I am experiencing a bug with the php omni-complete, where if i have
multiple buffers open, the omni-complete action (^X^O in insert mode)
will give the following errors on anything other than the first buffer
opened, and be missing quite a lot of the completion entries that are
usually available.

-----------------
Error detected while processing function phpcomplete#CompletePHP:
line  390:
E121: Undefined variable: b:php_keywords
Press ENTER or type command to continue
Error detected while processing function phpcomplete#CompletePHP:
line  390:
E116: Invalid arguments for function extend
Press ENTER or type command to continue
-----------------

As far as i could tell, this was working fine in 7.0aa, and has only
become apparent after upgrading to 7.0b.

--
Alan Briolat <alan@...>

Bram Moolenaar did utter on 31/03/2006 18:51:
> Benji Fisher wrote:
>
>>      I noticed a problem with :hardcopy today.  With my defaults,
>>
>> 'encoding' = utf-8
>> 'printencoding' =
>>
>> on Linux (FC2, GTK2, PS printer) I get an error (E673) when I do
>>
>> :hardcopy
>>
>> I can fix it with
>>
>> :set printencoding=latin1
>>
>> or
>>
>> :set encoding=latin1
>>
>> but it used to work smoothly on vim 6.4.  I do not see any changes
>> mentioned in version7.txt .  I am a few days behind:  vim 7.0c03.
>
> I see it too.  When 'printencoding' is empty it should use the default
> "latin1", as documented.  Mike?

It should.  However this error comes from an earlier check in
mch_print_init() in hardcopy.c.  I'll go back over the logic and supply
a patch in the next day or two.

TTFN

Mike
--
Taglines temporarily suspended - sorry.

>>> Clicking on "Save as..." in the File menu doesn't have any effect
>>> (except showing ":browse confirs saveas" in the status line).
>>>
>>> I run 7.0c02 BETA built with Visual Studio .Net 2003.
>> It works just fine for me.  I have no clue why it wouldn't work for you.
>
>      Let's try some basic diagnosis.  Try
>
> :redir @a
> :menu File.Save\ As\.\.\.
> :redir END
> :put a
>
> and send us the result.  (This is one way to get the output of the :menu
> command into an e-mail, use some other method if you like.)
>
>

Here's what I get:

--- Menus ---
350 Save &As...^I:sav
      n*   :browse confirm saveas<CR>
      v*   <C-C>:browse confirm saveas<CR><C-\><C-G>
      o*   <C-C>:browse confirm saveas<CR><C-\><C-G>
      i*   <C-C>:browse confirm saveas<CR><C-\><C-G>

--
Stefan

Addition of a couple of tab menu items under the Window menu.  More
could probably be added, such as move tab left/right, etc.


*** menu.vim-orig Sat Apr  1 10:31:00 2006
--- menu.vim Sat Apr  1 10:51:04 2006
***************
*** 780,795 ****
   an 70.320 &Window.Sp&lit\ To\ #<Tab>^W^^ <C-W><C-^>
   an 70.330 &Window.Split\ &Vertically<Tab>^Wv <C-W>v
   if has("vertsplit")
!   an <silent> 70.332 &Window.Split\ File\ E&xplorer :call MenuExplOpen()<CR>
!   if !exists("*MenuExplOpen")
!     fun MenuExplOpen()
!       if @% == ""
!  20vsp .
!       else
!  exe "20vsp " . expand("%:p:h")
!       endif
!     endfun
!   endif
   endif
   an 70.335 &Window.-SEP1- 		 <Nop>
   an 70.340 &Window.&Close<Tab>^Wc 	 :confirm close<CR>
--- 780,808 ----
   an 70.320 &Window.Sp&lit\ To\ #<Tab>^W^^ <C-W><C-^>
   an 70.330 &Window.Split\ &Vertically<Tab>^Wv <C-W>v
   if has("vertsplit")
!   let s:doMenuExplOpen=1
!   an <silent> 70.331 &Window.Split\ File\ E&xplorer :call MenuExplOpen(1)<CR>
! endif
! if exists('*tabpagenr')
!   let s:doMenuExplOpen=1
!   an 70.332 &Window.-SEP0- 	 <Nop>
!   an 70.333 &Window.New\ Tab 	 :tabnew<CR>
!   an 70.334 &Window.New\ Tab\ File\ Explorer :call MenuExplOpen(2)<CR>
! endif
! if !exists("*MenuExplOpen") && exists('s:doMenuExplOpen')
!   fun MenuExplOpen(type)
!     if @% == ""
!       let where='.'
!     else
!       let where=expand("%:p:h")
!     endif
!
!     if a:type == 2
!       exe 'tabnew ' . where
!     else
!       exe '20vsp ' . where
!     endif
!   endfun
   endif
   an 70.335 &Window.-SEP1- 		 <Nop>
   an 70.340 &Window.&Close<Tab>^Wc 	 :confirm close<CR>

