On Tue, 18 Dec 2012 01:09:31 +0100 Vincent Torri <vincent.to...@gmail.com> said:

> On Tue, Dec 18, 2012 at 12:41 AM, Carsten Haitzler <ras...@rasterman.com>
> wrote:
> > On Mon, 17 Dec 2012 22:43:36 +0100 Joerg Sonnenberger
> > <jo...@britannica.bec.de> said:
> >
> >> On Mon, Dec 17, 2012 at 07:10:01PM -0200, Gustavo Sverzut Barbieri wrote:
> >> > Jokes aside, we appreciate to have E to run well on that, but it
> >> > cannot block development because features that is in the standard for
> >> > 12 years without being implemented.
> >>
> >> First of all, shm_open is an optional part of POSIX. One of the reasons
> >> for reluctantly implementing it is that it is extremely lame. The way
> >> Linux implements it provides ~0 advantage of just using open() with a
> >> file in /tmp.
> >
> > that is total BUNK. a file in /tmp quite possibly lives ON DISK. it SURIVIES
> > REBOOTS...
> 
> http://pubs.opengroup.org/onlinepubs/009604499/functions/shm_open.html
> 
> "When a shared memory object is created, the state of the shared
> memory object, including all data associated with the shared memory
> object, persists until the shared memory object is unlinked and all
> other references are gone. It is unspecified whether the name and
> shared memory object state remain valid after a system reboot."
> 
> see the last sentence ?

a file on disk is specified to survive reboots.. UNLESS it is specifically in
part of the fs that is discarded on reboot. vast difference between them.

the shm_open impl on linux does ensure reboots == doesn't survive. beats me
what everyone else does - but by using shm_open() i get to not care. it's not
my problem. it's a "base os problem". if i use /tmp - it may be a ramdisk
(discarded). a reboot MAY do an rm -rf in it and have it on real disk. in fact
historically having /tmp on real disk has been the most common - and it
survives reboots... and you nicely missed the next bit which is... syncing to
disk. shm_open at least asks the impl to do the sensible thing without pushing
that problem back into each and every app or library higher up. shm_open can be
emulated elsewhwere and too bad then if it syncs to disk if there is no viable
way of choosing a "ramdisk".

> Vincent
> 
> >and it will get SYNCED TO DISK WHEN THE FS WANTS TO... an shm file
> > only hits disk when you need to swap- like any ram you may malloc().
> >
> > UNLESS /tmp happens by luck to be a ramdisk (tmpfs).
> >
> > so no. it's far from lame. it's EXACTLY what is needed. if shm_open HAPPENS
> > to just simply know that this is the os setup and then use /tmp = fine...
> > but then this is a detail each and every app doesn't have to be configured
> > separately for or go figure out.
> >
> > --
> > ------------- Codito, ergo sum - "I code, therefore I am" --------------
> > The Rasterman (Carsten Haitzler)    ras...@rasterman.com
> >
> >
> > ------------------------------------------------------------------------------
> > LogMeIn Rescue: Anywhere, Anytime Remote support for IT. Free Trial
> > Remotely access PCs and mobile devices and provide instant support
> > Improve your efficiency, and focus on delivering more value-add services
> > Discover what IT Professionals Know. Rescue delivers
> > http://p.sf.net/sfu/logmein_12329d2d
> > _______________________________________________
> > enlightenment-devel mailing list
> > enlightenment-devel@lists.sourceforge.net
> > https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
> 
> ------------------------------------------------------------------------------
> LogMeIn Rescue: Anywhere, Anytime Remote support for IT. Free Trial
> Remotely access PCs and mobile devices and provide instant support
> Improve your efficiency, and focus on delivering more value-add services
> Discover what IT Professionals Know. Rescue delivers
> http://p.sf.net/sfu/logmein_12329d2d
> _______________________________________________
> enlightenment-devel mailing list
> enlightenment-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
> 


-- 
------------- Codito, ergo sum - "I code, therefore I am" --------------
The Rasterman (Carsten Haitzler)    ras...@rasterman.com


------------------------------------------------------------------------------
LogMeIn Rescue: Anywhere, Anytime Remote support for IT. Free Trial
Remotely access PCs and mobile devices and provide instant support
Improve your efficiency, and focus on delivering more value-add services
Discover what IT Professionals Know. Rescue delivers
http://p.sf.net/sfu/logmein_12329d2d
_______________________________________________
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Reply via email to