> ‘.git/hooks/pre-applypatch.sample’ -> ‘.git/hooks/pre-applypatch’

These hooks check the format of the commits and verify it is
according to our standards.

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Why does make run git?

Tino Calancha-2
In reply to this post by Richard Stallman


On Mon, 31 Jul 2017, Richard Stallman wrote:

> [[[ To any NSA and FBI agents reading my email: please consider    ]]]
> [[[ whether defending the US Constitution against all enemies,     ]]]
> [[[ foreign or domestic, requires you to follow Snowden's example. ]]]
>
> I was surprised to see this output from `make'.  What does this do?
> It seems fundamentally unmodular for building Emacs from sources
> to concern itself in any way with how you got those sources.
>
> Configuring local git repository...
> ‘.git/config’ -> ‘.git/config.~1~’
> git config transfer.fsckObjects 'true'
> git config diff.elisp.xfuncname '^\(def[^[:space:]]+[[:space:]]+([^()[:space:]]+)'
> git config diff.m4.xfuncname '^((m4_)?define|A._DEFUN(_ONCE)?)\([^),]*'
> git config diff.make.xfuncname '^([$.[:alnum:]_].*:|[[:alnum:]_]+[[:space:]]*([*:+]?[:?]?|!?)=|define .*)'
> git config diff.shell.xfuncname '^([[:space:]]*[[:alpha:]_][[:alnum:]_]*[[:space:]]*\(\)|[[:alpha:]_][[:alnum:]_]*=)'
> git config diff.texinfo.xfuncname '^@node[[:space:]]+([^,[:space:]][^,]+)'
> Installing git hooks...
> ‘build-aux/git-hooks/commit-msg’ -> ‘.git/hooks/commit-msg’
> ‘build-aux/git-hooks/pre-commit’ -> ‘.git/hooks/pre-commit’
> ‘.git/hooks/applypatch-msg.sample’ -> ‘.git/hooks/applypatch-msg’
> ‘.git/hooks/pre-applypatch.sample’ -> ‘.git/hooks/pre-applypatch’
This stuff is run just once, the first time you build.  In some sense
you are doing more than just building Emacs from its sources: your
are cloning the Emacs repository; some configuration of this repository
might be expected.

Those hooks ensure that some minimum guidelines are fulfilled: for
instance, things like limit the number of characters for a line
within a commit message.
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Why does make run git?

Richard Stallman
In reply to this post by Eli Zaretskii
[[[ To any NSA and FBI agents reading my email: please consider    ]]]
[[[ whether defending the US Constitution against all enemies,     ]]]
[[[ foreign or domestic, requires you to follow Snowden's example. ]]]

  > > git config transfer.fsckObjects 'true'

  > This makes Git abort if it sees a bad object being pulled or pushed.

Isn't that a personal preference?  It seems wrong for Emacs to set it
arbitrarily.

  > > git config diff.elisp.xfuncname '^\(def[^[:space:]]+[[:space:]]+([^()[:space:]]+)'
  > > git config diff.m4.xfuncname '^((m4_)?define|A._DEFUN(_ONCE)?)\([^),]*'
  > > git config diff.make.xfuncname '^([$.[:alnum:]_].*:|[[:alnum:]_]+[[:space:]]*([*:+]?[:?]?|!?)=|define .*)'
  > > git config diff.shell.xfuncname '^([[:space:]]*[[:alpha:]_][[:alnum:]_]*[[:space:]]*\(\)|[[:alpha:]_][[:alnum:]_]*=)'
  > > git config diff.texinfo.xfuncname '^@node[[:space:]]+([^,[:space:]][^,]+)'

  > These tailor the Git's notion of a "function" for the various
  > languages.  It's handy in several commands that need to know where a
  > function begins and ends.

Do you mean, several Git commands, or several Emacs commands?
I would guess it is the former.

Running make on the Emacs checkout seems like an inadequate place to
do this.  There is no assurance that you will run make in that
checkout before committing anything.

Perhaps there should be a function in Emacs that does this, and some
VC commands could call it.

