Mail Archives: cygwin/2001/09/29/21:58:48
I think Christopher Faylor wrote:
> On Fri, Sep 28, 2001 at 03:16:27PM -0700, friedman_hill ernest j wrote:
> >
> > Here's the begining of a simple mkfifo implementation for Cygwin
> > [... I got some] rude and unhelpful responses [and] I abandoned
> > the project...
>
> I just searched the archives for some evidence of a "rude" response.
> ... I can't find any rude responses to your requests for help in
> the cygwin mailing list. [Perhaps] you are misremembering what
> happened. Maybe I somehow missed a rude response that soured you on
> the whole project or maybe you're just subject to hyperbole.
Well, let's see. It's entirely possible that I'm misremembering
precisely what happened, and I am indeed subject to hyperbole. I'm
basically a grumpy old man. The ultimate reason why I never went back
to the project was that I don't have a Win32 machine anymore. But I
remember being frustrated that I had figured out an idea for how to do
fifos, but I just couldn't get any advice on how to integrate the damn
thing into Cygwin. I also couldn't get the Cygwin DLL to build from
source, because the CVS top copy never seemed to compile at that
time. I was trying to figure out how to write and integrate an
fhandler, and I just couldn't seem to figure it out on my own. I was
enormously frustrated by the ORBS thing -- I couldn't get any advice
from the list because both my work mail and my ISP were blacklisted,
and nobody would talk to me one-on-one. I guess in retrospect I
probably could have gotten a Yahoo mail account or some other
subterfuge, but that was just too much monkey business for me to be a
part of (with apologies to Mr. Berry.)
Anyway, sorry to be a crank. I know no one else on this list ever acts
like a petulant child ;) I still read the list because I'm still
impressed by the amazing piece of work that is Cygwin -- I used it
constantly for something like four years, and I really can't thank
everyone involved enough. Thank you all again.
And Robert Collins wrote:
> I don't recall this... I certainly didn't see any
>
> > sufficiently rude and unhelpful responses
>
> to your emails - even though you apparently wrote your responses
> off-list! (see for example a reply to an email from you, but your
> email is no on the list archive...
> http://sources.redhat.com/ml/cygwin/2000-09/msg00447.html).
>
> I've also written a fifo implemention, feature complete except for a
> distressing tendency to die on fork, which I haven't looked at since
> february this year (and since then I've learnt a _lot_ more about
> fork()'s guts.
Rob, your implementation is a much more sophisticated version of
the same basic idea I had (which is at
http://herzberg.ca.sandia.gov/namedpipes) -- they both use Win32
memory-mapping objects for a header and data channel per fifo. Mine
uses semaphores instead of SignalObjectAndWait, otherwise, they
implement fifos in the same way. Now, yours deals with ntsec stuff and
lots of other nifty details that I never learned how to take care of.
It'd be great to get you version integrated into Cygwin.
The only Win32 machine I have now is Win95 installed in VMWare on
Linux. But that VM has Cygwin installed on it, and I still use it when
I have to do something on Windows. I'd like to help get your
implementation polished up, if there's something I could do. I'll try
to apply your patch and try it out for myself sometime this week.
---------------------------------------------------------
Ernest Friedman-Hill
Distributed Systems Research Phone: (925) 294-2154
Sandia National Labs FAX: (925) 294-2234
Org. 8920, MS 9012 ejfried AT ca DOT sandia DOT gov
PO Box 969 http://herzberg.ca.sandia.gov
Livermore, CA 94550
--
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/
- Raw text -