--
     Alternate definitions, #13: Pokemon (n), A Jamaican proctologist.
Christian J. Robinson <infynity@...> http://infynity.spodzone.com/
    PGP keys: 0x893B0EAF / 0xFB698360   http://infynity.spodzone.com/pgp

Benji Fisher skrev:
>
> P.S.  I see that
>
> :redir @*
>
> works, but
>
> :redir @*
>
> does not.
>

Do you mean that @a works but not @* (you wrote @* on both places)?

For me they both work. However, I'm on Windows where I think * and +
works differently than on Linux.

--
Stefan

Bram Moolenaar skrev:
> Stefan Karlsson wrote:
>
>> Clicking on "Save as..." in the File menu doesn't have any effect
>> (except showing ":browse confirs saveas" in the status line).
>>
>> I run 7.0c02 BETA built with Visual Studio .Net 2003.
>
> It works just fine for me.  I have no clue why it wouldn't work for you.
>
> - Bram
>

:browse set works as expected but none of the other :browse commands
have any effect.

--
Stefan

On Sat, Apr 01, 2006 at 07:56:34PM +0200, Stefan Karlsson wrote:
> Benji Fisher skrev:
> >
> >P.S.  I see that
> >
> >:redir @*
> >
> >works, but
> >
> >:redir @*
> >
> >does not.
> >
>
> Do you mean that @a works but not @* (you wrote @* on both places)?
>
> For me they both work. However, I'm on Windows where I think * and +
> works differently than on Linux.

      Thanks for helping me see what I wrote.  I meant that @* works but
@+ does not (Linux, FC2, GTK2).  "E475: Invalid argument: @+".

					 --Benji Fisher

On 3/31/06, Bram Moolenaar <Bram@...> wrote:
>
> Nicolas Weber wrote:
>
> > > > the 3rd line in runtime/autoload/pycomplete.vim in get_completions() has
> > > > to be
> > > >     stmt = str(stmt)+base
> > > > instead of
> > > >     stmt = stmt+base
> > >
> > > The version of pycomplete.vim that I have already is like that.
> >
> > I just did an 'cvs update', the change is not (yet?) in CVS.
>
> Sorry, I read your remark backwards.  It still is "stmt = stmt+base".
> Aaron?
>

Yeah there's 2 bugs like that that I know of, though the second one is
not so easy - let me work on it today and fix those.  I was planning
on rewriting the whole thing, as it's just dirty, but I don't have a
huge amount of time.

I'll reply here when it's fixed - sometime this weekend.

On Sat, Apr 01, 2006 at 07:49:03PM +0200, Stefan Karlsson wrote:
>
> >>>Clicking on "Save as..." in the File menu doesn't have any effect
> >>>(except showing ":browse confirs saveas" in the status line).
> >>>
> >>>I run 7.0c02 BETA built with Visual Studio .Net 2003.
> >>It works just fine for me.  I have no clue why it wouldn't work for you.
> >
> >     Let's try some basic diagnosis.  Try
> >
> >:menu File.Save\ As\.\.\.
>
> Here's what I get:
>
> --- Menus ---
> 350 Save &As...^I:sav
>     n*   :browse confirm saveas<CR>
>     v*   <C-C>:browse confirm saveas<CR><C-\><C-G>
>     o*   <C-C>:browse confirm saveas<CR><C-\><C-G>
>     i*   <C-C>:browse confirm saveas<CR><C-\><C-G>

      That looks right.  Starting in Normal mode, does

:browse confirm saveas

do anything?  From your other recent post on this thread, I guess not.
I guess that means either that there is something wrong in your system
or some incompatibility with VS.Net.  Are you up to trying a different
compiler?  I see instructions at

http://mywebpage.netscape.com/SharpPeople/vim/howto/index.html

(MSVC++, BCC (Borland?), GCC (MinGW or cygwin))

http://cream.sourceforge.net/vim.html

(Borland)

and

http://users.skynet.be/antoine.mechelynck/vim/compile.htm

(Borland BCC32 or cygwin gcc).  Tony Mechelynck often posts pre-compiled
binaries for Win32, but I guess he has been too busy lately, since the
latest version on his site is 7.0aa-0213 (alpha) and I have not seen any
posts from him in a while.

HTH 			 --Benji Fisher

I've spent most of today trying to build various Vims (6.2, 6.4, 6
series CVS) under OS X Intel.
I configured with "./configure --prefix=/opt/local --disable-darwin" I
only want the console version.
I was getting malloc/free problems in expand_files, it looked like
repl was being freed without being assigned to a valid pointer.
After much insanity, it appears that Apple's GCC causes some breakage
with -O2.  Reducing the optimisation level to -O, it works again.

gcc --version
i686-apple-darwin8-gcc-4.0.1 (GCC) 4.0.1 (Apple Computer, Inc. build 5250)

Can some other OS X Intel owners try this to confirm there is a
problem?  Any idea how to raise this with Apple?

Cheers
Brad

On Sat, 2006-04-01 at 18:31 -0500, Benji Fisher wrote:
>
> ....Are you up to trying a different compiler?  I see instructions
> at
[snip]
>
> http://cream.sourceforge.net/vim.html
>
> (Borland)

Just for the record, Cream now rolls with Cygwin as Tony Mechelynck
does. (We both found Borland had too many issues with Windows XP.)

SF.net CVS is having problems, mail me off-list for our scripts if
anyone's interested.


--
Steve Hall  [ digitect mindspring com ]

On 4/2/06, Brad Beveridge <brad.beveridge@...> wrote:
> Can some other OS X Intel owners try this to confirm there is a
> problem?

Yes, this is a known bug in GCC 4 shipping with OS X (for both
architectures, I'm pretty sure).

> Any idea how to raise this with Apple?

Nope, sorry.

HTH,
Chris Allen

vim7beta! I am using it.

Well, I sometimes am too dangerous, especially with a new thing. I
deliberately set the window size of gvim to a very small one, say, about
five lines. Then I tried out omni-completion, and then gvim began to
repeatedly flash. I guess it's because the window needed for
omnicompletion cannot be opened in such a small room. It doesn't stop. I
pressed CTRL-C and found that gvim began to do another repeating job. I
should say sorry to gvim, for a window that small hardly needs any
omni-completion.

If necessary, please ask me to give a sequence of reproduction steps. I
forgot to record the exact steps.

> Any idea how to raise this with Apple?

You may try submitting a bug report:

   http://developer.apple.com/bugreporter/

You'll have to sign up for a Developer Connection account, first.

-- Dave
Email: david.w.h.chin@...

Hi Bram Moolenaar, you wrote:

>
> That is right, Vim 7 continues to find matches, so that the popup menu
> can be filled.  But you don't have to wait for it to finish, you can
> type something while it's still busy.
>

My 5 cents: I have similar problem with C++ files. I set up 'include' so that it
points to Windows SDK directories so that I can auto-complete standard
functions. But the problem is that 99% of cases I want to auto-complete some
file-local name. And it is really inconvenient that Vim starts to scan
everything over and over again each time I press C-P/C-N. I keep on pressing ESC
constantly to stop that. And in fact it looks like you have to wait - the cursor
stops blinking so that user will feel that some modal thing is going on.

Now I stop wining and would like to propose a solution. I think it would be
great to stop the scanning if Vim doesn't need to build the menu, that is if
completeopt doesn't contain 'menu'.

P.S. Personally, I like all these new features, like spell checking etc., but I
like Vim because it is fast-as-light and I can do everything with a few key
strokes. I can live doing spell checking occasionally outside of Vim, but I
can't live pressing ESC all the time... :-)

P.P.S. Thanks for Vim. I'm really impressed how those Google guys manage to hire
the most talented guys in the world...

--
Alexei Alexandrov

Dnia sobota, 1 kwietnia 2006 16:19, Alan Briolat napisał:
> I am experiencing a bug with the php omni-complete, where if i have
> multiple buffers open, the omni-complete action (^X^O in insert mode)
> will give the following errors on anything other than the first buffer
> opened, and be missing quite a lot of the completion entries that are
> usually available.
>
> -----------------
> Error detected while processing function phpcomplete#CompletePHP:
> line  390:
> E121: Undefined variable: b:php_keywords

Oops. Sorry. Fast fix is to do::

	 %s/b:php_keywords/g:php_keywords/g

Sending fixed version to Bram.

m.

Brad Beveridge wrote:

> I've spent most of today trying to build various Vims (6.2, 6.4, 6
> series CVS) under OS X Intel.
> I configured with "./configure --prefix=3D/opt/local --disable-darwin" I
> only want the console version.
> I was getting malloc/free problems in expand_files, it looked like
> repl was being freed without being assigned to a valid pointer.
> After much insanity, it appears that Apple's GCC causes some breakage
> with -O2.  Reducing the optimisation level to -O, it works again.
>
> gcc --version
> i686-apple-darwin8-gcc-4.0.1 (GCC) 4.0.1 (Apple Computer, Inc. build 5250)
>
> Can some other OS X Intel owners try this to confirm there is a
> problem?  Any idea how to raise this with Apple?

This is a gcc 4.0.1 problem.  The optimizer reorders the code, so that a
block of memory is freed before it is used.  I stepped through it with
the debugger to verify this actually happens.

I haven't reported it, the docs on how to report bugs is very confusing.

--
This message contains 78% recycled characters.

  /// Bram Moolenaar -- Bram@... -- http://www.Moolenaar.net   \\\
///        sponsor Vim, vote for features -- http://www.Vim.org/sponsor/ \\\
\\\        download, build and distribute -- http://www.A-A-P.org        ///
  \\\            help me help AIDS victims -- http://www.ICCF.nl         ///

Hi list,

vim exists with a floating point exception when i do (in bash):

echo "# vim:lines=1000" > test; vim test

tested with vim 7.0b beta on Mandrake 10.2.
Is this the right behaviour?

Tobi

Tobias Klein wrote:

> vim exists with a floating point exception when i do (in bash):
>
> echo "# vim:lines=1000" > test; vim test
>
> tested with vim 7.0b beta on Mandrake 10.2.
> Is this the right behaviour?

Are you kidding?

I'll fix it.

--
ARTHUR: It is I, Arthur, son of Uther Pendragon, from the castle of Camelot.
         King of all Britons, defeator of the Saxons, sovereign of all England!
    [Pause]
SOLDIER: Get away!
                  "Monty Python and the Holy Grail" PYTHON (MONTY) PICTURES LTD

  /// Bram Moolenaar -- Bram@... -- http://www.Moolenaar.net   \\\
///        sponsor Vim, vote for features -- http://www.Vim.org/sponsor/ \\\
\\\        download, build and distribute -- http://www.A-A-P.org        ///
  \\\            help me help AIDS victims -- http://www.ICCF.nl         ///

Dnia niedziela, 2 kwietnia 2006 13:27, Tobias Klein napisał:
> Hi list,
>
> vim exists with a floating point exception when i do (in bash):
>
> echo "# vim:lines=1000" > test; vim test
>
> tested with vim 7.0b beta on Mandrake 10.2.
> Is this the right behaviour?

Confirming with 7.0c03
Caught FPE
exiting with floating point error

GDB:
Program received signal SIGFPE. Arithmetic exception.
[Switching to Thread -1219458848 (LWP 20081)]
0x08166e5f in syntax_start (wp=0x81e4cf0, lnum=1) at syntax.c:554
554         dist = syn_buf->b_ml.ml_line_count / (syn_buf->b_sst_len - Ros)
+ 1;


m.

On RedHat & Fedota linux, grub.conf is symlinked as follows:
/etc/grub.conf -> ../boot/grub/grub.conf

File filetype.vim has this:
     au BufNewFile,BufRead /boot/grub/menu.lst,/boot/grub/grub.conf     setf grub

As a result, filetype of /etc/grub.conf if not detected properly. Vim
does not set
filetype to 'grub' on file /etc/grub.conf.
Can this be fixed ?