  > > Installing git hooks...
  > > ‘build-aux/git-hooks/commit-msg’ -> ‘.git/hooks/commit-msg’
  > > ‘build-aux/git-hooks/pre-commit’ -> ‘.git/hooks/pre-commit’
  > > ‘.git/hooks/applypatch-msg.sample’ -> ‘.git/hooks/applypatch-msg’
  > > ‘.git/hooks/pre-applypatch.sample’ -> ‘.git/hooks/pre-applypatch’

  > These hooks check the format of the commits and verify it is
  > according to our standards.

Running make on Emacs seems like an unreliable place to do this,
because there is no assurance that you will run make in that checkout
before committing anything.


--
Dr Richard Stallman
President, Free Software Foundation (gnu.org, fsf.org)
Internet Hall-of-Famer (internethalloffame.org)
Skype: No way! See stallman.org/skype.html.


Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Why does make run git?

Tino Calancha-2


On Tue, 1 Aug 2017, Richard Stallman wrote:

> [[[ To any NSA and FBI agents reading my email: please consider    ]]]
> [[[ whether defending the US Constitution against all enemies,     ]]]
> [[[ foreign or domestic, requires you to follow Snowden's example. ]]]
>
>  > > git config transfer.fsckObjects 'true'
>
>  > This makes Git abort if it sees a bad object being pulled or pushed.
>
> Isn't that a personal preference?  It seems wrong for Emacs to set it
> arbitrarily.
It's recommended for security.  It was discussed before being
added:
https://lists.gnu.org/archive/html/emacs-devel/2016-01/msg01802.html

>  > > git config diff.elisp.xfuncname '^\(def[^[:space:]]+[[:space:]]+([^()[:space:]]+)'
>  > > git config diff.m4.xfuncname '^((m4_)?define|A._DEFUN(_ONCE)?)\([^),]*'
>  > > git config diff.make.xfuncname '^([$.[:alnum:]_].*:|[[:alnum:]_]+[[:space:]]*([*:+]?[:?]?|!?)=|define .*)'
>  > > git config diff.shell.xfuncname '^([[:space:]]*[[:alpha:]_][[:alnum:]_]*[[:space:]]*\(\)|[[:alpha:]_][[:alnum:]_]*=)'
>  > > git config diff.texinfo.xfuncname '^@node[[:space:]]+([^,[:space:]][^,]+)'
>
>  > These tailor the Git's notion of a "function" for the various
>  > languages.  It's handy in several commands that need to know where a
>  > function begins and ends.
>
> Do you mean, several Git commands, or several Emacs commands?
> I would guess it is the former.
>
> Running make on the Emacs checkout seems like an inadequate place to
> do this.  There is no assurance that you will run make in that
> checkout before committing anything.
>
> Perhaps there should be a function in Emacs that does this, and some
> VC commands could call it.
I think is for convenience.  It is set there once and for all.  There
are plenty of Git interfaces for Emacs out of there: whatever an user
decide to use, s?he already has set up those hooks; then her/him
potential contributions will be easier accepted.

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Why does make run git?

Paul Eggert
In reply to this post by Richard Stallman
Richard Stallman wrote:
> I was surprised to see this output from `make'.  What does this do?
> It seems fundamentally unmodular for building Emacs from sources
> to concern itself in any way with how you got those sources.

Make configures Git only if you got the sources via Git and have not configured
Emacs. If you got the sources via some other means, or have already configured
Emacs, Make doesn't configure Git. So I don't see a violation of modularity
here; it's merely a convenience that Make supplies a default configuration if
you haven't already configured things.

If you prefer to not configure Git even though you got the sources via Git, you
can run 'make ALL_IF_GIT=' or './configure; make' after grabbing the sources.
However, the default Git configuration is typically more convenient.

> Running make on the Emacs checkout seems like an inadequate place to
> do this.  There is no assurance that you will run make in that
> checkout before committing anything.

That's true. However, these are only guidelines, and it's not a huge tragedy if
the guidelines are occasionally violated.

> Perhaps there should be a function in Emacs that does this, and some
> VC commands could call it.

That would be nice. However, some developers don't use Emacs VC to check in (I
typically use vc-dwim instead), and so it's good to have Git check things as well.

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Why does make run git?

Kaushal Modi
On Tue, Aug 1, 2017 at 11:11 AM Paul Eggert <[hidden email]> wrote:
That's true. However, these are only guidelines, and it's not a huge tragedy if
the guidelines are occasionally violated.

