makeinfo 4.7

makeinfo 4.7

[Kim F. Storm]

According to the README file, makeinfo 4.2 or later is needed:

`man'       holds the source code for the Emacs Manual.  If you modify the
            manual sources, you will need the `makeinfo' program to produce
            an updated manual.  `makeinfo' is part of the GNU Texinfo
            package; you need version 4.2 or later of Texinfo.

But rcirc.texi uses indicateurl which was added in 4.7

Didn't there use to be something in configure that checked that the
required version of makeinfo is installed??

-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk

Re: makeinfo 4.7

[Eli Zaretskii]

> From: storm@cua.dk (Kim F. Storm)
> Date: Fri, 25 Aug 2006 23:54:36 +0200
> 
> According to the README file, makeinfo 4.2 or later is needed:
> 
> `man'       holds the source code for the Emacs Manual.  If you modify the
>             manual sources, you will need the `makeinfo' program to produce
>             an updated manual.  `makeinfo' is part of the GNU Texinfo
>             package; you need version 4.2 or later of Texinfo.
> 
> But rcirc.texi uses indicateurl which was added in 4.7

Why does rcirc.texi need @indicateurl?  Can't it do without it?

I don't think we should gratuitously raise the minimal version of
support tool required to rebuild Emacs, even though Texinfo 4.7 was
released quite some time ago.

Re: makeinfo 4.7

[Kim F. Storm]

Eli Zaretskii <eliz@gnu.org> writes:

>> But rcirc.texi uses indicateurl which was added in 4.7
>
> Why does rcirc.texi need @indicateurl?  Can't it do without it?

I'm not the Texinfo expert here -- so ???

> I don't think we should gratuitously raise the minimal version of
> support tool required to rebuild Emacs, even though Texinfo 4.7 was
> released quite some time ago.

I agree (but just upgraded to 4.8 anyway)!

-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk

Re: makeinfo 4.7

[Alan Mackenzie]

Morning, Eli!

On Fri, Aug 25, 2006 at 07:03:02PM -0400, Eli Zaretskii wrote:
> > From: storm@cua.dk (Kim F. Storm)
> > Date: Fri, 25 Aug 2006 23:54:36 +0200

> > According to the README file, makeinfo 4.2 or later is needed:

> > `man'       holds the source code for the Emacs Manual.  If you modify the
> >             manual sources, you will need the `makeinfo' program to produce
> >             an updated manual.  `makeinfo' is part of the GNU Texinfo
> >             package; you need version 4.2 or later of Texinfo.

> > But rcirc.texi uses indicateurl which was added in 4.7

> Why does rcirc.texi need @indicateurl?  Can't it do without it?

> I don't think we should gratuitously raise the minimal version of
> support tool required to rebuild Emacs, even though Texinfo 4.7 was
> released quite some time ago.

cc-mode.texi uses @headitem, which was also first introduced in Texinfo
4.7.  How many other manuals also use 4.[78] features?

I suggest we just mandate 4.7 (or even 4.8), and be done.

-- 
Alan.

Re: makeinfo 4.7

[Eli Zaretskii]

> Date: Sat, 26 Aug 2006 09:11:49 +0100
> From: Alan Mackenzie <acm@muc.de>
> Cc: "Kim F. Storm" <storm@cua.dk>, emacs-devel@gnu.org
> 
> cc-mode.texi uses @headitem, which was also first introduced in Texinfo
> 4.7.  How many other manuals also use 4.[78] features?

If someone can run "make info" with makeinfo 4.2, we will know for
sure.

> I suggest we just mandate 4.7 (or even 4.8), and be done.

Why, because it's the easy way out?

