This is Mathematica in view mode; [Up]
Date: Sun 22-Sep-1989 03:41:33 From: Unknown Subject: Re: Mathematica (NeXT 1.0) I have some interesting problems with mathematica to report. I'd be interested to hear from anyone who has found others. My machine is the 8 MB od only version. I've noticed problems with memory before when running mathematica. Today I tried finding the inverse of a certain symbolic matrix. I watched the memory with ps -xva. The program seems to hang onto disk space once it is used for swap and never lets go. When the thing really got going, it wouldn't even let me kill it with the skull and crossbones. I got some failure panel for the abattoir. The swapfile increased in size to about 32 MB and when I was finally able to kill the process from shell it kept the space anyway (not quite the definition of swap is it? But wait...) I tried to copy some of the ps output to writenow but there was not enough space to open it(wn). I tried edit, but not enough space to save the file. I logged out. Then back in to get a message to the effect that my disk space was very low. The doc tried to do its thing opening term, shell and wn but not enough space. I logged out and turned the machine off. When I got back up, the doc was empty except for the next block at the top...It seems to me this is a very LARGE bug that makes it well nigh immpossible to use mathematica for anything significant. Any hope of fixing it soon? Are you listening NeXT?? And yes I did repeat this with the same results on another cube. HELP! I need this thing to do symbolic stuff. It doesn't even have to do it very fast...just correctly. How about porting MAPLE to the cube?? I'm stuck with the cube for a while since I had to do some ranting to get one in the first place. Come on guys, make me happy...just a little bit?? Oh yeah, mathematica has some other bugs. The Eigensystem command fails on most anything you want to try... For instance the matrix {{1,2,3},{a,0,0},{1,2,3}}. Gee, this one I can do by hand in a few minutes...HHHHEEEEEELLLLLLLPPPPPP!!!!!!! Bill Smith smithw@yvax.byu.edu >From: avie@wb1.cs.cmu.edu (Avadis Tevanian)
Date: Sun 23-Sep-1989 00:04:21 From: Unknown Subject: Re: Mathematica (NeXT 1.0) In article <811smithw@yvax.byu.edu> smithw@yvax.byu.edu writes: > > I have some interesting problems with mathematica to report. I'd be >interested to hear from anyone who has found others. My machine is >the 8 MB od only version. I've noticed problems with memory before >when running mathematica. Today I tried finding the inverse of a >certain symbolic matrix. I watched the memory with ps -xva. The program >seems to hang onto disk space once it is used for swap and never lets go. >When the thing really got going, it wouldn't even let me kill it with >the skull and crossbones. I got some failure panel for the abattoir. >The swapfile increased in size to about 32 MB and when I was finally able >to kill the process from shell it kept the space anyway (not quite the >definition of swap is it? But wait...) You are running into a problem where you just plain need more disk space due to the size of Mathematica. If you need to perform Mathematica computations that use that much memory, then you should free up space on your disk. To do this, look at the printed release notes, which suggest several files/directories that you may want to remove (e.g., TeX, documentation, ...). Also, the operating system is capable of shrinking the swapfile. It will shrink the size of the swapfile to correspond to the maximum page left in the swapfile. Unfortunately, what often happens is that a large program will extend the swapfile to size n, another long lived background process will cause a new page to be swapped to n+1, the original large program will exit, but the swapfile will remain at size n+1 because it is the largest. Although the space freed in the middle of the file will be reused, the fact of the matter is that the file is still large. The sure fire way to reclaim the swap space is to reboot the system, which truncates the swapfile to a predetermined size (16 meg, I believe).
These are the contents of the former NiCE NeXT User Group NeXTSTEP/OpenStep software archive, currently hosted by Marcel Waldvogel and Netfuture.ch.