> Perhaps there should be a function in Emacs that does this, and some
> VC commands could call it.

That would be nice. However, some developers don't use Emacs VC to check in (I
typically use vc-dwim instead), and so it's good to have Git check things as well.

+1. I don't use Emacs VC and I actually like the checks make is doing at the moment.
--

Kaushal Modi

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Why does make run git?

Radon Rosborough
In reply to this post by Richard Stallman
> It seems fundamentally unmodular for building Emacs from sources
> to concern itself in any way with how you got those sources.

But there are also .dir-locals.el files in the source tree. It seems
equally unmodular for building Emacs from sources to concern itself in
any way with how you edit those sources. It just so happens that
configuring Git can't be done by including files in the source tree
(for security reasons); it has to be done by running Git commands.

If your reaction to the .dir-locals.el argument is "but obviously it
makes sense for Emacs' source tree to have particular support for
Emacs, since that is the software it is building", then I would say it
makes equal sense for Emacs' source tree to have particular support
for Git, since that is the version-control system it is using. Much
like people include Travis configuration files, Docker configuration
files, and—well—.gitignore and .gitattributes files in their source
tree.

IMO, it's really not worth it to aspire to complete modularity if that
comes at the cost of significant practical inconvenience.

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Why does make run git?

Richard Stallman
In reply to this post by Tino Calancha-2
[[[ To any NSA and FBI agents reading my email: please consider    ]]]
[[[ whether defending the US Constitution against all enemies,     ]]]
[[[ foreign or domestic, requires you to follow Snowden's example. ]]]

  > This stuff is run just once, the first time you build.  In some sense
  > you are doing more than just building Emacs from its sources: your
  > are cloning the Emacs repository; some configuration of this repository
  > might be expected.

I'm not sure concretely what "might be expected" means here.
How do other programs configure git?

The question is not whether these configurations are good ones,
but rather how properly to install them.

--
Dr Richard Stallman
President, Free Software Foundation (gnu.org, fsf.org)
Internet Hall-of-Famer (internethalloffame.org)
Skype: No way! See stallman.org/skype.html.


Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Why does make run git?

Richard Stallman
In reply to this post by Tino Calancha-2
[[[ To any NSA and FBI agents reading my email: please consider    ]]]
[[[ whether defending the US Constitution against all enemies,     ]]]
[[[ foreign or domestic, requires you to follow Snowden's example. ]]]

  > > Perhaps there should be a function in Emacs that does this, and some
  > > VC commands could call it.
  > I think is for convenience.

But it is unreliable.  You can't rely on the user's having done 'make'
in the repository before doing the git operations that use these
settings.


--
Dr Richard Stallman
President, Free Software Foundation (gnu.org, fsf.org)
Internet Hall-of-Famer (internethalloffame.org)
Skype: No way! See stallman.org/skype.html.


Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Why does make run git?

Richard Stallman
In reply to this post by Radon Rosborough
[[[ To any NSA and FBI agents reading my email: please consider    ]]]
[[[ whether defending the US Constitution against all enemies,     ]]]
[[[ foreign or domestic, requires you to follow Snowden's example. ]]]

  > But there are also .dir-locals.el files in the source tree. It seems
  > equally unmodular for building Emacs from sources to concern itself in
  > any way with how you edit those sources.

.dir-locals.el isn't related specifically to "building Emacs from
sources".  It applies to all editing of the checked-out Emacs sources
with Emacs.

Whether it is good or bad to have these .dir-locals.el files is
another issue and I see no need to discuss it now.  My point is only
that it is a not similar to the issue at hand, and not directly
pertinent.

--
Dr Richard Stallman
President, Free Software Foundation (gnu.org, fsf.org)
Internet Hall-of-Famer (internethalloffame.org)
Skype: No way! See stallman.org/skype.html.


Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Why does make run git?

Radon Rosborough
In reply to this post by Richard Stallman
> You can't rely on the user's having done 'make'
> in the repository before doing the git operations that use these
> settings.

That is true but it's much worse to rely on the user using VC. Hooking
into the build process isn't foolproof, but it's much better than
hooking into every possible version-control package for Emacs,
including ones like Magit which aren't built in to Emacs.

I don't think there really is any perfect solution to the problem of
configuring Git. You just have to accept that in some circumstances,
the configuration won't be done correctly, so that's why it's
important for there to be more robust checks run on CI when possible.