Re: makeinfo 4.7

[Eli Zaretskii]

> From: Eli Zaretskii <eliz@gnu.org>
> Date: Sat, 26 Aug 2006 08:15:46 -0400
> Cc: emacs-devel@gnu.org, storm@cua.dk
> 
> > Date: Sat, 26 Aug 2006 09:11:49 +0100
> > From: Alan Mackenzie <acm@muc.de>
> > Cc: "Kim F. Storm" <storm@cua.dk>, emacs-devel@gnu.org
> > 
> > cc-mode.texi uses @headitem, which was also first introduced in Texinfo
> > 4.7.  How many other manuals also use 4.[78] features?
> 
> If someone can run "make info" with makeinfo 4.2, we will know for
> sure.

I just did try that, and these two manuals are the only ones that fail
to build with makeinfo 4.2.  cc-mode uses @headitem and @tie, and rcirc
uses @indicateurl.

So I think it should be fairly easy to fix the manuals and avoid
forcing more people to gratuitously upgrade their Texinfo
installations.  Personally, I hate it when a new package I install (in
this case Emacs) requires me to upgrade a whole slew of other packages
for no good reason; I certainly wouldn't want us to do that to others.

Re: makeinfo 4.7

[Alan Mackenzie]

Hi, Eli!

On Sat, Aug 26, 2006 at 08:46:53AM -0400, Eli Zaretskii wrote:
> > From: Eli Zaretskii <eliz@gnu.org>
> > Date: Sat, 26 Aug 2006 08:15:46 -0400
> > Cc: emacs-devel@gnu.org, storm@cua.dk

> > > Date: Sat, 26 Aug 2006 09:11:49 +0100
> > > From: Alan Mackenzie <acm@muc.de>
> > > Cc: "Kim F. Storm" <storm@cua.dk>, emacs-devel@gnu.org

> > > cc-mode.texi uses @headitem, which was also first introduced in
> > > Texinfo 4.7.  How many other manuals also use 4.[78] features?

> > If someone can run "make info" with makeinfo 4.2, we will know for
> > sure.

> I just did try that, and these two manuals are the only ones that fail
> to build with makeinfo 4.2.  cc-mode uses @headitem and @tie, and rcirc
> uses @indicateurl.

> So I think it should be fairly easy to fix the manuals and avoid
> forcing more people to gratuitously upgrade their Texinfo
> installations.  Personally, I hate it when a new package I install (in
> this case Emacs) requires me to upgrade a whole slew of other packages
> for no good reason; I certainly wouldn't want us to do that to others.

I can certainly empathise wholeheartedly with that.  ;-)

@w can be used instead of @tie.  Is there a canonical backwards
compatible equivalent of @headitem?  Perhaps @b round each word?

-- 
Alan.

Re: makeinfo 4.7

[Eli Zaretskii]

> Date: Sat, 26 Aug 2006 15:23:31 +0100
> From: Alan Mackenzie <acm@muc.de>
> Cc: emacs-devel@gnu.org
> 
> Is there a canonical backwards compatible equivalent of @headitem?
> Perhaps @b round each word?

It's @b (but only @iftex) and a "-------------------" line (only
@ifnottex).

Re: makeinfo 4.7

[Richard Stallman]

    `man'       holds the source code for the Emacs Manual.  If you modify the
		manual sources, you will need the `makeinfo' program to produce
		an updated manual.  `makeinfo' is part of the GNU Texinfo
		package; you need version 4.2 or later of Texinfo.

    But rcirc.texi uses indicateurl which was added in 4.7

The statement that Texinfo 4.2 is needed was added in 2002.
Maybe we should just update that to say "4.7".
Does anyone see any reason why not?

When was Texinfo 4.7 released?

Re: makeinfo 4.7

[Karl Berry]

    Maybe we should just update that to say "4.7".
    Does anyone see any reason why not?

No.  For that matter, it's pretty common to simply require the "latest
released version" to rebuild stuff, if you make changes.

    When was Texinfo 4.7 released?

April 2004.

Re: makeinfo 4.7

[Richard Stallman]

	When was Texinfo 4.7 released?

    April 2004.

That is over 2 years.  So it is ok for Emacs to need it.

Re: makeinfo 4.7

[Eli Zaretskii]

> From: Richard Stallman <rms@gnu.org>
> Date: Sun, 27 Aug 2006 10:34:32 -0400
> Cc: emacs-devel@gnu.org, storm@cua.dk
> 
> 	When was Texinfo 4.7 released?
> 
>     April 2004.
> 
> That is over 2 years.  So it is ok for Emacs to need it.

For only 3 directives that are very easy to change back to
back-compatible ones, I think such a requirement would be annoyingly
unjustified.

Re: makeinfo 4.7

[David Kastrup]

Eli Zaretskii <eliz@gnu.org> writes:

>> From: Richard Stallman <rms@gnu.org>
>> Date: Sun, 27 Aug 2006 10:34:32 -0400
>> Cc: emacs-devel@gnu.org, storm@cua.dk
>> 
>> 	When was Texinfo 4.7 released?
>> 
>>     April 2004.
>> 
>> That is over 2 years.  So it is ok for Emacs to need it.
>
> For only 3 directives that are very easy to change back to
> back-compatible ones, I think such a requirement would be annoyingly
> unjustified.

Who would be annoyed by it?  IIRC, the distribution tarballs come with
premade info files, anyway.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: makeinfo 4.7

[Eli Zaretskii]

> Cc: rms@gnu.org,  emacs-devel@gnu.org,  storm@cua.dk,  karl@freefriends.org
> From: David Kastrup <dak@gnu.org>
> Date: Sun, 27 Aug 2006 18:26:39 +0200
> 
> > For only 3 directives that are very easy to change back to
> > back-compatible ones, I think such a requirement would be annoyingly
> > unjustified.
> 
> Who would be annoyed by it?

People who don't have Texinfo 4.7 or later installed.

> IIRC, the distribution tarballs come with premade info files,
> anyway.

I thought the GNU project was about being able to modify the manual,
not only use whatever is shipped with the tarballs.

On top of that, a large number of users who aren't developers are
using the CVS code, where you must have Texinfo to produce the manual.

Re: makeinfo 4.7

[David Kastrup]

Eli Zaretskii <eliz@gnu.org> writes:

>> Cc: rms@gnu.org,  emacs-devel@gnu.org,  storm@cua.dk,  karl@freefriends.org
>> From: David Kastrup <dak@gnu.org>
>> Date: Sun, 27 Aug 2006 18:26:39 +0200
>> 
>> > For only 3 directives that are very easy to change back to
>> > back-compatible ones, I think such a requirement would be
>> > annoyingly unjustified.
>> 
>> Who would be annoyed by it?
>
> People who don't have Texinfo 4.7 or later installed.

_And_ need to run makeinfo in the first place.

>> IIRC, the distribution tarballs come with premade info files,
>> anyway.
>
> I thought the GNU project was about being able to modify the manual,
> not only use whatever is shipped with the tarballs.

If you are modifying the manual, you can be expected to be using
development tools that are not from the stone age.  That is nothing an
inexperienced user would do.

> On top of that, a large number of users who aren't developers are
> using the CVS code, where you must have Texinfo to produce the
> manual.

Again, this is nothing an inexperienced user would do.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: makeinfo 4.7

[Eli Zaretskii]

> Cc: emacs-devel@gnu.org,  karl@freefriends.org
> From: David Kastrup <dak@gnu.org>
> Date: Mon, 28 Aug 2006 00:30:01 +0200
> 
> > I thought the GNU project was about being able to modify the manual,
> > not only use whatever is shipped with the tarballs.
> 
> If you are modifying the manual, you can be expected to be using
> development tools that are not from the stone age.

If it is really necessary to ask them to upgrade, I agree.  But this
is not our case.  Are you saying we should ask people to upgrade for
no good reason at all?

> That is nothing an inexperienced user would do.

There's nothing in "make info" that requires any particular
experience, IMO.

