SSS homepage--program list--postings log
General Information Updates
Here are updates to our general information concerning the SSS Project. Postings regarding individual programs may be accessed from the program list. Click on the title to see the posting and addenda to it. You may also check the postings log to see a list of the recent additions to the postings.
We urge you to send comments, bug reports, suggested program modifications, novel ways to use the programs, and interesting presets using this posting form (i.e., e-mail to sss@ruph.cornell.edu).
1.19 -- 08/12/05 --More Linux installation problems
Here is Barry Robinsons's response to a reported problem with a Linux installation:
I'm running Fedora Core 1 at home and a mixture of Redhat 7.3 and Redhat Enterprise 3 update 4 at work. I do not have access to any 2.6 kernel based Linux distributions, but I don't think that matters. The errno problem came up a few years ago, and incompatability with current versions of Tcl/Tk/Blt were with us almost from the beginning. If you carefully follow the instructions in README_COMPILE in sss/linuxbin, you should be able to rebuild all the simulations.
I rebuilt them at home already. Let me write down a procedure and try it on RHEL3.
Currently there are 4 tar files on the main SSS site and you need to download them all to rebuild. Go to: http://www.physics.cornell.edu/sss/download.html and under the unix section get "full linux", "tcl", "tk", and "blt" which are linuxall_new.tar.gz, tcl7.6p2_sss.tar, tk4.2p2.tar, and blt19sss.tar respectively. Assuming they are all in the same directory:
tar -xzvf linuxall_new.tar.gz
# This will give you an sss directory.
cd sss
tar -xvf ../tcl7.6p2_sss.tar
tar -xvf ../tk4.2p2.tar
tar -xvf ../blt19sss.tar
# The above will yeild tcl7.6/, tk4.2/, and blt-1.9/
# Grab the update tclPosixStr.c file from
#http://www.physics.cornell.edu/sss/data/updates.html#1.14
# and overwrite the versiion in tcl7.6/generic/
cd tcl7.6/unix
./configure --disable-load --enable-gcc
make
# Do not "make install", rather
cp libtcl7.6.a ../../linuxlib/
# Repeat for tk
cd ../../tk4.2/unix
./configure --disable-load --enable-gcc
make
cp libtk4.2.a ../../linuxlib/
# use ranlib to update index
cd ../../linuxlib ranlib libtcl7.6.a
ranlib libtk4.2.a
# You also need to make symlinks here for blt
ln -s libtcl7.6.a libtcl76.a
ln -s libtk4.2.a libtk42.a
#Repeat mostly for blt
cd ../blt-1.9
./configure --disable-load --with-gcc
# You may have to tell it that the tcl and tk libs live in ../linuxlib/
# when it asks.
# Edit the resulting Makefile both here in the blt-1.9 directory and in
# the src subdirectory thereof. Look for the line beginning with
#
# DEFINES =
# # and add the -DNO_BGEXEC and -DNDEBUG flags after the = separated by a
# space from = and any other options.
make
cp src/libBLT.a ../linuxlib
cd ../linuxlib
ranlib libBLT.a
# Now you can go to linuxbin and rebuild the simulations. You might need
# edit the makefile to remove the -fenum-int-equiv from CFLAGS. It
# doesn't apply anymore.
cd ../linuxbin
make clean
make
Well, that worked for me. Let me know if it works for you.
Thanks,
-Barry.
And let me know if it doesn't!
-----RHS
postings log.
1.18 -- 04/06/04 -- Windows Installation
In response to a couple of questions concerning installation of SSS onto newer Windows systems, Barry writes:
On my XP pro system at home all the simulations work fine. The installer
runs fine, but seems to have the location of the start menu hard-wired --
C:\Windows\Start Menu\Programs\SSS. Both my home system upgraded to XP
from NT/2000 and my work system running 2000 are in C:\Winnt not
C:\Windows. But, simply moving the shortcuts from the above location into
the Start menu for All Users (Right click on the Start button as an
administrator and select explore all users) fixes this.
I don't know why the other installation failed to run. Perhaps the user
doing the installation was not a member of the Administrators group. I
have not yet installed the Simulations on my Laptop running XP Home.
Hopefully, I can get to this this afternoon. I expect the programs to run
fine as XP Pro and XP Home are the same basic operating system. The
installer, I'm not certain about. There are subtle permission and policy
differences between Home and Pro.
The simulations run fine under XP home. The installer ran with no problem,
but I work with an account that is in the administrators group. The start
menu did not get the SSS group. Again it created C:\windows\start
menu\programs\sss. One needs to move the sss directroy to the start menu
for all users. One way to do this is to right click on the start button
and select Explore all users and copy the sss directory above there.
-----RHS
postings log.
1.17 -- 09/13/02 -- Linux installation problems
If you're having problems with installing from the CD you could go to posting 1.14. I'd recommend instead, however, going to the Download page and downloading the patched linux version which should work. If it doesn't work either, then e-mail promptly to rhs1@cornell.edu and we'll see what we can figure out.
-----RHS
postings log.
1.16 -- 03/01/01 -- Updating SSS for Windows
I have the dos version which is working well with Win 98.
When I try to update the program, I get the error message :
Can not find SSS directory !
Even, if I browse to C:\program files\sss.
Got any solution ?
-----PT
Barry suggests two things.
1) You may be selecting the wrong directory in which to install the updating program: try one level higher.
2) If you copied the original programs from the CD, they will be write protected ("attributes" will show R). To remove the write protection, cd to the sss directory and try the command
attrib -R * /S
First see whether this does indeed remove the R attribute. If not, try to figure out the right way to do it. Once the write protection is removed, then see if the updater will work.
-----RHS
postings log.
1.15 -- 03/01/01 -- Source code availability
Dear SSS,
I have downloaded your trial version of
sss for a "testdrive" - neat stuff! A question though:
Does the source code for your programs
appear in either the CD-ROM, or in the appendices of
the hardcover version of "Solid State Simulations"?
I am not as much interested in exploring "LAUE'S"
explicit predictions as in studying it, and making use
of it as a starting point for my own simulation project.
Also, how thorough is the description of the physics behind
the model in the text?
Thanks in advance, from an experimentalist!
-----PP
The source code IS on the CD that comes with the hardcover edition, but is not in the appendices. You are welcome to do anything you like with it, short of selling it as your own. I would be interested to see what you develop.
At the end of each chapter there is a brief sketch of the algorithms used for that chapter. These should be helpful in figuring out what is going on in the source codes even though they are not very detailed.
The main text of each chapter is designed to develop a certain body of physical ideas using the simulations to illustrate these ideas. I.e., the simulations are to support the text (or better, the text is designed to help the student explore the physics illustrated by the simulations).
The text is NOT designed to support the development of the simulations. The development of simulation programs IS a good way to learn the relevant physics, but I deliberately steered away from that approach feeling that it would be too easy for students to become bogged down with programming issues and to lose sight of the physics.
-----RHS
postings log.
1.14 -- 02/20/01 -- Installing SSS on new Linux systems
We have a new Suse Linux 7.0, with Linux kernel "2.2.17-pre".
I have installed the linuxall.tar package. When I go into the
linuxbin directory and call any of the simulation programs, I get
a message like "bash: ./bloch: File or directory not found".
--FM--
{Try first going to posting 1.17.}
Here are the instructions I received from Barry Robinson who seems to know what's going on--
You can download any of the following four files by clicking on them:
The first file is the fixed version of tclPosixStr.c that should be placed
in the tcl7.6/generic directory after the tcl tar file is extracted. Users
can then build their own new libtcl7.6.a file. The tclPosixStr.c.diff
file
is a context diff showing the minor changes made.
The newtcllib.tar
file is an archive of one file:
./sss/linuxlib/libtcl7.6.a
Users can extract this at the top level of their sss installation, and
then a make all from the linuxbin directory will work.
The linuxbinupd.tar.gz
file contains the new tcl library and all 14
simulations. Again, extracting in the parent directory
of the sss directory is the correct procedure.
{{{If you're curious about what the problem was, the following from Barry, a preliminary fix to the problem, may help you more than it helps me:
The libtcl7.6.a library needs to be recompiled before you can recompile
the simulations. Something has changed in libX11.so. Simply follow the
instructions for Tcl in the README_COMPILE file. Unfortunately, you will
get an error in tclPosixStr.c complaining about a "duplicate case value".
This is because of the lines in the linux file /usr/include/bits/errno.h.
/* Linux has no ENOTSUP error code. */
# define ENOTSUP EOPNOTSUPP
Edit the tclPosixStr.c file to comment out the ENOTSUP cases, and tcl will
compile. Placing the new libtcl7.6.a in linuxlib fixes link problems in
the simulation compilation, and the new simulations will run.}}}
-----RHS
postings log.
1.13 -- 11/08/99-- Printing from the SSS programs
I am having difficulty printing output from the SSS programs. The programs
are installed on a PC running Windows on a LAN. Neither I nor our local
computer guru can figure it out without (according to the guru) having to
develop a special program. I should mention that I am using the original
version of the programs without upgrades.
Any advice or help would be most welcome.
--DLH
You didn't say what kind of output you wanted to print. At the moment I think of four possibilities:
1. pictures (i.e., the windows in which the simulations are displayed as moving electrons, atom arrangements, etc.): A variety of screen capture programs are available for Macintosh, Unix, and, I presume, Windows. (For the Mac, there is a screen capture as part of the OS.) You need to find and use one of these.
2. graphs: click the "copy graph" button on the graph you wish to print. The new graph window has a "save postscript ..." button which allows you to save the graph, as a file in postscript format, to a location you define in the dialog box which appears. In order to get this file to a printer, you can
download the freeware Ghostscript from www.Aladdin.com. (This worked for me on my Mac and presumably the Windows version also works.) Ghostscript will let you print the graph: simply open the file in Ghostscript and ask it to print it.
3. raw numerical data: again, click "copy graph" but now click the "save data ..." button to write the numerical data to a file which you can open with an appropriate word processing, editing or spreadsheet program.
4. the "band list" table in "bloch": I doubt there's an easy way to access these data. You can certainly use (1. pictures) above to copy the table to a printer.
5. see also earlier postings on the same question, with similar answers,
1.6, and 1.7.
-----RHS
postings log.
1.12 -- 11/08/99 -- Updating Windows Files to New Version
Have you had difficulties upgrading your SSS Windows software to the new version? We have modified slightly the updater to correct a spelling of "bravais" and to remove confusion about where to load the new files. (If all programs work except "bravais", find the file labeled "bravis.tcl", change it to "bravais.tcl", and use it to replace the old version of "bravais.tcl".) If the updater does not work, it's probably because the old SSS files are write protected. You must remove the write protection before trying to update the files.
-----RHS
postings log.
1.11 -- 6/23/99 -- False alarm on bugs
Delays in publication have given us the opportunity to submit a revised CD to Cambridge Press. This final version has had the bugs removed. The remedial files have been left on this site for those few who have downloaded an early version of the update which still contained the bugs. (Still uncorrected is the slowing of some Mac programs when the screen resolution is changed. A double-click on the "display" menu choice restores the normal speed.)
-----RHS
postings log.
1.10 -- 3/19/99 -- Windows Bug List--but see update 1.11
With the revision of the software which allows use with lower resolution screens we have also managed to add a few bugs into the programs.
The Windows bug list:
-
1) In "debye", you may receive an error message when you try to open the "configuration" window. To fix this problem download debye.tcl and replace the peierls.tcl file in your sss/lib/peierls directory with this corrected file.
This bug is on the CD with the revised program and on earlier versions of the updater which can be used to replace the original (high resolution only) version of the software with the newer version.
-
2) In "peierls", when you switch between "show holes" and "show electrons" the graph switching will not work and you will receive an error message. To fix this problem download
peierls.tcl
and replace the peierls.tcl file in your sss/lib/peierls directory with this corrected file.
This bug is on the CD with the revised program and on earlier versions of the updater which can be used to replace the original (high resolution only) version of the software with the newer version.
-
3) In "bloch" the program will crash if you ask for a band calculation with the "E-max" set to too low a value. If you are using the revised version (allowing low resolution) of the SSS, to fix this problem download bloch.exe and replace the bloch.exe file in your sss/winbin directory with this corrected file. If you wish to recompile the program, you should also replace the bloch.c file in the compile/sss/src/bloch directory with the corrected bloch.c file.
This bug is on the CD with the revised program and on earlier versions of the updater which could be used to replace the original (high resolution only) version of the software with the newer version. This bug is ALSO in the original version (Windows only) of the SSS software. We have not written a fix for the original version and suggest a full update to the newer version of the SSS if this bug is causing you problems.
-
4) None of these bugs should appear if you have used the current version of the Windows updater. The updater can also be used to correct the files from the revised CD, if you wish (but we have heard a few reports of troubles with this updater).
-----RHS
postings log.
1.9 -- 3/19/99 -- Unix Bug List --but see update 1.11
With the revision of the software which allows use with lower resolution screens we have also managed to add a few bugs into the programs.
The Unix bug list:
1) In "debye", you may receive an error message when you try to open the "configuration" window. To fix this problem download debye.tcl and replace the debye.tcl file in your sss/lib/debye directory with this corrected file.
2) In "peierls", when you switch between "show holes" and "show electrons" the graph switching will not work and you will receive an error message. To fix this problem download peierls.tcl and replace the peierls.tcl file in your sss/lib/peierls directory with this corrected file.
These bugs are on the CD with the revised program and possibly on early copies of the replacement files used to update the original (high resolution only) version of the software to the newer version.
4) None of these bugs should appear if you have updated your original SSS files with the current versions of the updated Unix files. These files may also be used to correct the files from the revised CD, if you wish.
-----RHS
postings log.
1.8 -- 3/19/99 -- Macintosh Bug List --but see update 1.11
With the revision of the software which allows use with lower resolution screens we have also managed to add a few bugs into the programs.
The Macintosh bug list:
-
1) In many of the programs, if you change screen resolution with the "display" menu, the programs will run more slowly. To restore the program speed, simply select the new resolution a second time.
This bug will always be with you.
-
2) In "debye", you may receive an error message when you try to open the "configuration" window. To fix this problem download debye.tcl and replace the debye.tcl file in your sss/lib/debye folder with this corrected file.
This bug is on the CD with the revised program and on earlier versions of SSSMacUpdate, the updater which replaces the original (high resolution only) version of the software with the newer version.)
-
3) In "peierls", when you switch between "show holes" and "show electrons" the graph switching will not work and you will receive an error message. To fix this problem download peierls.tcl and replace the peierls.tcl file in your sss/lib/peierls folder with this corrected file.
This bug is on the CD with the revised program and on earlier versions of SSSMacUpdate, the updater which replaces the original (high resolution only) version of the software with the newer version.)
-
4) In "drude", when you switch between the "mean value" and "standard deviation" graphs the switching will not work and you will receive an error message. To fix this problem download drude.tcl and replace the drude.tcl file in your sss/lib/drude folder with this corrected file.
This bug appeared only in an early version of the updater.
-
4) Only the first of these bugs should appear if you have used the current version of the Macintosh updater. The updater can also be used to correct the files from the revised CD, if you wish.
-----RHS
postings log.
1.7 -- 10/5/98 -- Printing Figures
I am a student at Georgetown University and we are using your Solid State
Simulations Program in our Solid State class. We had one question for
you. We are unable to print the images that are created by your program.
Is there some command that we could use?
-----KB
We couldn't figure a simple way to get printing capability into the programs that would work for all platforms, so didn't try.
If you want to print out graphs, they can be saved to your disk as postscript files. Use the "copy graph" command to get a new window with the graph. Then the "save postscript ..." command will save the graph as a postscript file. You then need a separate program which will allow you to print the postscript file.
If you wish to print one of the simulation displays, or want an alternative to the postscript route for the graphs, you need a screen capture program such as xv for Unix systems; I have an old share-ware program, Flash-it, for the Macintosh; I'm not familiar with what's currently available for the Mac or for PC's. These allow you to "capture" a selected portion of the monitor display, save it as a file, and print.
Find someone who's faster than I at finding things on the Web and you should be able to find the programs you need quickly and at very little cost.
-----RHS
postings log.
1.6 -- 5/29/98 -- Printing Postscript
Students should be able to use the software at home as well, but
many of them (mostly the students using Win95 or WinNT, which is the
majority) do not know how to print PostScript files. I would NOT suggest
to change the way SSS produces printouts, but a hint on the homepage and
the installation files would be helpful to this group of students. To my
knowledge, the GhostScript-Suite (http://www.cs.wisc.edu/~ghost/) is one
of the most complete PS-Interpreters which is freely available, but
there may be alternatives.
-----CG
Great! Indeed in many of the problems it would be useful for students to work with printouts of the graphs. Ghostscript gives a useful way to print graphs that have been saved as PostScript files using the "save postscript ..." button in a "copied graph" window.
-----RHS
postings log.
1.5 -- 3/31/98 -- Missing buttons and sliders
I obtained the CD and hardcover book in August. Everything worked fine
except "sommer", for which no dialog box appeared. After a long delay
I got a replacement from Cambridge Press. This time "sommer" worked once
but then the dialog box again disappeared, BOTH from the hard disk-installed
version and from the CD itself (!!!).
JF
I bought two days ago your package with CD and found that no presets were
available for "laue" and "sommerfeld". Is it that I got an old CD or
program hasn't been completed yet?
EN
___________________
In both instances, the problem was best characterized by the statement "the column of buttons and sliders between the two displays was missing." The problem is a monitor screen with inadequate (800x600) resolution. Increasing the resolution to 1024x768 ended the problem. If the tcl program can't find room on the screen to place all the widgets, it just leaves them out.
Similarly, it is possible to lose a button or slider at the bottom of the set of controls if there is inadequate vertical height on the screen. For some programs, using Windows, you may have to hide the task bar (is that what it's called?) to give room for the full program window.
RHS
postings log.
1.4 -- 7/25/97 (modified 8/21/97)-- Bug fix for Windows systems
We have found several minor problems with the SSS simulations on Windows systems. (Mac and Unix systems seem to be immune from these problems.)
A) If you attempt to copygraph a graph containing no data (why would you want to do that?) the program will crash.
B) The colors of some of the backgrounds and buttons can come out strange.
Fixes for these problems will soon be available for downloading from
sssfix1.zip.
C) On some machines, the graphs in the "additional exercises" get strange color mappings which can result in some lines not showing. If you have problems, try changing to a different color depth.
postings log.
1.3 -- 7/23/97 -- Screen resolution again
Although I understand the logic, it is unfortunate that so many of the
programs require the 1024 x 768 resolution. I teach all of my classes in
an electronic classroom equipped with a high-power Barco digital projector
--it projects on a 9 x 12 ft screen in a room large enough to seat 80.
These devices are still limited to VGA--640x480. Until new projectors with
the higher resolution are available (and we can afford to replace the ones
we have at $18K each), I will not be able to use your simulations for
in-class demos. This is a pity--but I see no way out.
___________________
Sorry about the screen resolution! We had hoped for 640x480 originally, but it's hard to fit that much information into that few pixels. If you display at that resolution, either the text is nearly unreadable or it all just doesn't fit on the screen. 800x600 works for many of the programs though not all. We're working on revisions to allow use on lower resolution screens, but make no promises about when it will get done.
With some computers you may be able to display part of the image. E.g., with my Mac, the program works perfectly well at 640x480, though I see only a portion of the display. By moving the window, I am able to access much, though not all, of the display. Also I've used an IBM notebook to drive a VGA display which could quite conveniently access the full area of the display by dragging the window, and which gave readable text. It worked quite well for lecture demos.
RHS
postings log.
1.2 -- 3/13/97 -- Screen Resolution
WARNING: roughly half of the SSS simulations require a screen resolution of 1024x768 or higher. The full display will not show on an 832x624 or 800x600 screen. On a Mac you can access the full display by scrolling so these programs are usable but not convenient. On Windows the missing portion of the display is inaccessible.
postings log.
SSS homepage--program list--postings log