What about putting CC Mode 5.32 into Emacs?

What about putting CC Mode 5.32 into Emacs?

[Alan Mackenzie]

Hi, Emacs.

A few weeks ago, I announced the release of CC Mode 5.32.  More
precisely, in this email:

    Date: Thu, 25 Aug 2011 08:50:46 +0000
    To: emacs-devel@gnu.org
    Subject: CC Mode 5.32 has been released.
    Message-ID: <20110825085046.GA2737@acm.acm>

There was no reaction whatsoever from anybody here.

On the 4th October, Carsten Mattner posted a request for CC Mode 5.32 to
be incorporated into Emacs 24.  His post also drew no reaction.

Is there some sort of filter in effect which stops anything about 5.32
being answered?  ;-)

So, is there any chance of getting 5.32 into Emacs 24 soon?  The
changes, over and above what is already in Emacs, are few (2 to be
precise) and small.  One of them is important (making C-M-a/e DTRT in
languages like C++ which have nested defuns).

-- 
Alan Mackenzie (Nuremberg, Germany).

Re: What about putting CC Mode 5.32 into Emacs?

[Christoph Scholtes]

On 10/8/2011 7:14 AM, Alan Mackenzie wrote:

> So, is there any chance of getting 5.32 into Emacs 24 soon?  The
> changes, over and above what is already in Emacs, are few (2 to be
> precise) and small.  One of them is important (making C-M-a/e DTRT in
> languages like C++ which have nested defuns).

I'd vote for inclusion right now, especially since the changes are small.

One other thing to consider: is it possible to make CC Mode available 
via ELPA to make it possible to upgrade to newer versions easily even 
after the solidest of all feature freezes?

Christoph

Re: What about putting CC Mode 5.32 into Emacs?

[Alan Mackenzie]

Hi, Christoph,

On Sat, Oct 08, 2011 at 07:37:17AM -0600, Christoph Scholtes wrote:
> On 10/8/2011 7:14 AM, Alan Mackenzie wrote:

> > So, is there any chance of getting 5.32 into Emacs 24 soon?  The
> > changes, over and above what is already in Emacs, are few (2 to be
> > precise) and small.  One of them is important (making C-M-a/e DTRT in
> > languages like C++ which have nested defuns).

> I'd vote for inclusion right now, especially since the changes are small.

> One other thing to consider: is it possible to make CC Mode available 
> via ELPA to make it possible to upgrade to newer versions easily even 
> after the solidest of all feature freezes?

It would be possible, but would mean yet more work.  I've already got
three places to be updated (the project at SourceForge, Emacs, and
XEmacs).  There are download instructions, both for released versions
and the CVS head at <http://cc-mode.sourceforge.net/release.php>.

Would it perhaps be possible to put some sort of forwarding message into
ELPA?

> Christoph

-- 
Alan Mackenzie (Nuremberg, Germany).

Re: What about putting CC Mode 5.32 into Emacs?

[Chong Yidong]

Alan Mackenzie <acm@muc.de> writes:

> So, is there any chance of getting 5.32 into Emacs 24 soon?  The
> changes, over and above what is already in Emacs, are few (2 to be
> precise) and small.  One of them is important (making C-M-a/e DTRT in
> languages like C++ which have nested defuns).

Well, you are the CC mode maintainer, so feel free to exercise your
judgement about what constitutes a bugfix and commit accordingly.
Changes to fix incorrect behavior are surely fine (assuming they don't
involve a total overhaul of the CC mode engine ;-)

Re: What about putting CC Mode 5.32 into Emacs?

[Stefan Monnier]

> Would it perhaps be possible to put some sort of forwarding message into
> ELPA?

We have an `org' package built out of Org's own repository rather than from
the `elpa' branch.  We can definitely do the same for other packages.


        Stefan

Re: What about putting CC Mode 5.32 into Emacs?

[Burton Samograd]

Chong Yidong <cyd@stupidchicken.com> writes:

> Alan Mackenzie <acm@muc.de> writes:
>
>> So, is there any chance of getting 5.32 into Emacs 24 soon?  The
>> changes, over and above what is already in Emacs, are few (2 to be
>> precise) and small.  One of them is important (making C-M-a/e DTRT in
>> languages like C++ which have nested defuns).
>
> Well, you are the CC mode maintainer, so feel free to exercise your
> judgement about what constitutes a bugfix and commit accordingly.
> Changes to fix incorrect behavior are surely fine (assuming they don't
> involve a total overhaul of the CC mode engine ;-)

I certainly hope that 5.32 fixes the indentation issues that I have been
experiencing with the lastest builds of emacs.  It's been quite
unreliable and annoying, although I haven't put up the effort to write a
bug report as I am running bzr emacs...

--
Burton Samograd

Re: What about putting CC Mode 5.32 into Emacs?

[Alan Mackenzie]

Hello, Burton.

On Tue, Oct 11, 2011 at 09:55:20AM -0600, Burton Samograd wrote:

> I certainly hope that 5.32 fixes the indentation issues that I have
> been experiencing with the lastest builds of emacs.  It's been quite
> unreliable and annoying, although I haven't put up the effort to write
> a bug report as I am running bzr emacs...

At the moment, that seems unlikely, since nearly all of 5.32's code is
already in bzr Emacs.  Sorry.

A bug report would be welcome.  Please create the meat of this with C-c
C-b from your buffer - that dumps the entire CC Mode configuration and
often assists massively in tracking down a bug.

> --
> Burton Samograd

-- 
Alan Mackenzie (Nuremberg, Germany).

Re: What about putting CC Mode 5.32 into Emacs?

[Michael Welsh Duggan]

Alan Mackenzie <acm@muc.de> writes:

> Hello, Burton.
>
> On Tue, Oct 11, 2011 at 09:55:20AM -0600, Burton Samograd wrote:
>
>> I certainly hope that 5.32 fixes the indentation issues that I have
>> been experiencing with the lastest builds of emacs.  It's been quite
>> unreliable and annoying, although I haven't put up the effort to write
>> a bug report as I am running bzr emacs...
>
> At the moment, that seems unlikely, since nearly all of 5.32's code is
> already in bzr Emacs.  Sorry.
>
> A bug report would be welcome.  Please create the meat of this with C-c
> C-b from your buffer - that dumps the entire CC Mode configuration and
> often assists massively in tracking down a bug.

Have you looked at bug 9560 recently?

  http://debbugs.gnu.org/cgi/bugreport.cgi?bug=9560

I have added as much debugging information as I could.  I currently have
an emacs -Q on my home machine which is in a bad-indentation state.  I
can leave it up for a few days.

-- 
Michael Welsh Duggan
(md5i@md5i.com)

Re: What about putting CC Mode 5.32 into Emacs?

[Burton Samograd]

"Michael Welsh Duggan" <md5i@md5i.com> writes:

> Alan Mackenzie <acm@muc.de> writes:
>
>> Hello, Burton.
>>
>> On Tue, Oct 11, 2011 at 09:55:20AM -0600, Burton Samograd wrote:
>>
>>> I certainly hope that 5.32 fixes the indentation issues that I have
>>> been experiencing with the lastest builds of emacs.  It's been quite
>>> unreliable and annoying, although I haven't put up the effort to write
>>> a bug report as I am running bzr emacs...
>>
>> At the moment, that seems unlikely, since nearly all of 5.32's code is
>> already in bzr Emacs.  Sorry.
>>
>> A bug report would be welcome.  Please create the meat of this with C-c
>> C-b from your buffer - that dumps the entire CC Mode configuration and
>> often assists massively in tracking down a bug.
>
> Have you looked at bug 9560 recently?
>
>   http://debbugs.gnu.org/cgi/bugreport.cgi?bug=9560

No I haven't, but I did just look it over.

>
> I have added as much debugging information as I could.  I currently have
> an emacs -Q on my home machine which is in a bad-indentation state.  I
> can leave it up for a few days.

It sounds a bit different than what I've been experiencing.  From
memory, it seems like when I'm editing code in a function, it loses the
idea that I am in a function and loses a level of indentation for all
code below a certain point.  Sometimes going up a few lines before the
problem and re-indenting every line fixes it, but it can take a couple
of tries to get it to actually work.  Other times I have to go to near
the start of the function and re-indent everything.  Sometimes nothing
works and I just re-indent it manually.

I don't want to complain without offering a solution, but in the 15+
years I've been using emacs I have *never* had a problem with C code
indentation and have grown to rely on it to help me find syntax errors
in my code.  Why it has unreliable is beyond me when it's worked so well
for so long...I hope this wasn't just change for the sake of change.

--
Burton Samograd

Re: What about putting CC Mode 5.32 into Emacs?

[Michael Welsh Duggan]

Burton Samograd <bsamograd@interalia.com> writes:

> "Michael Welsh Duggan" <md5i@md5i.com> writes:
>
>> Alan Mackenzie <acm@muc.de> writes:
>>
>>> Hello, Burton.
>>>
>>> On Tue, Oct 11, 2011 at 09:55:20AM -0600, Burton Samograd wrote:
>>>
>>>> I certainly hope that 5.32 fixes the indentation issues that I have
>>>> been experiencing with the lastest builds of emacs.  It's been quite
>>>> unreliable and annoying, although I haven't put up the effort to write
>>>> a bug report as I am running bzr emacs...

