This is the mail archive of the cygwin-developers mailing list for the Cygwin project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Updated: 1.7.0-34 (...and a plea)


On Dec  5 11:51, cygwin@cwilson.fastmail.fm wrote:
> Corinna Vinschen wrote on Fri, 5 Dec 2008 12:07:33 +0100
> > On the not so bright side, I'm a bit disappointed that only few
> > maintainers are actually rebuilding their applications for 1.7.  All
> > applications will probably profit at least from the now long pathnames
> > up to 32K.  And many network related applications will profit from IPv6
> > support and the new getaddrinfo family of functions.
> 
> With which gcc? [...]
> The reason I ask is because such a compiler switch would ALSO require
> rebuilding many of the packages, yet again. (C++ definitely, C? maybe;
> possible ABI change, as well as the shared libgcc thing). See this thread:
> http://cygwin.com/ml/cygwin-apps/2008-10/msg00090.html
> which I started two days before Dave Korn disappeared. (Err...his "a
> week to a fortnight" has become "a month and counting" Has anybody heard
> from Dave?)

No, unfortunately not.  As for C, I don't care if the application has
been built with gcc3 or gcc4, but that's just me.  I don't think it
makes a lot of sense to enforce shared libgcc and I don't think it makes
a lot of sense to worry about sjlj for C applications.

I'm not (better: even less) qualified to answer this question for C++
but afaiu C++ apps are affected and should all be built using gcc4.

> > You can even run them in parallel in two different console windows
> > without any bad interaction.  Isn't that a good reason to start porting
> > to 1.7?
> 
> I've been burned by this twice (or three times, depending on how you count)
> 
> Setup-1.7 (build 2.606)
> http://cygwin.com/ml/cygwin-apps/2008-08/msg00128.html (update #2a)
> http://cygwin.com/ml/cygwin-apps/2008-08/msg00129.html (new install #2b
> or #3)

Puzzles me.  I've installed on 2K8 with UAC enabled and the permissions
were as I wanted them.

Can't you debug this?  The permission code in setup is really simple, it
just uses the current user and primary group and creates the permissions
according to the permissions in the tar file.  The predefined dirs are
created with 755 and in some cases 1777 permissions.  The problem with
terminfo should be fixed in the meantime by changing the dependency order,
partly in the hint files and partly in setup-1.7.exe.

> [...]
> Concerning setup-1.7, and hassles concerning setup-1.7("new" install or
> upgrade) + cygwin-1.7sidebyside + vista + UAC + non-adminstrative user,
> MAYBE they've all been resolved now. I don't know.

Me neither, but as I said above...  vista == 2K8, at least related to
NTFS.

> [...]
> However, there was an
> additional change in the source code on 29 Aug, dealing with the
> dependency order resolution, that is not yet reflected in the released
> binary.  This change was the result of a giant subthread here:
> http://cygwin.com/ml/cygwin-apps/2008-08/msg00087.html
> 
> Maybe we should have a new 2.610 for those changes?

I see to it.  Maybe you want to give the current setup-1.7 a try,
just for testing a temporary installation?

> FYI, I never did get a definitive OK to proceed in this thread:
> http://cygwin.com/ml/cygwin-apps/2008-07/msg00024.html

I don't see how that's relevant for 1.7 testing and packaging.

> But, if I can get a definitive answer to the gcc4 question, I'll try to
> more proactive updating my (binary) packages for 1.7.

Well, for your pure C packages...


Corinna

-- 
Corinna Vinschen                  Please, send mails regarding Cygwin to
Cygwin Project Co-Leader          cygwin AT cygwin DOT com
Red Hat


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]