But really, hooking into the build process is probably the easiest way
to reach the most users with the least maintenance overhead and
inconvenience.

On Wed, Aug 2, 2017 at 10:26 AM, Richard Stallman <[hidden email]> wrote:

> [[[ To any NSA and FBI agents reading my email: please consider    ]]]
> [[[ whether defending the US Constitution against all enemies,     ]]]
> [[[ foreign or domestic, requires you to follow Snowden's example. ]]]
>
>   > > Perhaps there should be a function in Emacs that does this, and some
>   > > VC commands could call it.
>   > I think is for convenience.
>
> But it is unreliable.  You can't rely on the user's having done 'make'
> in the repository before doing the git operations that use these
> settings.
>
>
> --
> Dr Richard Stallman
> President, Free Software Foundation (gnu.org, fsf.org)
> Internet Hall-of-Famer (internethalloffame.org)
> Skype: No way! See stallman.org/skype.html.
>
>

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Why does make run git?

Paul Eggert
In reply to this post by Richard Stallman
On 08/02/2017 10:26 AM, Richard Stallman wrote:
> But it is unreliable.  You can't rely on the user's having done 'make'
> in the repository before doing the git operations that use these
> settings.

You can pretty much rely on users running either Make or ./autogen.sh;
otherwise they can't easily build or run Emacs. So the main problem here
is ./autogen.sh.

Although last year I arranged for ./autogen.sh (when invoked with no
arguments) to do Git settings like Make does, some other developers
objected at the time that this might conflict with their own Git
settings so I changed autogen.sh to not alter Git settings except on
explicit request. Given the confusion on this topic on the meantime,
though, I think the change was mistaken, and in that sense I believe
that you and I are in agreement.

To help improve the situation I propose the attached patch, which
reverts the default ./autogen.sh behavior to be './autogen.sh all', the
way it used to be (briefly). Developers who want autogen.sh to leave
their Git settings alone can continue to use './autogen.sh autoconf'.


0001-Default-autogen.sh-to-all.patch (1K) Download Attachment
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Why does make run git?

Eli Zaretskii
> Cc: [hidden email], [hidden email]
> From: Paul Eggert <[hidden email]>
> Date: Wed, 2 Aug 2017 10:56:53 -0700
>
> To help improve the situation I propose the attached patch, which
> reverts the default ./autogen.sh behavior to be './autogen.sh all', the
> way it used to be (briefly). Developers who want autogen.sh to leave
> their Git settings alone can continue to use './autogen.sh autoconf'.

AFAIU, doing that will silently force the installation of the
Git-related stuff when I run just "make", and Make decides it needs to
invoke autogen.sh, because the fact that I originally used the
non-default argument isn't recorded anywhere, and thus cannot be
replayed.  Right?  If so, I don't think I will like this change.

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Why does make run git?

Paul Eggert
On 08/02/2017 11:31 AM, Eli Zaretskii wrote:
>> To help improve the situation I propose the attached patch, which
>> reverts the default ./autogen.sh behavior to be './autogen.sh all', the
>> way it used to be (briefly). Developers who want autogen.sh to leave
>> their Git settings alone can continue to use './autogen.sh autoconf'.
> AFAIU, doing that will silently force the installation of the
> Git-related stuff when I run just "make", and Make decides it needs to
> invoke autogen.sh, because the fact that I originally used the
> non-default argument isn't recorded anywhere, and thus cannot be
> replayed.  Right?

No, if you run Make from a freshly-checked-out directory, Make calls
"./autogen.sh all", which sets up Git. The proposed patch doesn't affect
this.

And if you run "./autogen.sh autoconf" and then run Make, Make won't set
up Git. The proposed patch doesn't affect this either.

The only thing that changes is what happens if you run plain "./autogen.sh".


Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Why does make run git?

