Users of the SCRC analysis software have many ways of running this software,
on many different platforms.
Each different way of running the software can present its own unique set
of wrinkles or challenges, as something that has become familiar in one
environment can be strangely different, or not work at all, in another.
This guide should help you work through these various quirks, by explaining
each of the platforms on which our software runs, and what potential pitfalls
you may encounter there.
Please help us keep this tutorial up to date.
If you run into any other quirks or problems that are not explained here,
please let us know so we can include a description and fix or workaround.
Of all the
supported operating systems
on which the analysis software runs, the Linux operating system
(and especially Red Hat derived distributions) is the primary platform.
Although there are many different flavours of Linux around, and several
desktop environments available under Linux, all with their own unique
peculiarities, most of these are almost inconsequential to how the
analysis software runs.
This software uses only the most basic X Window System graphics, which
all of the various Linux distributions and desktops provide, so running
the analysis software is almost the same on all of these systems.
Neuro icon on Linux desktop
Gnome top panel menu and launch bar, with Terminal icon (3rd from left after System menu)
Once logged into Linux, you can launch the terminal application from
the the icon on the top panel menu bar, and use the command line from the
terminal window.
Alternatively, if you prefer using the Neuro front-end program, you can
double-click its icon on the desktop to launch it.
Most Linux systems will be equiped with a 3-button or wheel mouse, so
using these should be straightforward as all
mouse buttons
should work as expected.
The analysis software runs on Mac OS X, but doesn't make use of the native
"Cocoa" user interface of OS X.
Instead, it runs under the X11 system, which is available on Mac OS X 10.3 and
later, or as a beta verson for 10.2.8.
The X11 application must be running for the analysis software to be able to
open its windows.
Also, the analysis software is not installed in the standard directories that
are part of the default execution path, so a little bit of setup of environment
variables is needed to run this software.
For these reasons, there is a command script, Neuro.command, included with
the Mac OS X version of the software to automate the startup process.
Neuro.command script in /Applications folder
You can find Neuro.command in the /Applications folder of your Mac,
but for convenience you may want to copy it to your Desktop, for easy access.
Unfortunately, command scripts can't run from the application dock on OS X.
Double-click on Neuro.command to start it up.
It will make sure the X11 application is installed and running, and will
start it up if it isn't already running.
It will also start the neuro front-end program with the correct
command execution path and other environment variables set to allow the
analysis software to run.
Any other application launched by X11, including xterm, should also have all
the right environment variables set if X11 was launched by the Neuro.command
script.
This script also adds an item to neuro's menu to set the default
printer, using the Print Center application.
If you'd rather run the analysis software from OS X's Terminal application,
you can do that (as long as you've launched the X11 applications)
if you add these two lines to the .profile file in
your home directory:
export DISPLAY=:0.0
. /usr/neuro/lib/setup.sh
If you use a 3-button mouse, wheel mouse or Apple's "Mighty Mouse" on your Mac,
using it should be straightforward as all mouse buttons
should work as expected.
But if you're using a single button mouse or trackpad, be sure to read about
Using the Mouse Buttons
so you can access all the choices that the analysis software allows for
mouse input.
Any USB mouse should be plug-and-play under Mac OS X, so pick one that will
make your life easier.
It is often handy to access a Linux-based data capture or analysis system
remotely, via a networked X Window session, rather than logging in directly
to the system's own console.
This allows you to work from your own desktop or laptop system.
Here at the University of Manitoba, we use
StarNet's X-Win32 package,
as we have a site license for it.
Other X Window packages
are available for Windows systems, and are included with Mac and Linux systems.
One free alternative to X-Win32 for Windows systems is
Xming+PuTTY on a Stick.
Task bar icon and session menu, X-Win32 version 7 or older
Task bar icon and session menu, X-Win32 version 8 or later
To launch an X session with X-Win32, first make sure it is running.
You should have an X icon in your task bar if it is.
If not, start X-Win32 from the Start menu, or desktop or quick-launch icon.
