Re: Release candidate 1: /etc/hosts

[Igor Pechtchanski <pechtcha at cs dot nyu dot edu>]

On Thu, 12 Sep 2002, Nicholas Wourms wrote:

> --- Igor Pechtchanski <pechtcha@cs.nyu.edu> wrote:
> > On Thu, 12 Sep 2002, Paul Johnston wrote:
> >
> > > Hi,
> > > Thanks to Corinna, Joe, Nicholas, Warren and especially Igor, this
> > > script should now be good enough. I've successfully tested it on XP
> > > only.
> >
> > This works on Windows 98 (sort of):
> >
> > BASH-2.05b$ uname -a
> > CYGWIN_98-4.10 FAETON 1.3.12(0.54/3/2) 2002-07-06 02:16 i686
> > unknown
> > BASH-2.05b$ ./make-etc-links.sh
> > create symbolic link `/etc/hosts' to `/cygdrive/c/WINDOWS/hosts'
> > create symbolic link `/etc/protocols' to
> > `/cygdrive/c/WINDOWS/protocol'
> > create symbolic link `/etc/services' to
> > `/cygdrive/c/WINDOWS/services'
> > create symbolic link `/etc/networks' to
> > `/cygdrive/c/WINDOWS/networks'
> > BASH-2.05b$
> >
> > However, two problems:
> >
> > 1) When the script has run, but created a link to a non-existent
> > file, and then run again:
> >
> > BASH-2.05b$ ./make-etc-links.sh
> > /bin/ln: `/etc/protocols': File exists
> > BASH-2.05b$ ls /etc/protocols
> > /etc/protocols
> > BASH-2.05b$ [ ! -e /etc/protocols ]; echo $?
> > 0
> > BASH-2.05b$ [ ! -L /etc/protocols ]; echo $?
> > 1
> > BASH-2.05b$
> >
> > The -e test apparently fails if the file is a symbolic link to a
> > non-existent file (is this a bug?).  I've attached the correction.
> >
> > 2) CYGWIN="check_case:strict"
> > As I suspected earlier, this fails -- the links are created, but an
> > attempt to cat the files results in "no such file or directory",
> > and an attempt to save the file after editing results in a write
> > error.  On my Windows 98, cygwin interprets the filenames for
> > c:\windows\hosts, etc, as all caps.  I don't know how important this
> > is to pursue.
>
> Why not just check for all the possible combinations [HOSTS, Hosts,
> hosts]?  I use check_case:strict on a daily basis.  If you don't use
> it, then you cannot compile gcj java programs at times.  Also, you
> cannot bootstrap gcc with java enabled.  I think this is important
> and heads of questions in the long run.

Ahem, there are 32 possible combinations for hosts, and 256 for networks,
services and protocols (windows only uses the first 8 letters).  I realize
that most of them are improbable, but for the script to be robust, it
should be able to handle any of them.

I also constantly use check_case:strict, which is why I raised the concern
in the first place.  On my 2k system, the following will return the
exact case: 'cmd /c "dir /b "`cygpath -w $file`'...  Can someone verify
that this (with the appropriate correction of "cmd" to "command.com", of
course) also works on 9x/ME systems?
	Igor
-- 
				http://cs.nyu.edu/~pechtcha/
      |\      _,,,---,,_		pechtcha@cs.nyu.edu
ZZZzz /,`.-'`'    -.  ;-;;,_		igor@watson.ibm.com
     |,4-  ) )-,_. ,\ (  `'-'		Igor Pechtchanski
    '---''(_/--'  `-'\_) fL	a.k.a JaguaR-R-R-r-r-r-.-.-.  Meow!

It took the computational power of three Commodore 64s to fly to the moon.
It takes a 486 to run Windows 95.  Something is wrong here. -- SC sig file


--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Bug reporting:         http://cygwin.com/bugs.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/