Thanks
Yakov

Yakov Lerner wrote:

> On RedHat & Fedota linux, grub.conf is symlinked as follows:
> /etc/grub.conf -> ../boot/grub/grub.conf
>
> File filetype.vim has this:
>     au BufNewFile,BufRead /boot/grub/menu.lst,/boot/grub/grub.conf     setf=
>  grub
>
> As a result, filetype of /etc/grub.conf if not detected properly. Vim
> does not set
> filetype to 'grub' on file /etc/grub.conf.
> Can this be fixed ?

Sure.

--
The problem with political jokes is that they get elected.

  /// Bram Moolenaar -- Bram@... -- http://www.Moolenaar.net   \\\
///        sponsor Vim, vote for features -- http://www.Vim.org/sponsor/ \\\
\\\        download, build and distribute -- http://www.A-A-P.org        ///
  \\\            help me help AIDS victims -- http://www.ICCF.nl         ///

Mikolaj Machowski wrote:
> Dnia sobota, 1 kwietnia 2006 16:19, Alan Briolat napisał:
>> I am experiencing a bug with the php omni-complete, where if i have
>> multiple buffers open, the omni-complete action (^X^O in insert mode)
>> will give the following errors on anything other than the first buffer
>> opened, and be missing quite a lot of the completion entries that are
>> usually available.
>>
>> -----------------
>> Error detected while processing function phpcomplete#CompletePHP:
>> line  390:
>> E121: Undefined variable: b:php_keywords
>
> Oops. Sorry. Fast fix is to do::
>
>  %s/b:php_keywords/g:php_keywords/g
>
> Sending fixed version to Bram.
>
> m.
>

Aha, many thanks, I'll do that for now then!  And thanks for fixing it :P

On Fri, Mar 31, 2006 at 09:43:08PM +0200, Bram Moolenaar wrote:
> And what if you have the same file name in two or more different
> directories?  E.g., "Makefile".

What about reducing the path to the least part that make possible to
disambiguate file names?

As long as all are not ambiguous you just use the basename. In the
example above something lile ".../foo/Makefile" vs ".../bar/Makefile" (3
dots or something different that helps understanding that the full path
has been truncated).

Cheers.

--
Stefano Zacchiroli -*- Computer Science PhD student @ Uny Bologna, Italy
zack@{cs.unibo.it,debian.org,bononia.it} -%- http://www.bononia.it/zack/
If there's any real truth it's that the entire multidimensional infinity
of the Universe is almost certainly being run by a bunch of maniacs. -!-

Hello,

1.

Syntax of /etc/mtab file is identical to that of fstab. Hence suggestion
for addition in filetype.vim (line 607, vim7c03)::


     " FStab
     au BufNewFile,BufRead fstab,mtab  setf fstab

2.

In Mandriva config file for man has name man.config. filetype.vim
recognizes only man.conf. Suggested change (line 883)::

     " Man config
     au BufNewFile,BufRead /etc/man.conf* setf manconf

m.

Mikolaj Machowski wrote:

> 1.
>
> Syntax of /etc/mtab file is identical to that of fstab. Hence suggestion
> for addition in filetype.vim (line 607, vim7c03)::
>
>
>     " FStab
>     au BufNewFile,BufRead fstab,mtab  setf fstab

OK.

> 2.
>
> In Mandriva config file for man has name man.config. filetype.vim
> recognizes only man.conf. Suggested change (line 883)::
>
>     " Man config
>     au BufNewFile,BufRead /etc/man.conf* setf manconf

A "*" at the end is dangerous, this could pickup man.conf.readme or
whatever.  Adding man.config should be OK.


--
SOLDIER: Where did you get the coconuts?
ARTHUR:  Through ... We found them.
SOLDIER: Found them?  In Mercea.  The coconut's tropical!
                  "Monty Python and the Holy Grail" PYTHON (MONTY) PICTURES LTD

  /// Bram Moolenaar -- Bram@... -- http://www.Moolenaar.net   \\\
///        sponsor Vim, vote for features -- http://www.Vim.org/sponsor/ \\\
\\\        download, build and distribute -- http://www.A-A-P.org        ///
  \\\            help me help AIDS victims -- http://www.ICCF.nl         ///

Vim7.0b, MSWin32 GUI with OLE, big version, CVS 20060327, Win2K pro
(Someone says it recurs in Vim7.0c too.)

