Emacs 22.2.92 pretest

Emacs 22.2.92 pretest

[Chong Yidong]

Emacs pretest version 22.2.92 is now available.  This is the third
pretest for Emacs 22.3, which will be a bugfix release.

I'd like to urge everyone on emacs-devel (and on the BCC'd pretesters
list) to build, install, and run the pretest.  Those of you who are
running the CVS trunk may find it painful to switch back to
non-antialiased text, but please help us get as much testing done as
possible.

Please send me a private email letting me know what platform you built
on, and whether or not there were problems with the build.  If you
encounter any problems, especially regressions relative to Emacs 22.2,
please email me ASAP.


The source tarball is available here:

http://alpha.gnu.org/gnu/emacs/pretest/emacs-22.2.92.tar.gz

The xdelta against pretest 22.2.91 is here:

http://alpha.gnu.org/gnu/emacs/pretest/emacs-22.2.91-22.2.92.xdelta

The CVS tag is EMACS_PRETEST_22_2_92.

Re: Emacs 22.2.92 pretest

[Jason Rumney]

Chong Yidong wrote:
> Emacs pretest version 22.2.92 is now available.  This is the third
> pretest for Emacs 22.3, which will be a bugfix release.
>   

Precompiled binaries for Windows are available at the usual location:

http://alpha.gnu.org/gnu/emacs/pretest/windows/

If you have a development environment available, please consider Chong's 
advice below and compile from source yourself, to ensure that 
compilation of the pretest is well tested.

> I'd like to urge everyone on emacs-devel (and on the BCC'd pretesters
> list) to build, install, and run the pretest.  Those of you who are
> running the CVS trunk may find it painful to switch back to
> non-antialiased text, but please help us get as much testing done as
> possible.
>
> Please send me a private email letting me know what platform you built
> on, and whether or not there were problems with the build.  If you
> encounter any problems, especially regressions relative to Emacs 22.2,
> please email me ASAP.
>
>
> The source tarball is available here:
>
> http://alpha.gnu.org/gnu/emacs/pretest/emacs-22.2.92.tar.gz
>
> The xdelta against pretest 22.2.91 is here:
>
> http://alpha.gnu.org/gnu/emacs/pretest/emacs-22.2.91-22.2.92.xdelta
>
> The CVS tag is EMACS_PRETEST_22_2_92.
>   

format-time-string (was Emacs 22.2.92 pretest)

[Kazu Yamamoto]

Hello,

First of all, thank you for releaseing Emacs 22.2.92 pretest
and preparing Windows binary!

Unfortunately, Emacs on Windows (not including UNIX and Mac) have a
bug on format-time-string. (format-time-string "%Z") does not return
anything.

(let ((system-time-locale "C"))
  (format-time-string "%Z"))
-> ""

In Japanese environment, this should return "JST".

Some mailreaders on Emacs use this function to produce the Date:
field. (See the header of this message.) So, this bug has a big
impact.

I hope this bug will be fixed before Emacs 22.3 will be released.

--Kazu

RE: format-time-string (was Emacs 22.2.92 pretest)

[Drew Adams]

> First of all, thank you for releaseing Emacs 22.2.92 pretest
> and preparing Windows binary!
> 
> Unfortunately, Emacs on Windows (not including UNIX and Mac) have a
> bug on format-time-string. (format-time-string "%Z") does not return
> anything.
> 
> (let ((system-time-locale "C"))
>   (format-time-string "%Z"))
> -> ""
> 
> In Japanese environment, this should return "JST".
> 
> Some mailreaders on Emacs use this function to produce the Date:
> field. (See the header of this message.) So, this bug has a big
> impact.
> 
> I hope this bug will be fixed before Emacs 22.3 will be released.

This has been reported:

http://emacsbugs.donarmstrong.com/cgi-bin/bugreport.cgi?bug=641

Unfortunately, it was relegated to the wish list.

Re: format-time-string (was Emacs 22.2.92 pretest)

[Jason Rumney]

> In Japanese environment, this should return "JST".
>   

On Windows, that has never been the case unless you explicitly set the 
TZ environment variable to JST.

> Some mailreaders on Emacs use this function to produce the Date:
> field.

Those mailreaders already have a bug if they do that. RFC 822 only 
allowed a small subset of timezone abbreviations (US timezones and GMT, 
plus single character Milatary designations that I've never seen in real 
use), and its successor RFC 2822 makes even these obsolete, requiring 
numeric timezone specifiers in all cases.

Re: format-time-string

[Kazu Yamamoto]

Hello,

> On Windows, that has never been the case unless you explicitly set
> the TZ environment variable to JST.

What I wanted to say is that Emacs 22.2.92 works differently from 
Emacs 22.2.

If this is a spec change, I need to change my mailreader.

> > Some mailreaders on Emacs use this function to produce the Date:
> > field.
> Those mailreaders already have a bug if they do that. RFC 822 only
> allowed a small subset of timezone abbreviations (US timezones and
> GMT, plus single character Milatary designations that I've never
> seen in real use), and its successor RFC 2822 makes even these
> obsolete, requiring numeric timezone specifiers in all cases.

No. Please look at the Date: field above. JST is inserted as
*comment*.  Of course, it is not required. But again, I would point
out that mailreaders which work on Emacs 22.2 does not work on Emacs
22.2.92.

--Kazu

Re: format-time-string

[Kazu Yamamoto]

Hello,

> This has been reported:
> 
> http://emacsbugs.donarmstrong.com/cgi-bin/bugreport.cgi?bug=641

Thank you for your response.

> Unfortunately, it was relegated to the wish list.

So, should I interpret that this is a spec change?

--Kazu

Re: format-time-string

[Jason Rumney]

Kazu Yamamoto (山本和彦) wrote:
> What I wanted to say is that Emacs 22.2.92 works differently from 
> Emacs 22.2.
>   

I'm not aware of any change made since 22.2 that would affect this.

> No. Please look at the Date: field above. JST is inserted as
> *comment*.

Then the mail reader is hardly "not working" if it is not present.

Re: format-time-string

[Kazu Yamamoto]

Hello,

> I'm not aware of any change made since 22.2 that would affect this.

A friend identified the change. It is "2007-06-06".

EMACS_22_BASE src/Changelog
| 2008-03-26  Chong Yidong  <cyd@stupidchicken.com>
|
|         * Version 22.2 released.
- snip -
| 2007-06-06  Jason Rumney  <jasonr@gnu.org>
|
|         * s/ms-w32.h: Don't define HAVE_TZNAME.
- snip -
| 2007-06-02  Chong Yidong  <cyd@stupidchicken.com>
|
|        * Version 22.1 released.

I need to correct my previous statement.
This was introduced between 22.1 and 22.2.

--Kazu