See y'all there.
PS: Note that it looks pretty stuffy over there; you might want to think about bringing water.
Click HERE for a printer friendlier version of the map.
To print to the colour printer, log into physics and execute, e.g.
% lpr -Plp2 -Zsimplex your-project-name-here.psPlease use -Zsimplex on whatever printer you use (again best to always print your final copy from physics, since that's where control of your printing is easiest), so that your term project is printed one-sided.
The solution to this problem is simply to choose another port number, and inform the application(s) that use this information (xvs, DV, sdftoxvs, sdftoDV etc. of the new choice. This is done via setting of the environment variables XVSPORT and DVPORT respectively. I suggest that you choose the port numbers simply by adding your userid (the third entry in /etc/passwd for your account name) to the default port numbers. For example, as you can verify by greping for 'matt' in /etc/passwd, matt's UID is 243. Thus I would set
setenv XVSPORT 5243 setenv DVPORT 5248and, logically, place these lines in my ~/.cshrc file. (Please, please, don't put the above lines in YOUR ~/.cshrc. If 2 or more of you do so, you will defeat the disctinctiveness mechanism!!!).
On the lnx machines, there are instructor-scripts getXVSPORT and getDVPORT that should be in your path, and that will return your UID added to 5000 or 5005 respectively, so that you can use
setenv XVSPORT `getXVSPORT` setenv DVPORT `getDVPORT`Note that all of the quotes in the above are backticks, a.k.a. the most powerful punctuation known to humankind.
We hope to be sending your grades/comments for Homework 4 by midnight tonight, and hope to do the same for Homework 5 by noon on Saturday.
Please remember that
Given the time of year, and in order to minimize frustration, from now until the exam, if you have problems accessing the course machines, you should feel free to phone me either at
You can also continue to e-mail, and I will endeavour to try to keep things alive as much as possible.
Sorry again for the inconvenience caused earlier in the day.
However, I see that some folk other than the 555 registrants are attempting the whole assignment. Thus to minimize frustration, please note that the script lino alluded to in Problem 2e) should be lines. (lino is an antecedent of lines, with, as you may have discovered, a minimalist usage message.)
The online handout has again been updated.
I have updated the aforementioned source file in ~phys410's account accordingly.
I once again apologize for any confusion this has called.
I will distribute hardcopies in tomorrow's class.
Jason has found another typo in the Homework 4 handout. Equation (2.8) should begin
n
( t , x )
j
rather than
n
( t , u )
j
The online handout has again been updated.
Jon Benjamin has pointed out that in the source files ~phys410/hw4/a[23]/diffusion[24]-priv.f, the comment lines
c u0(x) := u(t,0) = sin(2 Pi x)should be replaced with
c u0(x) := u(0,x) = sin(2 Pi x)The on-line code has been updated accordingly.
C u = fwhere C is a n x n tridiagonal matrix, u is the unknown vector, and f is the right hand side of the linear system that u satisfies.
Arguably (he pleads), for a physicist this is a good notation, since except at the boundaries we then have
f = f(x )
j j
where one must understand that the f's that appear on the two sides of
the above equation are NOT identical; one is the right hand side vector for a
linear system satisfied by a particular vector, u, the other is a vector
resulting from the "restriction" of a continuum function, f, to a vector of
x-values, x.
Pedagogically, however, there can be NO argument that the original prescription left a lot to be desired! Thus, I have thus now modified the problem handout, as well as the template files ~phys410/hw4/a[23]/diffusion[24].mws so that you are now asked to set up a linear system of the form
C u = bwhere C is a n x n tridiagonal matrix, u is the unknown vector, and b is the right hand side of the linear system that u satisfies.
Management should probably be horsewhipped, and apologizes profusely for the confusion that was caused.
The online handout has been updated to reflect all three of these items.
execute diffusion2should read1 1 0.25 lambda 0 1 execute diffusion4 1 1 0.25 lambda 0 1
execute diffusion2The on-line handout has been updated.6 1 0.25 lambda 0 1 execute diffusion4 6 1 0.25 lambda 0 1
For those of you HAVE started the assignment, please note that I have made numerous typo-level modifications to the handout, as well as several small changes of various of the template files in ~phys410/hw4/a[23]. If possible, I would advise recopying any files that you have already copied from phys410, AFTER BACKING UP YOUR CURRENT WORK. You can then use, e.g. 'diff -r new-dir old-dir' to see the minor modifications in the template files.
Thanks to Jason Penner for pointing out a glitch in one of the template files.
Don't be alarmed by the length of the handout. To a paraphrase an old adage, it hurt me a lot more than it's going to hurt you.
As always, let me know if you have any undue problems with the homework, or if you find bugs in the handout.
% Mandelbrot 500should read
% Mandelbrot 501The online version has been updated.
TUESDAY, DECEMBER 14 2004, 3:30 PM LOCATION TBALet me know immediately if you have a conflict.
... called ~/hw2/a3/procs. I must be able to read ~/hw2/a3/procs ...
should read
... called ~/hw2/a2/procs. I must be able to read ~/hw2/a2/procs ...
i.e. the answer to Problem 2 should live in directory ~/hw2/a2. The on-line version has been updated to reflect
the bug-fix.
The Guide is an updated version of the one whose first chapter I distributed to you, so those of you interested in the full details of Maple programming should find it quite useful.
In deciding which of 9.2.1 or 10.0 (if either) you decide to install, note that 10.0 is, of course, the most recent version, but one with which Pal and I currently have no experience (i.e. we are somewhat more likely to be able to help you out with problems you have with 9.2.1).
Students who wish to avail themselves of this opportunity to install Linux on one or more home machines should feel reasonably comfortable with configuration of their PCs. Although the Mandrake distribution is quite user-friendly, and although installation of Mandrake on a system that already has a version of Windows installed should be straightforward in most cases (leaving you with a dual-boot system), there is still the possiblity that something could go awry, and that, e.g., you could end up trashing your Windows partition.
Best systems-administration practice dictates that one always make a complete backup of a system prior to commencing work such as Linux co-installation. I whole-heartedly commend this practice to you, and, unfortunately cannot except responsibility (nor can Pal) for any loss of data that may result from failed attempts to install Linux on Windows boxes.
Darn, I thought I'd be able to sneak this one past you. This a matter out of the instructor's control, and a request has been made to remedy the situation. If all of this means nothing to you, with luck it will in a week or two.
dastegir matthewe jonben tjblair gbosdet awbowie wochao szumeng chin manixc rickychu dhanesar jwd djeaton fameli lfan pgf jgallant jgodfrey benjamin arefa masrur kasian chinakt bayo lc xiangli mcliu martham tor ajpenner cpfuller robinr mayaro rspence sstorey dongen tswang mandyman wanyan mencywoo syewchukIf you wish, and if your account appears in the above list, you should now be able to ssh into the lnx machines from a remote host via any one of
ssh userid@lnx1.physics.ubc.ca ssh userid@lnx2.physics.ubc.ca ssh userid@lnx3.physics.ubc.cawhere userid is to be replaced with your own user id. You should also be able to login into the lnx machines via their consoles in Hennings 205. Your password should be the same as that for your account on physics.ubc.ca. If you have any difficulties logging in, and if the NOTES concerning machine usage are not sufficient to solve these problems, let me know via e-mail immediately.
As several of you are aware, the class is currently officially full. Should you still wish to register for the course, please send an email to the instructor, Matt Choptuik.