_vimrc contains only one line:
au BufRead * Sexplore

"vimfiles" directory has been renamed to avoid influence of plugins.

Open GVIM, then execute ":help" or ":help au" (can be any help topic),
it says "E434: Can't find tag pattern" and crash, Windows reports VIM
read some memory that is not readable.

Liy Yubao wrote:

> Vim7.0b, MSWin32 GUI with OLE, big version, CVS 20060327, Win2K pro
> (Someone says it recurs in Vim7.0c too.)
>
> _vimrc contains only one line:
> au BufRead * Sexplore
>
> "vimfiles" directory has been renamed to avoid influence of plugins.
>
> Open GVIM, then execute ":help" or ":help au" (can be any help topic),
> it says "E434: Can't find tag pattern" and crash, Windows reports VIM
> read some memory that is not readable.

That's a nasty autocommand, splitting the window and doing something
completely different while the user wants to see a help file.  But Vim
shouldn't crash.  I'll fix it.

--
"I've been teaching myself to play the piano for about 5 years and now write
most of my songs on it, mainly because I can never find any paper."
		 Jeff Lynne, ELO's greatest hits

  /// Bram Moolenaar -- Bram@... -- http://www.Moolenaar.net   \\\
///        sponsor Vim, vote for features -- http://www.Vim.org/sponsor/ \\\
\\\        download, build and distribute -- http://www.A-A-P.org        ///
  \\\            help me help AIDS victims -- http://www.ICCF.nl         ///

Hi!

I have seen a surprising (to me) behavior of Vim
today:

1. Define the mapping

onoremap <F3> <C-C>n

2. and do

:set insertmode

You are in Insertmode now

3. Do <C-O>
4. Do d

You are in Operator pending mode now

5. Do <F3>

The result: not the next pattern is searched for, but
'n' is inserted. I consider, that <C-C> of the mapping
switches back to Insert mode.

Is this the normal behavior or shouldn't Vim return to
Insert mode not before the mapping has finished? If
'insertmode' is not set, this is the actual behavior
of Vim when doing the same thing.

Best wishes,
Georg









___________________________________________________________
Yahoo! Messenger - NEW crystal clear PC to PC calling worldwide with voicemail
http://uk.messenger.yahoo.com

Georg Dahn wrote:

> I have seen a surprising (to me) behavior of Vim
> today:
>
> 1. Define the mapping
>
> onoremap <F3> <C-C>n
>
> 2. and do
>
> :set insertmode
>
> You are in Insertmode now
>
> 3. Do <C-O>
> 4. Do d
>
> You are in Operator pending mode now
>
> 5. Do <F3>
>
> The result: not the next pattern is searched for, but
> 'n' is inserted. I consider, that <C-C> of the mapping
> switches back to Insert mode.
>
> Is this the normal behavior or shouldn't Vim return to
> Insert mode not before the mapping has finished? If
> 'insertmode' is not set, this is the actual behavior
> of Vim when doing the same thing.

Well, the CTRL-C aborts the current command.  When 'insertmode' is set
Vim goes back to Insert mode.  I can't say this is wrong.  Setting
'insertmode' changes a lot of things.

Perhaps what should happen is that CTRL-C should forget about the rest
of the mapping.  But that would break existing mappings, thus we can't
do that.

--
CART DRIVER: Bring out your dead!
LARGE MAN:   Here's one!
CART DRIVER: Ninepence.
BODY:        I'm not dead!
                  "Monty Python and the Holy Grail" PYTHON (MONTY) PICTURES LTD

  /// Bram Moolenaar -- Bram@... -- http://www.Moolenaar.net   \\\
///        sponsor Vim, vote for features -- http://www.Vim.org/sponsor/ \\\
\\\        download, build and distribute -- http://www.A-A-P.org        ///
  \\\            help me help AIDS victims -- http://www.ICCF.nl         ///