[cgf@redhat.com: fhandler redesign]

Robert Collins robert.collins@itdomain.com.au
Fri May 18 05:07:00 GMT 2001

I'd really like to see your daemon made generic, with some sort of
stable API for doing cross-process/global stuff (so that things like
allocating shm id's and holding shm memory areas open... don't need to
reinvent the wheel.)


----- Original Message -----
From: "egor duda" <deo@logos-m.ru>
To: <cygwin-developers@cygwin.com>
Sent: Friday, May 18, 2001 10:04 PM
Subject: Re: [cgf@redhat.com: fhandler redesign]

> Hi!
> Friday, 18 May, 2001 Christopher Faylor cgf@redhat.com wrote:
> CF> Any thoughts?
> CF> cgf
> CF> ----- Forwarded message from Christopher Faylor
<cgf@redhat.com> -----
> CF> From: Christopher Faylor <cgf@redhat.com>
> CF> To: Corinna Vinschen <vinschen@redhat.com>
> CF> Subject: fhandler redesign
> CF> Date: Sun, 1 Apr 2001 22:58:24 -0400
> CF> I would like to start thinking about an fhandler redesign.  The
> CF> method is, IMO, too bulky and not layered enough.
> CF> One problem is that there is no real "device" layer so many things
> CF> copied around between fds that should only exist at the device
> CF> One example of this is your recent save/restore screen changes.
> CF> should only be one save screen buffer and it should only exist at
> CF> device layer.  Since we don't have a real device layer, we have to
> CF> save the screen buffer in an opened fd.
> there's a problem with persistent-across-address-space-boundaries
> information, and functionality, associated with it. we should either
> designate one of cygwin processes to be "master" for some device,
> like it's done with ttys, or use separate server.
> Egor.            mailto:deo@logos-m.ru ICQ 5165414 FidoNet

More information about the Cygwin-developers mailing list