remember(-diary).el

remember(-diary).el

[martin rudalics]

When doing a cvs update of the Emacs trunk I currently get:

cvs update: move away lisp/textmodes/remember.el; it is in the way
C lisp/textmodes/remember.el

My CVS version is 1.11.19.  CVS has created a file
textmodes/remember-diary.el along with the CVS entry

/remember-diary.el/1.3/Tue Oct 30 23:28:00 2007//

but _no_ file textmodes/remember.el.  Hence I conclude that this problem
is due to a short filenames clash on my operating system - WindowsME
with a FAT32 file-system.  In fact the 8+3 versions of remember-diary.el
and remember.el both resolve to remember.el.

Re: remember(-diary).el

[Miles Bader]

martin rudalics <rudalics@gmx.at> writes:
> Hence I conclude that this problem is due to a short filenames clash
> on my operating system - WindowsME with a FAT32 file-system.  In fact
> the 8+3 versions of remember-diary.el and remember.el both resolve to
> remember.el.

Please use a real filesystem.

-Miles

-- 
`The suburb is an obsolete and contradictory form of human settlement'

Re: remember(-diary).el

[Glenn Morris]

Miles Bader wrote:

> martin rudalics <rudalics@gmx.at> writes:
>
>> Hence I conclude that this problem is due to a short filenames
>> clash on my operating system - WindowsME with a FAT32 file-system.
>> In fact the 8+3 versions of remember-diary.el and remember.el both
>> resolve to remember.el.
>
> Please use a real filesystem.

As an alternative suggestion, let me say that remember-diary contains
only 2 functions and a defcustom, and could easily be merged into
remember.

Re: remember(-diary).el

[martin rudalics]

> Please use a real filesystem.

Do you mean support for Windows98 shall be discontinued?

Re: remember(-diary).el

[Miles Bader]

martin rudalics <rudalics@gmx.at> writes:
>> Please use a real filesystem.
>
> Do you mean support for Windows98 shall be discontinued?

Windows98 has support for long filenames doesn't it?
[I think even W95 did at some point.]

-Miles

-- 
Fast, small, soon; pick any 2.

Re: remember(-diary).el

[martin rudalics]

 > Windows98 has support for long filenames doesn't it?

Sure.  Maybe my cvs.exe doesn't support them or I should supply
some option.  I'm still searching for an explanation of this.  So
far the 8+3 one is the only I got.

Re: remember(-diary).el

[Sven Joachim]

On 2007-11-21 08:24 +0100, martin rudalics wrote:

> When doing a cvs update of the Emacs trunk I currently get:
>
> cvs update: move away lisp/textmodes/remember.el; it is in the way
> C lisp/textmodes/remember.el
>
> My CVS version is 1.11.19.  CVS has created a file
> textmodes/remember-diary.el along with the CVS entry
>
> /remember-diary.el/1.3/Tue Oct 30 23:28:00 2007//
>
> but _no_ file textmodes/remember.el.  Hence I conclude that this problem
> is due to a short filenames clash on my operating system - WindowsME
> with a FAT32 file-system.  In fact the 8+3 versions of remember-diary.el
> and remember.el both resolve to remember.el.

I suspect you have turned of the generation of numeric tail names in the
Windows Registry, otherwise remember-diary.el would resolve to
rememb~1.el.  Use regedit and search for a key named NameNumericTail.
If such a key exists and has a value of 0, change it to 1 and restart
Windows.  Google will tell you a lot more about the pros and cons of
NameNumericTail.

Also make sure that you do _not_ have a key named Win31FileSystem set to
1, that setting would turn off generation of long file names completely.

Hope this helps,
Sven

Re: remember(-diary).el

[Eli Zaretskii]

> From: Glenn Morris <rgm@gnu.org>
> Date: Wed, 21 Nov 2007 03:39:33 -0500
> Cc: martin rudalics <rudalics@gmx.at>, emacs-devel <emacs-devel@gnu.org>
> 
> As an alternative suggestion, let me say that remember-diary contains
> only 2 functions and a defcustom, and could easily be merged into
> remember.

A good idea, IMO.  John and Sacha, is it okay to do that?

Re: remember(-diary).el

[Eli Zaretskii]

> From: Miles Bader <miles@gnu.org>
> Date: Wed, 21 Nov 2007 22:55:37 +0900
> Cc: emacs-devel <emacs-devel@gnu.org>
> 
> Windows98 has support for long filenames doesn't it?

It does, but it also records the short 8+3 alias, and in this case
they clash.

Re: remember(-diary).el

[Eli Zaretskii]

> From: Sven Joachim <svenjoac@gmx.de>
> Date: Wed, 21 Nov 2007 19:24:33 +0100
> Cc: emacs-devel <emacs-devel@gnu.org>
> 
> I suspect you have turned of the generation of numeric tail names in the
> Windows Registry, otherwise remember-diary.el would resolve to
> rememb~1.el.

That's true, but even if numeric tails are not disabled, you could
overwrite remember-diary.el when you write to (an already existing)
remember.el.

It is simply not a good idea to have such clashes on FAT32 filesystems
under Windows 9x.

Re: remember(-diary).el

[John Wiegley]

On Nov 21, 2007, at 2:49 PM, Eli Zaretskii wrote:

>> From: Glenn Morris <rgm@gnu.org>
>> Date: Wed, 21 Nov 2007 03:39:33 -0500
>> Cc: martin rudalics <rudalics@gmx.at>, emacs-devel <emacs-devel@gnu.org 
>> >
>>
>> As an alternative suggestion, let me say that remember-diary contains
>> only 2 functions and a defcustom, and could easily be merged into
>> remember.
>
> A good idea, IMO.  John and Sacha, is it okay to do that?

I certainly don't mind, but it would be Sacha's call.

John

Re: remember(-diary).el

[Miles Bader]

Eli Zaretskii <eliz@gnu.org> writes:
>> Windows98 has support for long filenames doesn't it?
>
> It does, but it also records the short 8+3 alias, and in this case
> they clash.

I'm not sure what you mean....

As I understand it (it's been a fairly long time since I've used the
more primitive versions of windows), the "short name" chosen for a long
name is explicitly chosen to ensure it doesn't clash with any other
short names.  E.g., for the file "remember-diary.el", it might choose
"rememb~1.el" as a short name.

It sounds like either Martin has support for long names turned off, or
cvs.exe was built in some bad way.

-Miles

-- 
`Cars give people wonderful freedom and increase their opportunities.
 But they also destroy the environment, to an extent so drastic that
 they kill all social life' (from _A Pattern Language_)

Re: remember(-diary).el

[Richard Stallman]

    Do you mean support for Windows98 shall be discontinued?

I don't think we should desupport Windows 98.  Last I checked,
it had many users in the third world.

If this means we must do some renames, let's do them.

    As an alternative suggestion, let me say that remember-diary contains
    only 2 functions and a defcustom, and could easily be merged into
    remember.

In this case, that sounds like a fine solution.

Re: remember(-diary).el

[Eli Zaretskii]

> Cc: rudalics@gmx.at, emacs-devel@gnu.org
> From: Miles Bader <miles.bader@necel.com>
> Date: Thu, 22 Nov 2007 11:13:40 +0900
> 
> As I understand it (it's been a fairly long time since I've used the
> more primitive versions of windows), the "short name" chosen for a long
> name is explicitly chosen to ensure it doesn't clash with any other
> short names.  E.g., for the file "remember-diary.el", it might choose
> "rememb~1.el" as a short name.

In theory, yes, but in practice, the algorithm that chooses the short
8+3 alias has bugs that could well cause a clash, depending on what
other files are present in the directory whose names map to the same
strings after 8+3 truncation, and in what order Windows sees the files
created.

> It sounds like either Martin has support for long names turned off

I sincerely doubt that: no one in their right mind disables the long
file names support.

> cvs.exe was built in some bad way.

I'm not aware of a way one could build a Windows binary so as to
causes such an effect.  Of course, it's possible that I have something
to learn.

Re: remember(-diary).el

[martin rudalics]

 >>As I understand it (it's been a fairly long time since I've used the
 >>more primitive versions of windows), the "short name" chosen for a long
 >>name is explicitly chosen to ensure it doesn't clash with any other
 >>short names.  E.g., for the file "remember-diary.el", it might choose
 >>"rememb~1.el" as a short name.
 >
 > In theory, yes, but in practice, the algorithm that chooses the short
 > 8+3 alias has bugs that could well cause a clash, depending on what
 > other files are present in the directory whose names map to the same
 > strings after 8+3 truncation, and in what order Windows sees the files
 > created.

FWIW Eli's right:

(1) If remember.el is created _before_ remember-diary.el, the 8+3 alias
of the former is written to disk as REMEMBER.EL that of the latter as
REMEMB~1.EL (I did use the disk editor to verify that).

(2) If remember-diary.el is created first, its 8+3 alias is written to
disk as REMEMBER.EL.  When I now try to find a file remember.el in the
same directory, Windows finds remember-diary.el instead.  When I try to
store a file remember.el in that directory, Windows complains that such
a file already exists.

In general the problem occurs when the name of the second file I want to
write has eight characters _and_ is a prefix of the name of the first
file.  Hence, I get the same bug for remember_diary.el vs remember.el as
well as for rememberdiary.el vs remember.el.

With other words: When the name of a file A sans extension is a prefix
of the name of a file B sans extension and the length of the name of
file A sans extension equals 8, file A must be created before file B in
order for having both files coexist in the same directory.  I didn't
bother to check this rule for filenames containing a tilde.

My problem with the CVS repository thus could be resolved by renaming
remember.el to something like remember-core.el.

Re: remember(-diary).el

[joakim]

martin rudalics <rudalics@gmx.at> writes:

Would it be possible for emacs to read its files from some kind of
archive? That way the issue of different filesystems having
substandard support for long filenames wouldnt come up.

To elaborate:
- some systems work this way so there is precedence(javas jar files
for instance)
- the el files are already compressed, so also archiving them in a
archive blob of some form doesnt seem to be so far away
- the archiving system wouldn need to be default, it could be used on
old windoze systems and maybe on embedded systems

This assumes of course that emacs is built on some other os and the
resulting binary delivered to the target system, so maybe the idea
falls apart there, I dont know.

>>>As I understand it (it's been a fairly long time since I've used the
>>>more primitive versions of windows), the "short name" chosen for a long
>>>name is explicitly chosen to ensure it doesn't clash with any other
>>>short names.  E.g., for the file "remember-diary.el", it might choose
>>>"rememb~1.el" as a short name.
>>
>> In theory, yes, but in practice, the algorithm that chooses the short
>> 8+3 alias has bugs that could well cause a clash, depending on what
>> other files are present in the directory whose names map to the same
>> strings after 8+3 truncation, and in what order Windows sees the files
>> created.
>
> FWIW Eli's right:
>
> (1) If remember.el is created _before_ remember-diary.el, the 8+3 alias
> of the former is written to disk as REMEMBER.EL that of the latter as
> REMEMB~1.EL (I did use the disk editor to verify that).
>
> (2) If remember-diary.el is created first, its 8+3 alias is written to
> disk as REMEMBER.EL.  When I now try to find a file remember.el in the
> same directory, Windows finds remember-diary.el instead.  When I try to
> store a file remember.el in that directory, Windows complains that such
> a file already exists.
>
> In general the problem occurs when the name of the second file I want to
> write has eight characters _and_ is a prefix of the name of the first
> file.  Hence, I get the same bug for remember_diary.el vs remember.el as
> well as for rememberdiary.el vs remember.el.
>
> With other words: When the name of a file A sans extension is a prefix
> of the name of a file B sans extension and the length of the name of
> file A sans extension equals 8, file A must be created before file B in
> order for having both files coexist in the same directory.  I didn't
> bother to check this rule for filenames containing a tilde.
>
> My problem with the CVS repository thus could be resolved by renaming
> remember.el to something like remember-core.el.

-- 
Joakim Verona

Re: remember(-diary).el

[Leo]

On 2007-11-22 08:57 +0000, martin rudalics wrote:
>>>As I understand it (it's been a fairly long time since I've used the
>>>more primitive versions of windows), the "short name" chosen for a long
>>>name is explicitly chosen to ensure it doesn't clash with any other
>
>>>short names.  E.g., for the file "remember-diary.el", it might choose
>>>"rememb~1.el" as a short name.
>>
>> In theory, yes, but in practice, the algorithm that chooses the short
>> 8+3 alias has bugs that could well cause a clash, depending on what
>> other files are present in the directory whose names map to the same
>> strings after 8+3 truncation, and in what order Windows sees the files
>> created.
>
> FWIW Eli's right:
>
> (1) If remember.el is created _before_ remember-diary.el, the 8+3 alias
> of the former is written to disk as REMEMBER.EL that of the latter as
> REMEMB~1.EL (I did use the disk editor to verify that).
>
> (2) If remember-diary.el is created first, its 8+3 alias is written to
> disk as REMEMBER.EL.  When I now try to find a file remember.el in the
> same directory, Windows finds remember-diary.el instead.  When I try to
> store a file remember.el in that directory, Windows complains that such
> a file already exists.
>
> In general the problem occurs when the name of the second file I want to
> write has eight characters _and_ is a prefix of the name of the first
> file.  Hence, I get the same bug for remember_diary.el vs remember.el as
> well as for rememberdiary.el vs remember.el.
>
> With other words: When the name of a file A sans extension is a prefix
> of the name of a file B sans extension and the length of the name of
> file A sans extension equals 8, file A must be created before file B in
> order for having both files coexist in the same directory.  I didn't
> bother to check this rule for filenames containing a tilde.
>
> My problem with the CVS repository thus could be resolved by renaming
> remember.el to something like remember-core.el.

remember-diary probably should be included with planner (an external
package) instead of in Emacs. AFAIK, it has not been used in any
packages except planner.

-- 
.:  Leo  :.  [ sdl.web AT gmail.com ]  .:  [ GPG Key: 9283AA3F ]  :.

          Use the best OS -- http://www.fedoraproject.org/

Re: remember(-diary).el

[martin rudalics]

 > I suspect you have turned of the generation of numeric tail names in the
 > Windows Registry, otherwise remember-diary.el would resolve to
 > rememb~1.el.  Use regedit and search for a key named NameNumericTail.
 > If such a key exists and has a value of 0, change it to 1 and restart
 > Windows.

Applying the suggested change resolved the problem.

 > Hope this helps,
 > Sven

Indeed this helped.  Thank you very much.

Re: remember(-diary).el

[Juri Linkov]

>> I suspect you have turned of the generation of numeric tail names in the
>> Windows Registry, otherwise remember-diary.el would resolve to
>> rememb~1.el.  Use regedit and search for a key named NameNumericTail.
>> If such a key exists and has a value of 0, change it to 1 and restart
>> Windows.
>
> Applying the suggested change resolved the problem.

Maybe this problem should be described in etc/PROBLEMS?

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

Re: remember(-diary).el

[Juri Linkov]

> remember-diary probably should be included with planner (an external
> package) instead of in Emacs. AFAIK, it has not been used in any
> packages except planner.

Do you mean adding a new file remember-planner.el to Emacs?

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

Re: remember(-diary).el

[Leo]

On 2007-11-22 23:57 +0000, Juri Linkov wrote:
>> remember-diary probably should be included with planner (an external
>> package) instead of in Emacs. AFAIK, it has not been used in any
>> packages except planner.
>
> Do you mean adding a new file remember-planner.el to Emacs?

I mean remember-diary belongs to the external package planner-el. It
should be removed from Emacs for now.

