Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm
List-Subscribe: <mailto:cygwin-subscribe AT cygwin DOT com>
List-Archive: <http://sourceware.org/ml/cygwin/>
List-Post: <mailto:cygwin AT cygwin DOT com>
List-Help: <mailto:cygwin-help AT cygwin DOT com>, <http://sourceware.org/ml/#faqs>
Sender: cygwin-owner AT cygwin DOT com
Mail-Followup-To: cygwin AT cygwin DOT com
Delivered-To: mailing list cygwin AT cygwin DOT com
Date: Sun, 12 Sep 2004 14:22:37 -0400 (EDT)
From: Igor Pechtchanski <pechtcha AT cs DOT nyu DOT edu>
Reply-To: cygwin AT cygwin DOT com
To: cygwin AT cygwin DOT com
Subject: Re: cygserver service won't start
In-Reply-To: <20040912175541.GG18421@trixie.casa.cgf.cx>
Message-ID: <Pine.GSO.4.61.0409121415510.13114@slinky.cs.nyu.edu>
References: <20040909080829 DOT GC2953 AT cygbert DOT vinschen DOT de> <chqtp3$n05$1 AT sea DOT gmane DOT org> <20040910080758 DOT GE468 AT cygbert DOT vinschen DOT de> <chsjeg$9em$1 AT sea DOT gmane DOT org> <chslrm$g4i$1 AT sea DOT gmane DOT org> <20040911123916 DOT GA8906 AT cygbert DOT vinschen DOT de> <chvame$htv$1 AT sea DOT gmane DOT org> <Pine DOT GSO DOT 4 DOT 61 DOT 0409111258440 DOT 16985 AT slinky DOT cs DOT nyu DOT edu> <20040912144634 DOT GE17670 AT cygbert DOT vinschen DOT de> <Pine DOT GSO DOT 4 DOT 61 DOT 0409121129250 DOT 16985 AT slinky DOT cs DOT nyu DOT edu> <20040912175541 DOT GG18421 AT trixie DOT casa DOT cgf DOT cx>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
X-Scanned-By: MIMEDefang 2.39

On Sun, 12 Sep 2004, Christopher Faylor wrote:

> On Sun, Sep 12, 2004 at 11:36:42AM -0400, Igor Pechtchanski wrote:
> >On Sun, 12 Sep 2004, Corinna Vinschen wrote:
> >
> >> On Sep 11 13:06, Igor Pechtchanski wrote:
> >> > On Sat, 11 Sep 2004, Alder wrote:
> >> > > On a perfect platform, perhaps, but like I said, I really had to add
> >> > > it because without it there a Windows dialog would pop up when I ran
> >> > > cygserver telling me that 'Cygwin1.dll' could not be found.  That
> >> > > was also true on my previous installation, and one of the reasons I
> >> > > thought it best to reinstall.
> >>
> >> This is weird.  I never observed that and I installed a lot of services
> >> using cygrunsrv on a good number of machines.
> >>
> >> > I'm observing similar behavior, but it looks like our causes may be
> >> > different.
> >>
> >> Igor, out with your debugging experiences!
> >
> >I believe they, along with a description of the fix, are summarized in
> ><http://cygwin.com/ml/cygwin-developers/2004-09/msg00007.html>, which
> >seems to have been overlooked, as it never got any reply (the rest of the
> >thread deteriorated into a discussion of the need for check_case:strict).
> >I'd like to get some feedback on whether my analysis and the fix plan are
> >correct, before I go through the machinery of getting the patch approved.
>
> Given the number of problems we've had with strict case checking, I
> think it is time to start thinking about ripping it out.

That's your choice, of course, but I already stated my reasons for why I
find this useful.  This particular problem results from strict case
checking being overeager in some cases.  See below.

> Neither Corinna nor I are huge fans of the functionality and while I
> understand that some people like to use it, the amount of extra jumping
> hoops required in the code multiplied by the number of times we hear
> people complaining about problems seems to indicate that this is not a
> robust feature.  Corinna essentially implemented this feature under
> duress.
>
> It's difficult to see without the actual patch but your change seems to
> add more logic to cygwin and adds decision points in parts of the code
> which are called a lot.

Yes, my changes added extra logic, but mostly because I wanted to keep the
observable impact to a minimum.  Upon reflection, it seems like the
particular bit of code that causes this problem was an attempt at
optimization, so that instead of multiple accesses to the same path
failing, the path would simply not be converted at all.  IMO, there's no
need for this optimization at all, and that piece of code should just be
excised altogether.  I'll send a description of my proposed changes (which
will *simplify* the logic) to cygwin-developers.

> Any time I hear about who great this option is, it seems as though
> people raise the issue of finding multiple files with the same name in a
> tar file.  If this really represents the majority of usefulness for this
> feature then I think something like a tar-lint or adding an option to
> tar itself might be a better plan than burdening the cygwin DLL with
> this feature.
>
> cgf

Or CVS, or rsync, or scp, or a whole bunch of other interactions with
case-sensitive operating systems.
	Igor
-- 
				http://cs.nyu.edu/~pechtcha/
      |\      _,,,---,,_		pechtcha AT cs DOT nyu DOT edu
ZZZzz /,`.-'`'    -.  ;-;;,_		igor AT watson DOT ibm DOT com
     |,4-  ) )-,_. ,\ (  `'-'		Igor Pechtchanski, Ph.D.
    '---''(_/--'  `-'\_) fL	a.k.a JaguaR-R-R-r-r-r-.-.-.  Meow!

"Happiness lies in being privileged to work hard for long hours in doing
whatever you think is worth doing."  -- Dr. Jubal Harshaw

--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Problem reports:       http://cygwin.com/problems.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/