Mathematica 7.01 and Mac OS 10.6 (Snow Leopard)
in [Mathematica]
|
Prev: trading
Next: closing curly brace unrecognized
From: monochrome on 13 Oct 2009 23:25 I have experienced this once and was directed by customer service to do the same thing. It has not recurred. Bayard On Oct 13, 4:18 am, Andrzej Kozlowski <a...(a)mimuw.edu.pl> wrote: > Actually the situation is even worse than I described below. When I > try to restart Mathematica after force quitting it, I get the message > that the 2-user limit on this license has been reached. Contact > Wolfram Research or an authorized Mathematica distributor for > information on upgrding your license configuration. > > It seems that I have to restart my computer to be able to use > Mathematica again, which is really getting somewhat ridiculous. > > Andrzej Kozlowski > > On 13 Oct 2009, at 17:56, Andrzej Kozlowski wrote: > > > > > I have been experiencing rare but very irritating problems running > > Mathematica 7.01 under Mac OS 10.6 (Snow Leopard) on a MacBook. What = > > happens is that for some reason the kernel suddenly quits during an > > evaluation. When I try to restart it (by evaluating some expression) = > > I get the message: "The kernel failed to start because you attempted = > > to run more Mathematica kernels than you have licensed." There are > > two buttons, Purchase and Cancel. I find it rather annoying that the = > > Purchase button is selected by default but not half as annoying as > > the fact that clicking on either button produces no effect. > > Everything is actually frozen and there is no way out other than > > force quitting Mathematica. > > > Has anyone experienced anything similar? As I mentioned - this seems = > > fairly rare; it has happened twice since I installed Snow Leopard > > about a month ago and I can't reproduce it. > > > Andrzej Kozlowski
From: John Fultz on 13 Oct 2009 23:25 On Tue, 13 Oct 2009 07:18:38 -0400 (EDT), Andrzej Kozlowski wrote: > I have been experiencing rare but very irritating problems running > Mathematica 7.01 under Mac OS 10.6 (Snow Leopard) on a MacBook. What > happens is that for some reason the kernel suddenly quits during an > evaluation. When I try to restart it (by evaluating some expression) I > get the message: "The kernel failed to start because you attempted to > run more Mathematica kernels than you have licensed." There are two > buttons, Purchase and Cancel. I find it rather annoying that the > Purchase button is selected by default but not half as annoying as the > fact that clicking on either button produces no effect. Everything is > actually frozen and there is no way out other than force quitting > Mathematica. > > Has anyone experienced anything similar? As I mentioned - this seems > fairly rare; it has happened twice since I installed Snow Leopard > about a month ago and I can't reproduce it. > > Andrzej Kozlowski We've been investigating this problem internally. Our initial investigation suggests that this happens only when a Java runtime launched by the kernel has been left running (for reasons yet unknown), even though the kernel died. If you kill the Java runtime, then it'll free up the process. The best way I know of to kill the process is the old-fashioned, Unix-style method. I.e., open up a Terminal session, use 'ps' to finds its pid, and terminate it with 'kill -9'. We're working to resolve this and other Snow Leopard issues. We've been a bit slower on this then I would have liked...partly because of a collision of internal resources, and partly because a couple of the Snow Leopard issues have been difficult to reproduce and diagnose. But I'm hoping to have more to say on the issue in the near future. Sincerely, John Fultz jfultz(a)wolfram.com User Interface Group Wolfram Research, Inc.
From: DrMajorBob on 14 Oct 2009 07:56 My M7/Snow Leopard problem is that the kernel often freezes (while loading) the first time I evaluate a cell after opening Mathematica. Then I go to Evaluate>Stop Kernel, and trying again always seems to work. It's not a HUGE problem, but it wastes a minute of my time whenever it happens... about 20% of my Mathematica sessions. Bobby On Tue, 13 Oct 2009 22:18:07 -0500, monochrome <bayard.webb(a)gmail.com> wrote: > I have experienced this once and was directed by customer service to > do the same thing. It has not recurred. > > Bayard > > On Oct 13, 4:18 am, Andrzej Kozlowski <a...(a)mimuw.edu.pl> wrote: >> Actually the situation is even worse than I described below. When I >> try to restart Mathematica after force quitting it, I get the message >> that the 2-user limit on this license has been reached. Contact >> Wolfram Research or an authorized Mathematica distributor for >> information on upgrding your license configuration. >> >> It seems that I have to restart my computer to be able to use >> Mathematica again, which is really getting somewhat ridiculous. >> >> Andrzej Kozlowski >> >> On 13 Oct 2009, at 17:56, Andrzej Kozlowski wrote: >> >> >> >> > I have been experiencing rare but very irritating problems running >> > Mathematica 7.01 under Mac OS 10.6 (Snow Leopard) on a MacBook. What = > >> > happens is that for some reason the kernel suddenly quits during an >> > evaluation. When I try to restart it (by evaluating some expression) = > >> > I get the message: "The kernel failed to start because you attempted = > >> > to run more Mathematica kernels than you have licensed." There are >> > two buttons, Purchase and Cancel. I find it rather annoying that the = > >> > Purchase button is selected by default but not half as annoying as >> > the fact that clicking on either button produces no effect. >> > Everything is actually frozen and there is no way out other than >> > force quitting Mathematica. >> >> > Has anyone experienced anything similar? As I mentioned - this seems = > >> > fairly rare; it has happened twice since I installed Snow Leopard >> > about a month ago and I can't reproduce it. >> >> > Andrzej Kozlowski > > -- DrMajorBob(a)yahoo.com
From: Andrzej Kozlowski on 14 Oct 2009 07:56 On 14 Oct 2009, at 04:02, John Fultz wrote: > On Tue, 13 Oct 2009 07:18:38 -0400 (EDT), Andrzej Kozlowski wrote: >> I have been experiencing rare but very irritating problems running >> Mathematica 7.01 under Mac OS 10.6 (Snow Leopard) on a MacBook. What >> happens is that for some reason the kernel suddenly quits during an >> evaluation. When I try to restart it (by evaluating some >> expression) I >> get the message: "The kernel failed to start because you attempted to >> run more Mathematica kernels than you have licensed." There are two >> buttons, Purchase and Cancel. I find it rather annoying that the >> Purchase button is selected by default but not half as annoying as >> the >> fact that clicking on either button produces no effect. Everything is >> actually frozen and there is no way out other than force quitting >> Mathematica. >> >> Has anyone experienced anything similar? As I mentioned - this seems >> fairly rare; it has happened twice since I installed Snow Leopard >> about a month ago and I can't reproduce it. >> >> Andrzej Kozlowski > > We've been investigating this problem internally. Our initial > investigation > suggests that this happens only when a Java runtime launched by the > kernel has > been left running (for reasons yet unknown), even though the kernel > died. If > you kill the Java runtime, then it'll free up the process. Thanks a lot, that explains everything. Some people have written advising me to kill any additional Mathematica processes that were running but, of course, I already knew that though forgot to mention it in my post. There were no recognisable Mathematica processes (or anything beginning with Math) running. I normally use Apple's Activity Monitor to kill any processes that cause problems, and that was the first thing I looked at. This may be a bit naive question, but what it the Java runtime process name when it is launched by the Kernel? Right now the only Java related process that I see in the Activity Monitor is something called JavaApplicationStub, actually two of them, but none of them seems to be it (as they do to quit when I quit Mathematica). Andrzej Kozlowski > > The best way I know of to kill the process is the old-fashioned, > Unix-style > method. I.e., open up a Terminal session, use 'ps' to finds its > pid, and > terminate it with 'kill -9'. > > We're working to resolve this and other Snow Leopard issues. We've > been a bit > slower on this then I would have liked...partly because of a > collision of > internal resources, and partly because a couple of the Snow Leopard > issues have > been difficult to reproduce and diagnose. But I'm hoping to have > more to say on > the issue in the near future. > > Sincerely, > > John Fultz > jfultz(a)wolfram.com > User Interface Group > Wolfram Research, Inc. > >
From: Joel Reymont on 14 Oct 2009 07:57
On Oct 14, 4:25 am, John Fultz <jfu...(a)wolfram.com> wrote: > > We've been investigating this problem internally. Our initial investigation > suggests that this happens only when a Java runtime launched by the kernel has > been left running (for reasons yet unknown), even though the kernel died. If > you kill the Java runtime, then it'll free up the process. Why does the kernel launch a Java runtime? |