> > On top of that, a large number of users who aren't developers are
> > using the CVS code, where you must have Texinfo to produce the
> > manual.
> 
> Again, this is nothing an inexperienced user would do.

I think you should check your facts.  It is no longer true that only
hackers use the CVS code.

Re: makeinfo 4.7

[David Kastrup]

Eli Zaretskii <eliz@gnu.org> writes:

>> Cc: emacs-devel@gnu.org,  karl@freefriends.org
>> From: David Kastrup <dak@gnu.org>
>> Date: Mon, 28 Aug 2006 00:30:01 +0200
>> 
>> > I thought the GNU project was about being able to modify the manual,
>> > not only use whatever is shipped with the tarballs.
>> 
>> If you are modifying the manual, you can be expected to be using
>> development tools that are not from the stone age.
>
> If it is really necessary to ask them to upgrade, I agree.  But this
> is not our case.  Are you saying we should ask people to upgrade for
> no good reason at all?
>
>> That is nothing an inexperienced user would do.
>
> There's nothing in "make info" that requires any particular
> experience, IMO.

"make info" will not do anything in a released tarball unless you
touched the manual sources.

>> > On top of that, a large number of users who aren't developers are
>> > using the CVS code, where you must have Texinfo to produce the
>> > manual.
>> 
>> Again, this is nothing an inexperienced user would do.
>
> I think you should check your facts.  It is no longer true that only
> hackers use the CVS code.

Inexperienced users use CVS code somebody else compiled.  Basically
all GNU/Linux distributions provide some emacs-snapshot package, and
there are several precompiled offers for MacOSX and Windows.

I don't see somebody juggling with CVS and compiling himself and so on
use a makeinfo from several years ago and being unable to upgrade.

It just does not make sense.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: makeinfo 4.7

[Eli Zaretskii]

> Cc: emacs-devel@gnu.org,  karl@freefriends.org
> From: David Kastrup <dak@gnu.org>
> Date: Mon, 28 Aug 2006 01:03:59 +0200
> 
> >> That is nothing an inexperienced user would do.
> >
> > There's nothing in "make info" that requires any particular
> > experience, IMO.
> 
> "make info" will not do anything in a released tarball unless you
> touched the manual sources.

The manual is mostly plain text, so there's no special experience
required to touch that, either.  For example, suppose one finds a typo
or stale info.

> >> > On top of that, a large number of users who aren't developers are
> >> > using the CVS code, where you must have Texinfo to produce the
> >> > manual.
> >> 
> >> Again, this is nothing an inexperienced user would do.
> >
> > I think you should check your facts.  It is no longer true that only
> > hackers use the CVS code.
> 
> Inexperienced users use CVS code somebody else compiled.

I don't think this is true.  I'd be willing to change my mind if
presented with statistics, though.

Re: makeinfo 4.7

[David Kastrup]

Eli Zaretskii <eliz@gnu.org> writes:

>> Cc: emacs-devel@gnu.org,  karl@freefriends.org
>> From: David Kastrup <dak@gnu.org>
>> Date: Mon, 28 Aug 2006 01:03:59 +0200
>> 
>> >> That is nothing an inexperienced user would do.
>> >
>> > There's nothing in "make info" that requires any particular
>> > experience, IMO.
>> 
>> "make info" will not do anything in a released tarball unless you
>> touched the manual sources.
>
> The manual is mostly plain text, so there's no special experience
> required to touch that, either.  For example, suppose one finds a
> typo or stale info.

If one feels inclined to meddle with the manual, one will have the
required experience for using not completely outdated software for the
purpose of compiling brand new software.

Eli, it is abundantly clear that no matter what I'll say regarding
this topic, you will have the last word on this matter.

Suit yourself.  I guess everything that is not plain silly has been
said already, and then some.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: makeinfo 4.7

[Kim F. Storm]

Eli Zaretskii <eliz@gnu.org> writes:

> If it is really necessary to ask them to upgrade, I agree.  But this
> is not our case.  Are you saying we should ask people to upgrade for
> no good reason at all?

When I started (again) to hack on CVS emacs years ago, I also had to
upgrade makeinfo (to 4.3 at that time).  Now, years later, I had
to upgrade to makeinfo 4.8.  Fine!  It took just 10 minutes.

Just like we prefer that users upgrade to the latest version of emacs
(to avoid spending time on bug reports for older versions), I'm sure
the maintainers of makeinfo also prefer people to use the latest
version.

So why should emacs (of all projects) not use an up-to-date version of
makeinfo?  I don't see the problem at all.  


Still, wasn't there a time where the make process would barf with
an error message telling what version of makeinfo was required if
the installed version is too old???


BTW, isn't it about time [after the release] to get rid of the P_
macro, and assume that all systems have an ANSI-C complient compiler
these days?

-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk

Re: makeinfo 4.7

[Karl Berry]

    BTW, isn't it about time [after the release] to get rid of the P_
    macro, and assume that all systems have an ANSI-C complient compiler
    these days?

For what it's worth, many other packages have assumed an ANSI C compiler
for some time now, with no significant reported problems (as far as I
can tell).

Re: makeinfo 4.7

[Stefan Monnier]

> BTW, isn't it about time [after the release] to get rid of the P_
> macro, and assume that all systems have an ANSI-C complient compiler
> these days?

100% agreement.  As a matter of fact, I suspect that the current
code already cannot be compiled without an ANSI-C compiler.


        Stefan

Re: makeinfo 4.7

[Eli Zaretskii]

> Cc: Eli Zaretskii <eliz@gnu.org>,  emacs-devel@gnu.org,  karl@freefriends.org
> From: Stefan Monnier <monnier@iro.umontreal.ca>
> Date: Tue, 29 Aug 2006 12:49:52 -0400
> 
> > BTW, isn't it about time [after the release] to get rid of the P_
> > macro, and assume that all systems have an ANSI-C complient compiler
> > these days?
> 
> 100% agreement.  As a matter of fact, I suspect that the current
> code already cannot be compiled without an ANSI-C compiler.

I hope no one is considering to do that now.  September, when the
pretest is supposed to start, is only one day away.  If we are serious
about that goal, we should not start fixing that which ain't broken
before we release Emacs 22.1.

Re: makeinfo 4.7

[Kim F. Storm]

Eli Zaretskii <eliz@gnu.org> writes:

>> Cc: Eli Zaretskii <eliz@gnu.org>,  emacs-devel@gnu.org,  karl@freefriends.org
>> From: Stefan Monnier <monnier@iro.umontreal.ca>
>> Date: Tue, 29 Aug 2006 12:49:52 -0400
>> 
>> > BTW, isn't it about time [after the release] to get rid of the P_
>> > macro, and assume that all systems have an ANSI-C complient compiler
>> > these days?
>> 
>> 100% agreement.  As a matter of fact, I suspect that the current
>> code already cannot be compiled without an ANSI-C compiler.
>
> I hope no one is considering to do that now.  September, when the
> pretest is supposed to start, is only one day away.  If we are serious
> about that goal, we should not start fixing that which ain't broken
> before we release Emacs 22.1.

(search-backward "[after the release]")   :-)


-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk

Re: makeinfo 4.7

[Eli Zaretskii]

> Cc: David Kastrup <dak@gnu.org>,  karl@freefriends.org,  emacs-devel@gnu.org
> From: storm@cua.dk (Kim F. Storm)
> Date: Mon, 28 Aug 2006 10:20:36 +0200
> 
> Just like we prefer that users upgrade to the latest version of emacs
> (to avoid spending time on bug reports for older versions), I'm sure
> the maintainers of makeinfo also prefer people to use the latest
> version.
> 
> So why should emacs (of all projects) not use an up-to-date version of
> makeinfo?  I don't see the problem at all.  

I already explained why in this thread.  If you read that and it is
still not clear, I'm sorry, but I cannot say it any clearer than I
already did.

Re: makeinfo 4.7

[Kim F. Storm]

Eli Zaretskii <eliz@gnu.org> writes:

>> So why should emacs (of all projects) not use an up-to-date version of
>> makeinfo?  I don't see the problem at all.  
>
> I already explained why in this thread.  If you read that and it is
> still not clear, I'm sorry, but I cannot say it any clearer than I
> already did.

You said it very clearly -- I just don't buy the argument :-)

-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk

Re: makeinfo 4.7

[Juri Linkov]

>> Just like we prefer that users upgrade to the latest version of emacs
>> (to avoid spending time on bug reports for older versions), I'm sure
>> the maintainers of makeinfo also prefer people to use the latest
>> version.
>>
>> So why should emacs (of all projects) not use an up-to-date version of
>> makeinfo?  I don't see the problem at all.

Just today these new Texinfo directives caused trouble for me.
I successfully compiled CVS Emacs on quite an old GNU/Linux server
(Red Hat 9.0).  But there were only two errors in the whole build process:
in two Info files cc-mode.texi and rcirc.texi, since the version of Texinfo
on this server is 4.3:

makeinfo --force cc-mode.texi
cc-mode.texi:1628: Unknown command `headitem'.
cc-mode.texi:1661: Unknown command `headitem'.
make[1]: *** [../info/ccmode] Error 2

makeinfo --force rcirc.texi
rcirc.texi:122: Unknown command `indicateurl'.
rcirc.texi:122: Misplaced {.
rcirc.texi:122: Misplaced }.
rcirc.texi:428: Unknown command `indicateurl'.
rcirc.texi:428: Misplaced {.
rcirc.texi:428: Misplaced }.
make[1]: *** [../info/rcirc] Error 2

Due to these errors I was unable to consult the Info documentation from
the rest of Info files that should be created after the compilation that
was interrupted by makeinfo errors.  And I had no time/need to update
Texinfo on this server.

Please change back these two directives to backward-compatible ones.

-- 
Juri Linkov
http://www.jurta.org/emacs/

Re: makeinfo 4.7

[David Kastrup]

Juri Linkov <juri@jurta.org> writes:

>>> Just like we prefer that users upgrade to the latest version of emacs
>>> (to avoid spending time on bug reports for older versions), I'm sure
>>> the maintainers of makeinfo also prefer people to use the latest
>>> version.
>>>
>>> So why should emacs (of all projects) not use an up-to-date version of
>>> makeinfo?  I don't see the problem at all.
>
> Just today these new Texinfo directives caused trouble for me.  I
> successfully compiled CVS Emacs on quite an old GNU/Linux server
> (Red Hat 9.0).

Why would you not upgrade the server?  It has reached end of service
long ago.  There are not even security updates anymore.

At the current point of time, I get password crack attacks via ssh
about twice a week on my dialup account (so randomly assigned IP
numbers), usually several hundred login attempts about a second apart.

One of the infected machines was on a university network, and the
system administrators there were kind enough to tell me what computer
it was.  It was an embedded system running an old version of RedHat.

I really don't think that we should go out of our depth to support
installations like that.

> Due to these errors I was unable to consult the Info documentation
> from the rest of Info files that should be created after the
> compilation that was interrupted by makeinfo errors.  And I had no
> time/need to update Texinfo on this server.

You had the time to check out, configure and install Emacs from CVS,
but did not have the time to run

yum update texinfo

(or fetch and install a suitable RPM from one of many repositories).

Unhn.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: makeinfo 4.7

[Miles Bader]

David Kastrup <dak@gnu.org> writes:
>> Just today these new Texinfo directives caused trouble for me.  I
>> successfully compiled CVS Emacs on quite an old GNU/Linux server
>> (Red Hat 9.0).
>
> Why would you not upgrade the server?  It has reached end of service
> long ago.  There are not even security updates anymore.

When I have to install emacs on an old machine, and feel too annoyed to
update makeinfo, I just don't install the manuals!  Usually I only use
emacs casually on those machines anyway...

Doesn't seem like a big deal to me.

-Miles
-- 
Saa, shall we dance?  (from a dance-class advertisement)

Re: makeinfo 4.7

[Eli Zaretskii]

> Cc: Eli Zaretskii <eliz@gnu.org>,  karl@freefriends.org,  emacs-devel@gnu.org,
> 	  storm@cua.dk
> From: David Kastrup <dak@gnu.org>
> Date: Thu, 31 Aug 2006 01:33:57 +0200
> 
> Why would you not upgrade the server?  It has reached end of service
> long ago.  There are not even security updates anymore.

Didn't we agree to stop arguing about this?

I said I will change those two manuals to not use the offending
directives, and I will, in a couple of days.  Please stop adding more
and more of the same verbiage, about which we agreed to disagree.

Re: makeinfo 4.7

[Juri Linkov]

> Why would you not upgrade the server?  It has reached end of service
> long ago.  There are not even security updates anymore.

I can't decide what to install on every server where I have an account.

> You had the time to check out, configure and install Emacs from CVS,
> but did not have the time to run
>
> yum update texinfo
>
> (or fetch and install a suitable RPM from one of many repositories).

It is very easy to fetch and compile the latest version of Emacs from CVS
in the home directory on any GNU/Linux machine without root privileges
with just two commands: `cvs ... co emacs' and `./configure && make bootstrap'.
I don't want to get the latest version of any other program to be able
to build Emacs ;)

-- 
Juri Linkov
http://www.jurta.org/emacs/

Re: makeinfo 4.7

[Richard Stallman]

    > That is over 2 years.  So it is ok for Emacs to need it.

    For only 3 directives that are very easy to change back to
    back-compatible ones, I think such a requirement would be annoyingly
    unjustified.

I do not consider it a serious problem to depend on Texinfo 4.7 in
these circumstances.  However, if you want to take out those 3
commands, I don't mind.

Re: makeinfo 4.7

[Eli Zaretskii]

> From: Richard Stallman <rms@gnu.org>
> CC: karl@freefriends.org, emacs-devel@gnu.org, storm@cua.dk
> Date: Mon, 28 Aug 2006 18:09:42 -0400
> 
>     For only 3 directives that are very easy to change back to
>     back-compatible ones, I think such a requirement would be annoyingly
>     unjustified.
> 
> I do not consider it a serious problem to depend on Texinfo 4.7 in
> these circumstances.  However, if you want to take out those 3
> commands, I don't mind.

Will do.

Re: makeinfo 4.7

[Eli Zaretskii]

> Date: Wed, 30 Aug 2006 13:44:15 +0300
> From: Eli Zaretskii <eliz@gnu.org>
> Cc: emacs-devel@gnu.org, storm@cua.dk, karl@freefriends.org
> 
> > I do not consider it a serious problem to depend on Texinfo 4.7 in
> > these circumstances.  However, if you want to take out those 3
> > commands, I don't mind.
> 
> Will do.

Done.

Re: makeinfo 4.7

[Stefan Monnier]

> 	When was Texinfo 4.7 released?
>     April 2004.
> That is over 2 years.  So it is ok for Emacs to need it.

You're probably right.  Could we just check that all the major GNU/Linux
distributions come with such a "recent" version of Texinfo (I'm of course
mostly thinking of Debian stable which has the tradition of upgrading late).


        Stefan

Re: makeinfo 4.7

[Eli Zaretskii]

> From: Stefan Monnier <monnier@iro.umontreal.ca>
> Date: Mon, 28 Aug 2006 11:14:59 -0400
> Cc: emacs-devel@gnu.org, storm@cua.dk, Karl Berry <karl@freefriends.org>
> 
> Could we just check that all the major GNU/Linux distributions come
> with such a "recent" version of Texinfo

Are we going to make an assumption that Emacs is built only on
GNU/Linux systems?