Re: Quit versus Clear["Global`*"]
- To: mathgroup at smc.vnet.net
- Subject: [mg121273] Re: Quit versus Clear["Global`*"]
- From: DrMajorBob <btreat1 at austin.rr.com>
- Date: Wed, 7 Sep 2011 05:38:47 -0400 (EDT)
- Delivered-to: l-mathgroup@mail-archive0.wolfram.com
- References: <j3t5ga$58h$1@smc.vnet.net> <j3vc1k$ejo$1@smc.vnet.net>
- Reply-to: drmajorbob at yahoo.com
When in doubt, Quit Amen. Other remedies are always a kludge. Bobby On Tue, 06 Sep 2011 02:57:01 -0500, tenons at gmail.com <tenons at gmail.com> wrote: > On Sep 4, 4:14 am, David Bailey <d... at removedbailey.co.uk> wrote: >> On 03/09/2011 13:10, Themis Matsoukas wrote:> What is the difference >> between quitting the kernel and executing Clear["Global`*"]? >> >> > Is there a command to quit the kernel, instead of selecting >> Quit/Local from there Evaluate menu? >> >> > Thanks >> >> Quit[] will quit the kernel. >> >> Obviously Clear["Global`*"] would not clear information that might end >> up in other contexts, whereas Quit[] will clear everything. >> >> I never clear the global context - I just quit the kernel one way or >> another. Restarting the kernel used to be rather slow, but with the >> speed of a modern PC, there really doesn't seem much point doing >> anything else unless you specifically wish to preserve some information >> that is not in Global`. >> >> I tend to quit the kernel quite a lot - e.g. between successive runs of >> a program - why risk the confusion that can arise when two function >> definitions end up coexisting! >> >> David Baileyhttp://www.dbaileyconsultancy.co.uk > > I agree whole-heartedly with this post. I've found it the only way > to stay out of trouble. > > I have a related problem that I hope I can get some suggestions for. > > GIF.exe has a memory leak (reported previously, acknowledged and is on > someone's todo list at Wolfram, hopefully) which makes a certain set > evaluations (which use ExportString[#,"GIF"]& extensively) unstable. > (i.e., MathKernel crashes randomly and unceremoniously, or worse). > > What I would like to do is divide the set of evaluations into > segments, run them successively quitting the MathKernel after each, > and then restarting on the next segment. The Quit[] clears out the > buildup in GIF.exe. Obviously I can do this by babysitting and > restarting each segment by hand (which I do, at approximately 15 > minute intervals, painfully wasting time). > > [I've tried many things including Uninstall["path/GIF.exe"] 'ing > periodically and escaping to the shell to kill the process. Neither > improves stability. ] > > How can I carry this out automatically? > -- DrMajorBob at yahoo.com