[...]

>> Have you looked at bug 9560 recently?
>>
>>   http://debbugs.gnu.org/cgi/bugreport.cgi?bug=9560
>
> No I haven't, but I did just look it over.
>
>>
>> I have added as much debugging information as I could.  I currently have
>> an emacs -Q on my home machine which is in a bad-indentation state.  I
>> can leave it up for a few days.
>
> It sounds a bit different than what I've been experiencing.  From
> memory, it seems like when I'm editing code in a function, it loses the
> idea that I am in a function and loses a level of indentation for all
> code below a certain point.  Sometimes going up a few lines before the
> problem and re-indenting every line fixes it, but it can take a couple
> of tries to get it to actually work.  Other times I have to go to near
> the start of the function and re-indent everything.  Sometimes nothing
> works and I just re-indent it manually.

I have seen this as well.  The reason I have documented the one I have
is that I have been able to recreate it fairly regularly, and I am
hoping that the underlying cause is similar.

> I don't want to complain without offering a solution, but in the 15+
> years I've been using emacs I have *never* had a problem with C code
> indentation and have grown to rely on it to help me find syntax errors
> in my code.  Why it has unreliable is beyond me when it's worked so well
> for so long...I hope this wasn't just change for the sake of change.

Definitely not.  I don't recall all the performance problems the caching
code was added to support, but they were very real, and there were a
fair number of complaints because of them.  I applaud the work Alan has
done to attempt to improve cc-mode, although I do worry about the
complexity of the current system.

That said, I will file a bug every time I find a new indentation
problem after a new cc-mode update, whether or not I can recreate it,
especially now that we are in pre-release.  I want to make sure that
people using Emacs after the release don't encounter the same problems I
am encountering.

-- 
Michael Welsh Duggan
(md5i@md5i.com)

Re: What about putting CC Mode 5.32 into Emacs?

[Burton Samograd]

"Michael Welsh Duggan" <md5i@md5i.com> writes:

>> I don't want to complain without offering a solution, but in the 15+
>> years I've been using emacs I have *never* had a problem with C code
>> indentation and have grown to rely on it to help me find syntax errors
>> in my code.  Why it has unreliable is beyond me when it's worked so well
>> for so long...I hope this wasn't just change for the sake of change.
>
> Definitely not.  I don't recall all the performance problems the caching
> code was added to support, but they were very real, and there were a
> fair number of complaints because of them.  I applaud the work Alan has
> done to attempt to improve cc-mode, although I do worry about the
> complexity of the current system.

Which is unfortuate as *I* have never seen or had a problem with the
previous CC modes...  Again, this is open source and if I really wanted
I could roll back in my own version, if I could find it.  CC mode was
one of those things I've always taken for granted since it "just worked".

--
Burton Samograd

Re: What about putting CC Mode 5.32 into Emacs?

[Alan Mackenzie]

Hello again, Burton.

On Wed, Oct 12, 2011 at 09:34:09AM -0600, Burton Samograd wrote:
> "Michael Welsh Duggan" <md5i@md5i.com> writes:

> >> I don't want to complain without offering a solution, but in the 15+
> >> years I've been using emacs I have *never* had a problem with C code
> >> indentation and have grown to rely on it to help me find syntax errors
> >> in my code.  Why it has unreliable is beyond me when it's worked so well
> >> for so long...I hope this wasn't just change for the sake of change.

I can assure you there's been no "change for change's sake" in the
indentation code.  Have a wee look at cc-engine.el (for example,
c-guess-basic-syntax) and you'll understand why.  ;-)

> > Definitely not.  I don't recall all the performance problems the caching
> > code was added to support, but they were very real, and there were a
> > fair number of complaints because of them.  I applaud the work Alan has
> > done to attempt to improve cc-mode, although I do worry about the
> > complexity of the current system.

> Which is unfortuate as *I* have never seen or had a problem with the
> previous CC modes...  Again, this is open source and if I really wanted
> I could roll back in my own version, if I could find it.

There are old versions of CC mode available from
<http://cc-mode@sourceforge.net/release.php>.

> CC mode was one of those things I've always taken for granted since it
> "just worked".

It will just work again.

> --
> Burton Samograd

-- 
Alan Mackenzie (Nuremberg, Germany).