Eli Zaretskii
> Cc: [hidden email], [hidden email], [hidden email]
> From: Paul Eggert <[hidden email]>
> Date: Wed, 2 Aug 2017 11:42:43 -0700
>
> On 08/02/2017 11:31 AM, Eli Zaretskii wrote:
> >> To help improve the situation I propose the attached patch, which
> >> reverts the default ./autogen.sh behavior to be './autogen.sh all', the
> >> way it used to be (briefly). Developers who want autogen.sh to leave
> >> their Git settings alone can continue to use './autogen.sh autoconf'.
> > AFAIU, doing that will silently force the installation of the
> > Git-related stuff when I run just "make", and Make decides it needs to
> > invoke autogen.sh, because the fact that I originally used the
> > non-default argument isn't recorded anywhere, and thus cannot be
> > replayed.  Right?
>
> No, if you run Make from a freshly-checked-out directory, Make calls
> "./autogen.sh all", which sets up Git. The proposed patch doesn't affect
> this.
>
> And if you run "./autogen.sh autoconf" and then run Make, Make won't set
> up Git. The proposed patch doesn't affect this either.

Maybe I'm confused, but top-level Makefile.in has this:

  $(srcdir)/configure: $(srcdir)/configure.ac $(srcdir)/m4/*.m4
          cd $(srcdir) && ./autogen.sh

This seems to say that if configure.ac or one of the *.m4 files become
newer than configure, either because I edit them or because I just
pulled from upstream, then Make will run "./autogen.sh", which under
the default you propose will configure Git.  If that's not true, then
what am I missing?

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Why does make run git?

Paul Eggert
On 08/02/2017 12:10 PM, Eli Zaretskii wrote:
> Maybe I'm confused, but top-level Makefile.in has this:
>
>    $(srcdir)/configure: $(srcdir)/configure.ac $(srcdir)/m4/*.m4
>  cd $(srcdir) && ./autogen.sh
>
Ah, you're right, that should say './autogen.sh autoconf', not plain
'./autogen.sh', since the intent is only to build 'configure'. This is a
typo regardless of whether my recently-proposed patch is accepted, and I
fixed it with the attached.

0001-Clarify-when-autogen.sh-should-run-only-autoconf.patch (1K) Download Attachment
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Why does make run git?

Richard Stallman
In reply to this post by Radon Rosborough
[[[ To any NSA and FBI agents reading my email: please consider    ]]]
[[[ whether defending the US Constitution against all enemies,     ]]]
[[[ foreign or domestic, requires you to follow Snowden's example. ]]]

  > That is true but it's much worse to rely on the user using VC. Hooking
  > into the build process isn't foolproof, but it's much better than
  > hooking into every possible version-control package for Emacs,
  > including ones like Magit which aren't built in to Emacs.

Not necessarily.  We could make it a documented Emacs facility,
which could be used by any repository, and all front-ends ought
to call it.

--
Dr Richard Stallman
President, Free Software Foundation (gnu.org, fsf.org)
Internet Hall-of-Famer (internethalloffame.org)
Skype: No way! See stallman.org/skype.html.


Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Why does make run git?

Radon Rosborough
> We could make it a documented Emacs facility,
> which could be used by any repository, and all front-ends ought
> to call it.

I don't object, but this sounds a little bit like overkill to me. This
would be a significant amount of added complexity (e.g. notifying all
known authors of version-control packages [maybe not that many?] and
hoping they incorporate supporting changes; dealing with safety [is
safe-local-variable-values enough?]; working out the appropriate scope
[is this just for Git? can you also run arbitrary shell commands?];
designing an interface that will cover all common use cases; making
sure that this doesn't impact performance; deciding whether to
override existing configuration and hooks; etc.), and it still doesn't
really fix the problem since plenty of people will just use Git
directly.

I'd vote for leaving the situation as-is, since there's no general way
to prevent people from using Git without first configuring it. The
"right" solution is to not rely on the local Git configuration for
anything mission-critical.

OTOH, since this is a relatively common problem, maybe somebody has
already invented a general-purpose solution. In that case perhaps
Emacs should support that standard. I don't think Emacs should create
its own standard in this case.

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Why does make run git?

Paul Eggert
In reply to this post by Paul Eggert
Paul Eggert wrote:
> This is a typo regardless of whether my recently-proposed patch is accepted, and
> I fixed it with the attached.

No further comment after the fix, and as the fix means that the proposed patch
seems to be the right thing to do I installed it, along with a followup cleanup
patch. Both patches are attached.

0001-Default-autogen.sh-to-all.patch (1K) Download Attachment
0002-Simplify-re-and-document-autoconf.sh-all.patch (1K) Download Attachment
12
Loading...