Mailing-List: contact cygwin-apps-help AT sourceware DOT cygnus DOT com; run by ezmlm
Sender: cygwin-apps-owner AT sourceware DOT cygnus DOT com
List-Subscribe: <mailto:cygwin-apps-subscribe AT sources DOT redhat DOT com>
List-Archive: <http://sources.redhat.com/ml/cygwin-apps/>
List-Post: <mailto:cygwin-apps AT sources DOT redhat DOT com>
List-Help: <mailto:cygwin-apps-help AT sources DOT redhat DOT com>, <http://sources.redhat.com/lists.html#faqs>
Delivered-To: mailing list cygwin-apps AT sources DOT redhat DOT com
X-Originating-IP: [203.29.195.40]
From: "Gareth Pearce" <tilps AT hotmail DOT com>
To: cygwin-apps AT sources DOT redhat DOT com
Subject: Re: [Patch] setup.exe - no skip/keep option buggyness
Date: Sat, 10 Nov 2001 04:39:46 
Mime-Version: 1.0
Content-Type: text/plain; format=flowed
Message-ID: <F117mdMJOtd3rOE9TuH0000830b@hotmail.com>
X-OriginalArrivalTime: 10 Nov 2001 04:39:46.0704 (UTC) FILETIME=[B9914100:01C169A1]



>
>----- Original Message -----
>From: "Gareth Pearce" <tilps AT hotmail DOT com>
>
>
> >
> > hmm I hope you tested my patch, because although I tested it with no
>
>I did. It seemed fine.
>
> > problems, if your original patch didnt work - I wouldnt be supprised
>that
> > mine fails for the same reasons...
>
>Yours works because the state _SKIP is already handled. There is no
>state for 'keep' - just states for PREV/CURR/TEST && a test for those
>being the current installed version.

ahh okay I can see that ...
ugly hack forming in my head which says that since the package cant be 
SAME_PREV SAME_CUR or SAME_TEST - and since SAME_* are just a keep, it could 
be made SAME_TEST without repercussions to the outside world ... and with 
slightly better logic then my UI hack below ... still they are both hacks 
... and a nice clean fix would be nice ...

>
> > this was what I meant about not knowing
> > why it worked... I also thought being set to unknown should break my
>code
> > ... but only realised that after having tested it and having had it
>work jut
> > fine...
>
>Have you done much work with FSM's? set_action() is a fairly typically
>FSM transition function.

I know about DFA's which the equivelent pretty much.  and I didnt see 
anywhere that there was a transition to ACTION_SAME - but I didnt look at 
your original patch very closely.

>
> > if you want it to say keep instead of skip ... under the way my patch
> > works...
>
>...
>
>Yes, for the UI that will work, but the internal logic still needs to be
>straightened out. Keeping the gui unfixed will act as a reminder :} .
>Thoughts anyone? For the post-category release I'm hoping to have had a
>chance to do something on this, so that the impact shouldn't be huge.

*nod*
half a thought above...
>
>...
>
> > would submit a patch, but NTFS on win2k being read only mounted is
>stuffing
> > me arround... and I couldnt test it here in debian anyway...
>
>You might try plex86, I'm intending to use that for testing, if I can
>ever get it to work.

I used to use VMWare ... but free trials run out of time ... annoying 
that...
is plex86 free? I guess I could go find out...
>
> > cyg-apps for setup/winsup directory stuff? hmmm thought I read
>somewhere
> > that was not right... anyway ... I can do that :P
> > and cautious is a not wanting to tred on toes just yet :P I am sure
>that
> > will change after patch 3 or 4 ... :P
>
>Oh, don't worry about my toes, on public lists I tend to wear steel caps
>:}.
steel caps wont survive dynamite...
but I think I can keep from doing that much damage without too much trouble.

Gareth


_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp