Saturday, June 30, 2012

Re: New build of official Vim?

On 29/06/12 17:33, Ben Fritz wrote:
> On Friday, June 29, 2012 9:15:33 AM UTC-5, Robert wrote:
>> The current version at vim.org for Windows is 7.3_46. If I look at the Cream download it is at 7.3_556. So what determines if a newer build of the official gets put out there?
>>
>> Note: I rely on the installer, I don't build Vim myself.
>>
>> I am just curious really.
>>
>> Robert
>
> Normally Bram only releases a new installer with a new minor version, not for new patches. The 7.3.46 was an oddity, I think there were some big windows bugs fixed somewhere in those first 46 patches. The next "official" installer will most likely be 7.4.0 or 8.0.0.
>

Yes, and one of the reasons Bram doesn't feel the need to compile Vim
executables very often is that Steve Hall very regularly updates his
"Vim without Cream" distribution for Windows, Björn Winckler regularly
releases MacVim for the Macintosh (and maintains "snapshots" for MacVim
beta-testers), and (I suppose) Linux or other non-Mac "Unix-like" users
are more "technical" and not afraid of compiling Vim from source, or
those who are can get "reasonably recent" versions (though maybe still a
few months old) from whatever distribution they are using. (For example,
the latest gvim on the openSUSE Linux "stable" 12.1 release that I'm
using is a 7.3.322 about half a year old. I prefer compiling my own.) :-)

With the new patches that came out yesterday, the latest patchlevel of
Vim is 7.3.582. OTOH the first patch not in Steve's "Vim without Cream",
i.e. 7.3.557, is only about 10 days old, so I think you can confidently
use that unless you experience one of the bugs fixed by a later patch,
as shown near the end of http://ftp.vim.org/pub/vim/patches/7.3/README
I'm copying them here for your convenience:

> 2958 7.3.557 crash when an autocommand wipes out a buffer when it is hidden
> 2956 7.3.558 (after 7.3.552) memory access error
> 3483 7.3.559 home_replace() does not work with 8.3 filename
> 1551 7.3.560 get an error for a locked argument in extend()
> 1511 7.3.561 refresh: always in a complete function breaks the "." command
> 1659 7.3.562 ":profdel" works when the +profile feature is disabled
> 2742 7.3.563 (after 7.3.557) can't build with tiny features
> 1785 7.3.564 (after 7.3.559) warning for pointer conversion
> 1806 7.3.565 can't generate proto file for Python 3
> 2363 7.3.566 (after 7.3.561) redo works incorrectly without refresh:always
> 1739 7.3.567 missing copyright notice
> 3890 7.3.568 bad indents for #ifdefs
> 133265 7.3.569 evaluating Vim expression in Python is insufficient
> 4659 7.3.570 ":vimgrep" does not obey 'wildignore'
> 3915 7.3.571 duplicated condition
> 1915 7.3.572 duplicate statement in if and else
> 1419 7.3.573 using array index before bounds checking
> 2491 7.3.574 a CTRL-L character is not pasted on the search command line
> 1586 7.3.575 "ygt" tries to yank instead of giving an error
> 7301 7.3.576 formatting of lists inside comments is not right yet
> 6542 7.3.577 size of memory does not fit in 32 bit unsigned
> 2025 7.3.578 misplaced declaration.
> 7644 7.3.579 (after 7.3.569) can't compile with Python 2.5
> 1517 7.3.580 warning on 64 bit MS-Windows
> 4236 7.3.581 problems compiling with Python
> 1342 7.3.582 missing pieces in test OK file

As you can see, Vim is actively updated software. :-)

The reason Steve's builds are labeled "unofficial" is that Bram doesn't
compile them. However they are made from Bram's own sources using a good
C/C++ compiler (the Cygwin/MinGW gcc for "native" Windows, I think).
That compiler (the one I think Steve uses) is not made by Microsoft. I
never had trouble with it when I was still on Windows, and I haven't
noticed complaints about Vim that could be traced to a bug in that compiler.


Best regards,
Tony.
--
FATHER: Did you kill all those guards?
LAUNCELOT: Yes ... I'm very sorry ...
FATHER: They cost fifty pounds each!
"Monty Python and the Holy Grail" PYTHON (MONTY)
PICTURES LTD

--
You received this message from the "vim_use" maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php

No comments: