TCP Wrappers Port

Corinna Vinschen cygwin@cygwin.com
Fri Jul 27 15:21:00 GMT 2001


On Fri, Jul 27, 2001 at 05:44:12PM -0400, Prentis Brooks wrote:
> Please don't take my words the wrong way.  I am only considering this based
> on win32 operability.  sys_nerr and sys_errlist work in your patch now, but it
> is possible that it may change later, while the strerror() is part of the
> win32 platform.  Also, sys_nerr and sys_errlist do not contain any of the
> win32 error codes.  But, I am not saying we have to make this change, and I am
> not looking to do this without some idea as to how the Cygwin community wants
> to manage this.  If we stick with sys_nerr and sys_errlist (Unix basis) then
> the community is committing to keeping that functionality in the Cygwin
> environment, we keep this patch and move forward.  Personally, I like the
> simplicity of your patch.  However, if the community does not want to make that
> commitment and chooses instead to work more in line with the win32 based
> API, then we need to use the str_error() route.  Both work, two answers to the
> same problem, and both have their good points and bad.  It is just a matter
> of which way the community chooses to support.  I can work this either way,
> in fact, your patch is easier to maintain on the tcp_wrapper end.
> 
> I am waiting for some feedback from the other maintainers on this list,
> specifically those responsible for the heart of cygwin and I intend to follow
> their lead.  Once I have some response to this, I plan to contact the
> author and explain the patches as well.  If they choose to maintain the
> sys_nerr and sys_errlist, as someone mentioned earlier, then we just move
> forward.  I hope, I am wording this properly as I do not mean to hurt any
> feelings or suggest that your patch was not sufficient.  I just happened to
> find another way before I found yours.
> 
> On Fri, 27 Jul 2001, Mumit Khan wrote:
> [...]
> > I really don't see why you would want to change anything at all! My
> > trivial patch does the job just fine, and it's been accepted by the
> > author, which means that the next release should not need any changes
> > for Cygwin.
> >
> > Principle of minimum changes to existing packages is a good one.

Just my 2 ct:

- The principle of minimum changes is actually a good thing. The
  less intrusive Cygwin specific changes to existing packages are,
  the more attendance you'll get from the package maintainers to
  include the Cygwin patches into the main source tree. Which in turn
  lowers your need to patch packages again and again when new versions
  are released.

- Since the author has already included Mumit's Cygwin related patch
  it's better to follow that way.

- If you think to have a better solution feel free to work closer to
  the author on the next package release. It's usually better to
  have a new generic solution instead of a target specific patch.

Tell me when I shall stop bloviating.

Btw., the next Cygwin version will export sys_nerr and sys_errlist,
AFAICS. So the patch should be a temporary solution, anyway.

Corinna

-- 
Corinna Vinschen                  Please, send mails regarding Cygwin to
Cygwin Developer                                mailto:cygwin@cygwin.com
Red Hat, Inc.

--
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/



More information about the Cygwin mailing list