[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Bug#13287: less uses /usr/bin/editor without it necessarily being there.



On 23 Sep 1997, Darren/Torin/Who Ever... wrote:

> -----BEGIN PGP DECRYPTED MESSAGE-----
> Christian Schwarz, in an immanent manifestation of deity, wrote:
> >> If the envar EDITOR isn't set, less uses /usr/bin/editor.  This is
> >> problematic since there are no editor packages yet that register that as
> >> an alternative.  So, possibly it should depend on
> >> /usr/bin/sensible-editor since that's actually in a package and mentions
> >> that you should set $EDITOR.  But the way it is now, actually complies
> >> with the 2.3.0.0 packaging standards.  So, it might be a bug against
> >> debian-policy but not until it's been discussed.  Therefore, I'm cc'ing
> >> this to debian-policy (this could be considered a bug of this message.)
> >
> >I think this is actually a bug of less or debian-policy, it a bug of your
> >preferred editor not providing /usr/bin/editor. Please assign this bug to
> >the editor package.
> 
> I realized that I wasn't clear aft I sent that off.  The problem with
> less is that it doesn't have any dependencies so that something it
> depends on is there.  I'd be rather confused if I came from another unix
> OS where 'v' in less brings up vi and I get '/usr/bin/editor: not
> found'.  I might realize that I need to set $EDITOR but then I might not.
> On the other hand, if less invoked sensible-editor instead of editor, I
> would be told:
> echo -e "Couldn't find an editor!" 2>&1
> echo -e "Set the \$EDITOR environment variable to your desired editor." 2>&1
> 
> Using sensible-editor would also give a concrete dependency to less
> (debianutils (>= 1.6)) rather than a nebulous "your editor should
> provide this."  I can't tell dpkg this...

This discussion is getting too complicated for me :-) I don't know if
things are clear now, or not. So let me try to summarize. 

Feel free to add something or to start a discussion if you disagree.

1. The policy manual 2.3.0.0 contains a section `4.3 Editors and pagers'
which should explain everything.

2. Programs that launch an editor or pager should check for the
environment variables EDITOR/PAGER first. If these variables are not set,
they should fall back to /usr/bin/editor or /usr/bin/pager.

   Since the check for the env vars EDITOR/PAGER can be hard to implement
in some upstream sources, the /usr/bin/sensible-{editor,pager} scripts
have been created. They check for the env vars themselves and fall back to
/usr/bin/{editor,pager}. Thus, you can specify
/usr/bin/sensible-{editor,pager} as editor/pager in your program and the
user always gets a good editor/pager.

3. Packages providing an editor or pager should use "update-alternatives"
to "register" the programs. (The local sysadmin can adapt the links
/usr/bin/editor and pager if he/she prefers a different program.)

4. The base system includes an editor (ae) and a pager (I think less) and
installs the /usr/bin/sensible-{editor,pager} scripts. Thus, we don't need
any virtual packages or dependencies.


Now to the implementation:

1. The base system has to provide initial settings for
/usr/bin/{pager,editor}.

2. Each editor/pager package has to provide the "alternatives."

3. Each application launching an editor or pager has to be fixed to check
the env vars and fall back to /usr/bin/{editor,pager} _or_ use
/usr/bin/sensible-*. 

AFAIK, all three steps have not been implemented yet. I think we should
file bug reports against these packages. (Unfortunately, I don't have time
for this now. So if someone wants to do this... But please drop a note to
debian-policy so that noone else is doing the same and we end up with
several reports :-)


Thanks,

Chris

--                 Christian Schwarz
Do you know         schwarz@monet.m.isar.de, schwarz@schwarz-online.com,
Debian GNU/Linux?    schwarz@debian.org, schwarz@mathematik.tu-muenchen.de
      
Visit                  PGP-fp: 8F 61 EB 6D CF 23 CA D7  34 05 14 5C C8 DC 22 BA
http://www.debian.org   http://fatman.mathematik.tu-muenchen.de/~schwarz/


Reply to: