Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm List-Subscribe: List-Archive: List-Post: List-Help: , Sender: cygwin-owner AT cygwin DOT com Mail-Followup-To: cygwin AT cygwin DOT com Delivered-To: mailing list cygwin AT cygwin DOT com X-Authentication-Warning: eos.vss.fsi.com: ford owned process doing -bs Date: Wed, 28 Jan 2004 13:42:32 -0600 (CST) From: Brian Ford X-X-Sender: ford AT eos To: cygwin AT cygwin DOT com cc: David Balazic Subject: Re: Lost g++ after update In-Reply-To: Message-ID: References: <600B91D5E4B8D211A58C00902724252C01BC04D3 AT piramida DOT hermes DOT si> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-IsSubscribed: yes Reply-To: cygwin AT cygwin DOT com This is the last one from me on this topic, I promise. On Wed, 28 Jan 2004, Igor Pechtchanski wrote: > On Wed, 28 Jan 2004, David Balazic wrote: > > > A few days ago I downloaded the actual setup.exe ( version 2.416 ) and run > > it to update my cygwin installation. > > I just pressed "Next" on each page. After it finished , I rebooted ( it > > suggested so ). > > > > Then I noticed that I don't have g++ any more : > > > Install the "gcc-g++" package. A package search at > for "bin/g\+\+.exe" should have found it, > but there seems to be a bug (which I'm looking into) that truncates the > match list before it displays "gcc-g++", which may have prompted this > question. > No, the obvious reason for the question is that he simply did a normal update and lost the g++ functionality that he had previously had. > > Then I run setup again and set it to reinstall gcc and texinfo. After that > > texinfo is OK, but g++ is still missing. > > > > Any clue what happened ? > > > > I am now running setup.exe again and manually selecting gcc-g++. > > It just finished. g++ works again. > > > > Why did it disappear ? > > > > Because setup by default only upgrades the packages you had installed or > their dependences. "gcc-g++" was neither. Reading the release > announcement would have been helpful here. > In a way, it was installed. The problem here is that the previous gcc pacakge contained all of the gcc, gcc-g++, gcc-g77, etc. functionality. Relying on the release announcement for this information seems like asking for this kind of problem report to me. Additionally, when once chooses all->install to just get everything, the gcc-core and gcc-testsuite source packages are *always* re-installed. If the old gcc package had become an empty placeholder dependening on the new packages that replaced its functionality, this question would not have been posted. Further more, if there were empty binary packages for the source only ones (gcc-core and gcc-testsuite), my second observation would not occur. <\end rant> It seems to me that the gcc package and the gcc-core packaga are backward (at least in FSF normal terms). I wish I knew how to fix it now, but it seems too late to reverse the package meanings, or fix the dependencies. Suggestions are more than welcome. IMHO, the empty binary packages for gcc and gcc-core would still be "a good thing". -- Brian Ford Senior Realtime Software Engineer VITAL - Visual Simulation Systems FlightSafety International Phone: 314-551-8460 Fax: 314-551-8444 -- 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/