-- 
.:  Leo  :.  [ sdl.web AT gmail.com ]  .:  [ GPG Key: 9283AA3F ]  :.

          Use the best OS -- http://www.fedoraproject.org/

Re: remember(-diary).el

[Richard Stallman]

    Would it be possible for emacs to read its files from some kind of
    archive? That way the issue of different filesystems having
    substandard support for long filenames wouldnt come up.

I don't want to bother with new features to support semi-obsolete
systems.  Renaming files is less work.

Re: remember(-diary).el

[Stephen J. Turnbull]

Richard Stallman writes:

 >     Would it be possible for emacs to read its files from some kind of
 >     archive? That way the issue of different filesystems having
 >     substandard support for long filenames wouldnt come up.
 > 
 > I don't want to bother with new features to support semi-obsolete
 > systems.  Renaming files is less work.

While there's probably no hurry to implement the feature for Emacs,
the idea of reading files from archives rather than on-disk file
systems has many applications.

For example, Python and Java have the concept of "eggs" and "jar
files" respectively, which simplify the distribution of 3rd-party
libraries, both by removing the distinction between the distribution
medium and the installed package and by making the directory listing
of the installed packages cleaner.  These are basically just zip
archives.  Also, most free software OSes today provide file systems
which allow mounting archives as file systems, as well as more
primitive loopback file systems that emulate an on-disk file system's
layout in a file.

The advantage of implementing at the application level is that
directory access doesn't require a system call.

Re: remember(-diary).el

[David Kastrup]

"Stephen J. Turnbull" <turnbull@sk.tsukuba.ac.jp> writes:

> Also, most free software OSes today provide file systems which allow
> mounting archives as file systems, as well as more primitive loopback
> file systems that emulate an on-disk file system's layout in a file.
>
> The advantage of implementing at the application level is that
> directory access doesn't require a system call.

Why not call the system for a task it is optimized thoroughly for?

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: remember(-diary).el

[Michael Olson]

[-- Attachment #1.1: Type: text/plain, Size: 797 bytes --]

Leo <sdl.web@gmail.com> writes:

> On 2007-11-22 23:57 +0000, Juri Linkov wrote:
>>> remember-diary probably should be included with planner (an external
>>> package) instead of in Emacs. AFAIK, it has not been used in any
>>> packages except planner.
>>
>> Do you mean adding a new file remember-planner.el to Emacs?
>
> I mean remember-diary belongs to the external package planner-el. It
> should be removed from Emacs for now.

It in no way belongs to planner-el.  remember-diary is distinct, and it
belongs in Emacs.

-- 
       Michael Olson -- FSF Associate Member #652     |
 http://mwolson.org/ -- Jabber: mwolson_at_hcoop.net  |  /` |\ | | |
            Sysadmin -- Hobbies: Lisp, GP2X, HCoop    | |_] | \| |_|
Projects: Emacs, Muse, ERC, EMMS, ErBot, DVC, Planner |

[-- Attachment #1.2: Type: application/pgp-signature, Size: 188 bytes --]

[-- Attachment #2: Type: text/plain, Size: 142 bytes --]

_______________________________________________
Emacs-devel mailing list
Emacs-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/emacs-devel

Re: remember(-diary).el

[Michael Olson]

[-- Attachment #1.1: Type: text/plain, Size: 980 bytes --]

Richard Stallman <rms@gnu.org> writes:

>     Do you mean support for Windows98 shall be discontinued?
>
> I don't think we should desupport Windows 98.  Last I checked,
> it had many users in the third world.
>
> If this means we must do some renames, let's do them.

The problem was resolved by changing a setting in the Windows registry,
so I don't think we need to rename the file.

>     As an alternative suggestion, let me say that remember-diary
>     contains only 2 functions and a defcustom, and could easily be
>     merged into remember.
>
> In this case, that sounds like a fine solution.

I would strongly prefer for the code to remain in a separate file, for
the sake of modularity.

-- 
       Michael Olson -- FSF Associate Member #652     |
 http://mwolson.org/ -- Jabber: mwolson_at_hcoop.net  |  /` |\ | | |
            Sysadmin -- Hobbies: Lisp, GP2X, HCoop    | |_] | \| |_|
Projects: Emacs, Muse, ERC, EMMS, ErBot, DVC, Planner |

[-- Attachment #1.2: Type: application/pgp-signature, Size: 188 bytes --]

[-- Attachment #2: Type: text/plain, Size: 142 bytes --]

_______________________________________________
Emacs-devel mailing list
Emacs-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/emacs-devel

Re: remember(-diary).el

[Leo]

On 2007-11-30 06:17 +0000, Michael Olson wrote:
> Leo <sdl.web@gmail.com> writes:
>
>> On 2007-11-22 23:57 +0000, Juri Linkov wrote:
>>>> remember-diary probably should be included with planner (an external
>>>> package) instead of in Emacs. AFAIK, it has not been used in any
>>>> packages except planner.
>>>
>>> Do you mean adding a new file remember-planner.el to Emacs?
>>
>> I mean remember-diary belongs to the external package planner-el. It
>> should be removed from Emacs for now.
>
> It in no way belongs to planner-el.  remember-diary is distinct, and it
> belongs in Emacs.

Do you have an example that has the form "DIARY: 2003.08.12 Sacha's
birthday" that is not in planner files?

AFAIK, remember-diary recognize a specific format that is not of much
use elsewhere. It is not too late to include this file when need be.

-- 
.:  Leo  :.  [ sdl.web AT gmail.com ]  .:  [ GPG Key: 9283AA3F ]  :.

          Use the best OS -- http://www.fedoraproject.org/

Re: remember(-diary).el

[Michael Olson]

[-- Attachment #1.1: Type: text/plain, Size: 763 bytes --]

Leo <sdl.web@gmail.com> writes:

> Do you have an example that has the form "DIARY: 2003.08.12 Sacha's
> birthday" that is not in planner files?
>
> AFAIK, remember-diary recognize a specific format that is not of much
> use elsewhere. It is not too late to include this file when need be.

OK, it looks like the examples in the header need rewriting, and
possibly the date-munging algorithm in `remember-diary-convert-entry'
needs generalization.  I've added this to my weekend to-do list.

-- 
       Michael Olson -- FSF Associate Member #652     |
 http://mwolson.org/ -- Jabber: mwolson_at_hcoop.net  |  /` |\ | | |
            Sysadmin -- Hobbies: Lisp, GP2X, HCoop    | |_] | \| |_|
Projects: Emacs, Muse, ERC, EMMS, ErBot, DVC, Planner |

[-- Attachment #1.2: Type: application/pgp-signature, Size: 188 bytes --]

[-- Attachment #2: Type: text/plain, Size: 142 bytes --]

_______________________________________________
Emacs-devel mailing list
Emacs-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/emacs-devel

Re: remember(-diary).el

[Richard Stallman]

    I would strongly prefer for the code to remain in a separate file, for
    the sake of modularity.

Modularity of design is no reason to keep it a separate file.
A piece of code can still be modular even if it is in the same file
with other code.

Is there any more specific issue of modularity at stake here?

Re: remember(-diary).el

[Michael Olson]

[-- Attachment #1.1: Type: text/plain, Size: 886 bytes --]

Richard Stallman <rms@gnu.org> writes:

>     I would strongly prefer for the code to remain in a separate file,
>     for the sake of modularity.
>
> Modularity of design is no reason to keep it a separate file.
> A piece of code can still be modular even if it is in the same file
> with other code.
>
> Is there any more specific issue of modularity at stake here?

The `remember-diary-extract-entries' function needs `make-diary-entry',
and so remember-diary.el must do (require 'diary-lib).  It would be a
shame to make remember.el require 'diary-lib even for people who have no
intention of using Diary.

-- 
       Michael Olson -- FSF Associate Member #652     |
 http://mwolson.org/ -- Jabber: mwolson_at_hcoop.net  |  /` |\ | | |
            Sysadmin -- Hobbies: Lisp, GP2X, HCoop    | |_] | \| |_|
Projects: Emacs, Muse, ERC, EMMS, ErBot, DVC, Planner |

[-- Attachment #1.2: Type: application/pgp-signature, Size: 188 bytes --]

[-- Attachment #2: Type: text/plain, Size: 142 bytes --]

_______________________________________________
Emacs-devel mailing list
Emacs-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/emacs-devel

Re: remember(-diary).el

[Glenn Morris]

Michael Olson wrote:

> The `remember-diary-extract-entries' function needs `make-diary-entry',
> and so remember-diary.el must do (require 'diary-lib).  It would be a
> shame to make remember.el require 'diary-lib even for people who have no
> intention of using Diary.

-(require 'diary-lib)
+(autoload 'make-diary-entry "diary-lib")

Re: remember(-diary).el

[Eli Zaretskii]

> From: Michael Olson <mwolson@gnu.org>
> Date: Fri, 30 Nov 2007 01:22:06 -0500
> 
> Richard Stallman <rms@gnu.org> writes:
> 
> >     Do you mean support for Windows98 shall be discontinued?
> >
> > I don't think we should desupport Windows 98.  Last I checked,
> > it had many users in the third world.
> >
> > If this means we must do some renames, let's do them.
> 
> The problem was resolved by changing a setting in the Windows registry,
> so I don't think we need to rename the file.

We cannot ask Emacs users on Windows 9x to enforce this setting.  Back
when I was using Windows 9x I also disabled numeric tails, because
there are very good reasons to do that (it's a long story, and I no
longer remember all the details, but one benefit is that programs
continue to work if you switch to plain DOS).

> >     As an alternative suggestion, let me say that remember-diary
> >     contains only 2 functions and a defcustom, and could easily be
> >     merged into remember.
> >
> > In this case, that sounds like a fine solution.
> 
> I would strongly prefer for the code to remain in a separate file, for
> the sake of modularity.

Then let's rename the file.

Re: remember(-diary).el

[Michael Olson]

[-- Attachment #1.1: Type: text/plain, Size: 866 bytes --]

Eli Zaretskii <eliz@gnu.org> writes:

>> I would strongly prefer for the code to remain in a separate file,
>> for the sake of modularity.
>
> Then let's rename the file.

If I am forced to choose between renaming and merging, it would be
better to merge, because it is more irritating to the end user (and we
have quite a few users of Remember before its inclusion to Emacs) to
hunt for a new name, as opposed to just removing their
 (require 'remember-diary)
line.

I loathe having to do either of these, and wish that we would find a
more elegant fix in the load or require mechanisms.

-- 
       Michael Olson -- FSF Associate Member #652     |
 http://mwolson.org/ -- Jabber: mwolson_at_hcoop.net  |  /` |\ | | |
            Sysadmin -- Hobbies: Lisp, GP2X, HCoop    | |_] | \| |_|
Projects: Emacs, Muse, ERC, EMMS, ErBot, DVC, Planner |

[-- Attachment #1.2: Type: application/pgp-signature, Size: 188 bytes --]

[-- Attachment #2: Type: text/plain, Size: 142 bytes --]

_______________________________________________
Emacs-devel mailing list
Emacs-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/emacs-devel

Re: remember(-diary).el

[Michael Olson]

[-- Attachment #1.1: Type: text/plain, Size: 1053 bytes --]

Glenn Morris <rgm@gnu.org> writes:

> Michael Olson wrote:
>
>> The `remember-diary-extract-entries' function needs
>> `make-diary-entry', and so remember-diary.el must do (require
>> 'diary-lib).  It would be a shame to make remember.el require
>> 'diary-lib even for people who have no intention of using Diary.
>
> -(require 'diary-lib)
> +(autoload 'make-diary-entry "diary-lib")

I suppose that wouldn't be too horrible.  *sigh*.  I'll do the merge
this weekend.

I do hope that future Emacs packages will be allowed to advertise
themselves as incompatible with Windows 98 rather than submit to this
dreadful mangling of the namespace.  Is this information about the
underlying filesystem available in some way to Emacs Lisp programs?  If
not, I'd like for it to be.

-- 
       Michael Olson -- FSF Associate Member #652     |
 http://mwolson.org/ -- Jabber: mwolson_at_hcoop.net  |  /` |\ | | |
            Sysadmin -- Hobbies: Lisp, GP2X, HCoop    | |_] | \| |_|
Projects: Emacs, Muse, ERC, EMMS, ErBot, DVC, Planner |

[-- Attachment #1.2: Type: application/pgp-signature, Size: 188 bytes --]

[-- Attachment #2: Type: text/plain, Size: 142 bytes --]

_______________________________________________
Emacs-devel mailing list
Emacs-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/emacs-devel

Re: remember(-diary).el

[Leo]

On 2007-11-30 22:11 +0000, Michael Olson wrote:
> Glenn Morris <rgm@gnu.org> writes:
>
>> Michael Olson wrote:
>>
>>> The `remember-diary-extract-entries' function needs
>>> `make-diary-entry', and so remember-diary.el must do (require
>>> 'diary-lib).  It would be a shame to make remember.el require
>>> 'diary-lib even for people who have no intention of using Diary.
>>
>> -(require 'diary-lib)
>> +(autoload 'make-diary-entry "diary-lib")
>
> I suppose that wouldn't be too horrible.  *sigh*.  I'll do the merge
> this weekend.
>
> I do hope that future Emacs packages will be allowed to advertise
> themselves as incompatible with Windows 98 rather than submit to this
> dreadful mangling of the namespace.  Is this information about the
> underlying filesystem available in some way to Emacs Lisp programs?  If
> not, I'd like for it to be.

Agree.

Maybe many third world people still use win 9x; but I just did a survey
and +90% don't intend to use any GNU software and and among the rest, 7%
do not want to upgrade their software.

HTH,
-- 
.:  Leo  :.  [ sdl.web AT gmail.com ]  .:  [ GPG Key: 9283AA3F ]  :.

          Use the best OS -- http://www.fedoraproject.org/

Re: remember(-diary).el

[Eli Zaretskii]

> From: Michael Olson <mwolson@gnu.org>
> Date: Fri, 30 Nov 2007 17:11:40 -0500
> Cc: rms@gnu.org, emacs-devel@gnu.org
> 
> I do hope that future Emacs packages will be allowed to advertise
> themselves as incompatible with Windows 98 rather than submit to this
> dreadful mangling of the namespace.

IMO, that's not possible at this time: lots of machines in the 3rd
world use Windows 9x, so limiting their support by Emacs would be not
nice on our part, especially if the reasons are aesthetic.  Note that
Emacs already fails to support some of the features that cannot be
supported on Windows 9x for technical reasons, like hard links,
Unicode-based clipboard operations, etc.

> Is this information about the underlying filesystem available in
> some way to Emacs Lisp programs?

We could make it available quite easily, but what do you suggest the
application code does with this information?

Re: remember(-diary).el

[Eli Zaretskii]

> From: Leo <sdl.web@gmail.com>
> Date: Fri, 30 Nov 2007 22:22:43 +0000
> 
> Maybe many third world people still use win 9x; but I just did a survey
> and +90% don't intend to use any GNU software and and among the rest, 7%
> do not want to upgrade their software.

But we _do_ want to convince them to use GNU software, don't we?

Re: remember(-diary).el

[David Kastrup]

Eli Zaretskii <eliz@gnu.org> writes:

>> From: Leo <sdl.web@gmail.com>
>> Date: Fri, 30 Nov 2007 22:22:43 +0000
>> 
>> Maybe many third world people still use win 9x; but I just did a
>> survey and +90% don't intend to use any GNU software and and among
>> the rest, 7% do not want to upgrade their software.
>
> But we _do_ want to convince them to use GNU software, don't we?

And the point of free software is that you are _free_ to use it, not
that you _have_ to use it.  At any given point of time, for the "normal
citizen" restrictive government regulations are not relevant.  But that
is a consolation only as long as you are not affected.

And GNU/Linux with its much more modest system requirements offers a
migration path (a path you can start with Emacs) to uptodate software
for older hardware, the kind of hardware people in the third world use.

Incidentally, if 10% intend to use GNU software, I would guess that is
at least double the percentage of people in the first and second world.
Even if the amount would have been more like 0.5%, for the sake of those
it would have been worth it.  10% already is a mass phenomenon, if you
ask me...

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: remember(-diary).el

[Richard Stallman]

    The `remember-diary-extract-entries' function needs `make-diary-entry',
    and so remember-diary.el must do (require 'diary-lib).

`remember-diary-extract-entries' could do (require 'diary-lib).

Re: remember(-diary).el

[Michael Olson]

[-- Attachment #1.1: Type: text/plain, Size: 754 bytes --]

Eli Zaretskii <eliz@gnu.org> writes:

>> Is this information about the underlying filesystem available in
>> some way to Emacs Lisp programs?
>
> We could make it available quite easily, but what do you suggest the
> application code does with this information?

I want to be able to pop up a warning window upon loading a file (this
window can be disabled by the user) which tells people of the issue, and
which registry key to change in order for it not to be a problem
anymore.

-- 
       Michael Olson -- FSF Associate Member #652     |
 http://mwolson.org/ -- Jabber: mwolson_at_hcoop.net  |  /` |\ | | |
            Sysadmin -- Hobbies: Lisp, GP2X, HCoop    | |_] | \| |_|
Projects: Emacs, Muse, ERC, EMMS, ErBot, DVC, Planner |

[-- Attachment #1.2: Type: application/pgp-signature, Size: 188 bytes --]

[-- Attachment #2: Type: text/plain, Size: 142 bytes --]

_______________________________________________
Emacs-devel mailing list
Emacs-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/emacs-devel

Re: remember(-diary).el

[Michael Olson]

[-- Attachment #1.1: Type: text/plain, Size: 1037 bytes --]

Eli Zaretskii <eliz@gnu.org> writes:

>> From: Michael Olson <mwolson@gnu.org>
>> Is this information about the underlying filesystem available in
>> some way to Emacs Lisp programs?
>
> We could make it available quite easily, but what do you suggest the
> application code does with this information?

An additional use case for this code is to detect whether the underlying
file system supports file modification times.  I'm not entirely certain
of this, but I think that at least some variants of FAT, perhaps all, do
not support them.  One of the programs I maintain (not part of Emacs
yet) relies on file modification times in order to accurately determine
which files need to have their contents updated after a corresponding
change to a "source" file.

-- 
       Michael Olson -- FSF Associate Member #652     |
 http://mwolson.org/ -- Jabber: mwolson_at_hcoop.net  |  /` |\ | | |
            Sysadmin -- Hobbies: Lisp, GP2X, HCoop    | |_] | \| |_|
Projects: Emacs, Muse, ERC, EMMS, ErBot, DVC, Planner |

[-- Attachment #1.2: Type: application/pgp-signature, Size: 188 bytes --]

[-- Attachment #2: Type: text/plain, Size: 142 bytes --]

_______________________________________________
Emacs-devel mailing list
Emacs-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/emacs-devel

Re: remember(-diary).el

[Eli Zaretskii]

> From: Michael Olson <mwolson@gnu.org>
> Cc: emacs-devel@gnu.org
> Date: Sun, 02 Dec 2007 16:33:05 -0500
> 
> >> Is this information about the underlying filesystem available in
> >> some way to Emacs Lisp programs?
> >
> > We could make it available quite easily, but what do you suggest the
> > application code does with this information?
> 
> I want to be able to pop up a warning window upon loading a file (this
> window can be disabled by the user) which tells people of the issue, and
> which registry key to change in order for it not to be a problem
> anymore.

That's be harassing the user, a very un-Emacsy thing to do, IMO.

Re: remember(-diary).el

[Eli Zaretskii]

> From: Michael Olson <mwolson@gnu.org>
> Cc: emacs-devel@gnu.org
> Date: Sun, 02 Dec 2007 20:05:23 -0500
> 
> An additional use case for this code is to detect whether the underlying
> file system supports file modification times.

That is unrelated to the numeric tails, even if it were a problem (see
below).

> I'm not entirely certain
> of this, but I think that at least some variants of FAT, perhaps all, do
> not support them.  One of the programs I maintain (not part of Emacs
> yet) relies on file modification times in order to accurately determine
> which files need to have their contents updated after a corresponding
> change to a "source" file.

I'm not aware of any filesystem that does not support file
modification time.  Even plain DOS supports them, and Windows 9x as
well, of course.  Maybe you mean creation and access times?  Those are
supported by Windows 9x VFAT filesystem, though (but not by DOS).

Re: remember(-diary).el

[Stefan Monnier]

> I'm not aware of any filesystem that does not support file
> modification time.  Even plain DOS supports them, and Windows 9x as
> well, of course.  Maybe you mean creation and access times?

There's no "creation time" under POSIX.
The `ctime' is often confused for it, but it's actually different.


        Stefan

Re: remember(-diary).el

[Michael Olson]

[-- Attachment #1.1: Type: text/plain, Size: 1302 bytes --]

Eli Zaretskii <eliz@gnu.org> writes:

>> From: Michael Olson <mwolson@gnu.org>
>> I want to be able to pop up a warning window upon loading a file
>> (this window can be disabled by the user) which tells people of the
>> issue, and which registry key to change in order for it not to be a
>> problem anymore.
>
> That's be harassing the user, a very un-Emacsy thing to do, IMO.

I don't think of it as harassing the user -- if something has
nondeterministic behavior whenever its source code gets checked out, I
want to warn people about it.  Obviously the message that appears would
need to be invective-free and informative.  And since it appears during
load, it needs to contain instructions for how to disable the display of
the message.

As for un-Emacsy, perhaps.  Nonetheless, since the particular program
that I have in mind writes a lot of files, which each produce a message,
it is the case that warning windows are the best way to indicate that an
error has occurred, or that an error may be likely to occur.

-- 
       Michael Olson -- FSF Associate Member #652     |
 http://mwolson.org/ -- Jabber: mwolson_at_hcoop.net  |  /` |\ | | |
            Sysadmin -- Hobbies: Lisp, GP2X, HCoop    | |_] | \| |_|
Projects: Emacs, Muse, ERC, EMMS, ErBot, DVC, Planner |

[-- Attachment #1.2: Type: application/pgp-signature, Size: 188 bytes --]

[-- Attachment #2: Type: text/plain, Size: 142 bytes --]

_______________________________________________
Emacs-devel mailing list
Emacs-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/emacs-devel

Re: remember(-diary).el

[Lennart Borgman (gmail)]

Michael Olson wrote:
> Eli Zaretskii <eliz@gnu.org> writes:
> 
>>> From: Michael Olson <mwolson@gnu.org>
>>> I want to be able to pop up a warning window upon loading a file
>>> (this window can be disabled by the user) which tells people of the
>>> issue, and which registry key to change in order for it not to be a
>>> problem anymore.
>> That's be harassing the user, a very un-Emacsy thing to do, IMO.
> 
> I don't think of it as harassing the user -- if something has
> nondeterministic behavior whenever its source code gets checked out, I
> want to warn people about it.  Obviously the message that appears would
> need to be invective-free and informative.  And since it appears during
> load, it needs to contain instructions for how to disable the display of
> the message.


One may feel differently about this, but for me it sounds like the right 
thing to do in a case like this. It is however very important that the 
message is (very) easy to turn off - and that the information in the 
message is easy to find again.

Actually I believe that this way of handling the problem is what 
computer users expects today. However I also agree with Eli that it 
could be harassing the user. It must be used with caution.

I have tried to avoid this as much as possible (though I felt a bit 
differently in the beginning). In one important case I use it now. It is 
for a package where you do not get the full benefit of it unless you 
turn on some things that may influence Emacs outside of that package 
too. I then inform the user of that this way, using display-buffer to 
show the message and from there the user can turn on those things or 
just choose not to see the message any more.

Re: remember(-diary).el

[Eli Zaretskii]

> From: Stefan Monnier <monnier@iro.umontreal.ca>
> Cc: Michael Olson <mwolson@gnu.org>,  emacs-devel@gnu.org
> Date: Mon, 03 Dec 2007 09:40:13 -0500
> 
> > I'm not aware of any filesystem that does not support file
> > modification time.  Even plain DOS supports them, and Windows 9x as
> > well, of course.  Maybe you mean creation and access times?
> 
> There's no "creation time" under POSIX.

I was talking about Windows and DOS, which are not Posix platforms.

Re: remember(-diary).el

[Michael Olson]

[-- Attachment #1.1: Type: text/plain, Size: 682 bytes --]

Richard Stallman <rms@gnu.org> writes:

> If this means we must do some renames, let's do them.
>
>     As an alternative suggestion, let me say that remember-diary
>     contains only 2 functions and a defcustom, and could easily be
>     merged into remember.
>
> In this case, that sounds like a fine solution.

I have now merged remember-diary.el into remember.el and have updated
its manual accordingly.

-- 
       Michael Olson -- FSF Associate Member #652     |
 http://mwolson.org/ -- Jabber: mwolson_at_hcoop.net  |  /` |\ | | |
          Programmer -- Hobbies: Lisp, HCoop          | |_] | \| |_|
Projects: Emacs, Muse, ERC, EMMS, ErBot, DVC, Planner |

[-- Attachment #1.2: Type: application/pgp-signature, Size: 188 bytes --]

[-- Attachment #2: Type: text/plain, Size: 142 bytes --]

_______________________________________________
Emacs-devel mailing list
Emacs-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/emacs-devel

Re: remember(-diary).el

[Eli Zaretskii]

> From: Michael Olson <mwolson@gnu.org>
> Date: Tue, 04 Dec 2007 20:22:32 -0500
> Cc: emacs-devel@gnu.org
> 
> Richard Stallman <rms@gnu.org> writes:
> 
> > If this means we must do some renames, let's do them.
> >
> >     As an alternative suggestion, let me say that remember-diary
> >     contains only 2 functions and a defcustom, and could easily be
> >     merged into remember.
> >
> > In this case, that sounds like a fine solution.
> 
> I have now merged remember-diary.el into remember.el and have updated
> its manual accordingly.

Thank you.