Re: Repost, different list...File::Spec, cygwin, Syntactic vs. Semantic path analysis

["lhall at pop dot ma dot ultranet dot com" <lhall at pop dot ma dot ultranet dot com>]

Ah, yes.  Teeth gnashing!  Those were the good old days! ;-)

Larry


Original Message:
-----------------
From: Rick Rankin rick_rankin@yahoo.com
Date: Fri, 10 Jan 2003 11:44:47 -0800 (PST)
To: cygwin@cygwin.com
Subject: Re: Repost, different list...File::Spec, cygwin, Syntactic vs. 
Semantic path analysis



--- Christopher Faylor <cgf@redhat.com> wrote:
> On Fri, Jan 10, 2003 at 10:30:23AM -0800, Shankar Unni wrote:
> >linda w (cyg) wrote:
> >>What were the _original_ design goals of Cygwin -- i.e.  as sponsored
> >>by "RedHat"?
> >
> >Cygwin predates RedHat.  See http://cygwin.com/history.html (the
> >earliest date in the file is Dec 1995).  RedHat bought Cygnus Solutions
> >(which was a shop for commercial support for GNU software, especially
> >GCC ports to obscure and new platforms), which did the original Cygwin
> >work.
> >
> >Anyone at RedHat from the original Cygwin team (the last warriors of
> >the (in)famous "Beta 20" :-)?) wanna answer this?
> 
> Like me, for instance?  I came onboard in '98 and talked to most of the
> initial developers who had eventually stampeded away from the (to them)
> distasteful duty of working on Windows.  I'd been involved with cygwin
> (aka gnu-win32) since early '97.
> 
> >There's an interesting line in the early changelogs:
> >
> >   Release Beta 8
> >   [...]
> >   Much nicer way of describing paths, eg //c/foo is c:\foo.
> >
> >Suggests that the early goal *was* to provide a POSIX-y view, and the 
> >exposing of Windows paths was added as a convenience..
> 
> Posix paths were one of the main reasons for cygwin.  The goal was to to
> modify tools like gcc and make as little as possible so that Cygnus
> could have a Windows toolchain but not force tool developers to deal
> with modifying every line of code which assumed that '/foo' meant "the
> file foo in the root directory" rather than "the file foo at the root
> directory of the current drive" or "the foo option".
> 
> I've been managing support for cygwin and have had to answer the "Why
> doesn't gcc deal with my c:\include paths very well" questions for
> years now.  Most people get the concept once it is explained to them.
> YMMV.
> 
> So, anyway, fork, exec, and posix paths were the main motivations for
> cygwin.  Once I came onboard, you could add signals to that list, too.
> 
> But, hey, if you don't believe me, then maybe Larry Hall has more
> credibility.  He's been around longer than I.
> 

I think I picked up my first GNU-Win32 package around b3 maybe? I seem to
remember discussions about changing the name to GNU-Win32, so I don't even
think that's what it was called when I first got it. Anyway, I recall when
the
// convention was added so that //c mapped to C:/, for example, and there
was
*much* discussion around the posix standard and its interpretation of the
leading //.  I also recall that there was much weeping, wailing, and
gnashing
of teeth when the interpretation of // was changed to what it is now, but
that's another story...

FWIW,

Rick

--
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/


--------------------------------------------------------------------
mail2web - Check your email from the web at
http://mail2web.com/ .