Then, click on the X icon in the task bar to get X-Win32's session
menu, and from there, select the system to which you wish to login.
If the system isn't listed there, you can add one by right-clicking on
the icon and selecting X-Config from the menu.
In the X-Config window, there is a Wizard that guides you through adding
sessions to your list.
On the local network, you can use the rexec connection method for
launching sessions, but for accessing our servers from off-campus (e.g.
from a laptop system or home PC), you should use SSH instead.
Once you've opened a session, X-Win32 may or may not ask you for a password,
depending on whether you've saved the password in the session profile.
After that, an xterm window should open, allowing you to enter commands.
Using a remote X session, the mouse buttons should work as expected,
though it's likely that Windows won't support the middle button transparently,
so you may need to use the technique described in the section on
Using the Mouse Buttons below.
Any X Window applications that you run from the xterm window, including
most programs in the analysis software suite, should open their own windows
right on your desktop, co-existing with any Windows application windows you
may have running.
To start a remote X Window session from a Mac or another Linux system,
you need to run the ssh command from your xterm window (or any
terminal window on a Linux system), giving the arguments -X
followed by the name of the system to which you want to connect.
For example:
ssh -X dave2.scrc.umanitoba.ca
Note that when using rexec connection method for launching sessions,
your password is sent unencrypted over the network.
This method should only be used on trusted networks.
See the section on
password security
below for details.
The SCRC's analysis server, dave2, as well as one of our capture systems,
are set up to allow connections from the Windows application
Remote Desktop Connection.
This service is provided on the Linux side via a server called
xrdp, which could easily be set up on any other Linux capture or
analysis system as needed.
When you start Remote Desktop Connection, which is usually under
Start->All Programs->Accessories,
it will ask for the computer to which you wish to connect.
You should enter the full host name and domain, e.g.
dave2.scrc.umanitoba.ca.
The Vista version of this program will give you a warning about not being
able to confirm the identity of the computer you want to connect to:
just click Yes.
When the connection in made to the server, you will need to login by providing
the user name and password.
Once logged in, you may get another warning from
KDEInit that it couldn't launch /usr/bin/autorun, which you can ignore
as well.
Note that when connecting to our servers using Remote Desktop,
your password is sent unencrypted over the network.
This method should only be used on trusted networks.
See the section on
password security
below for details.
A version of Remote Desktop Connection is also available for the Mac,
from the microsoft.com web site's
Mactopia.
On Linux systems, you can use the rdesktop application, or the
front-end tsclient launcher.
KDE's panel, showing KDE start button and Terminal icons
Using a Remote Desktop session, the mouse buttons should work as expected,
though it's likely that Windows won't support the middle button transparently,
so you may need to use the technique described in the section on
Using the Mouse Buttons below.
The other thing that will be different from a usual Linux session is that
you will be using KDE as the desktop manager, rather than the usual default
Gnome desktop, but that will have no impact on the analysis software.
You can launch KDE's Terminal program by clicking on the Terminal icon
on the bottom panel.
Unlike running X-Win32, however, everything will run within Remote Desktop's
window, with your other Windows applications being concealled, rather than
X application windows and Windows application windows co-existing on the same
desktop.
When you're finished with your connection, you should logout
using the KDE start button ("K" over a gear) on the bottom left,
rather than just closing the RDP session without logging out (using the X
button on the top bar).
Combining Remote Desktop and X-Win32:
It is also possible to use Remote Desktop in combination with X-Win32 or
some other Windows-based X Window server.
This can come in handy when you need to connect to a Linux system that
doesn't directly support Remote Desktop through an xrdp server.
Just start up Remote Desktop Connection, as above, but connect to the
desktop PC in your office, by its host name or IP address.
From that remote desktop session, you can open
X-Win32 to begin a remote X Window session, as
described in the previous session.
Note that when going through two layers of systems in this way, there is
an even higher likelyhood that not all
mouse buttons will go straight through to the
analysis program intact, so you may need to apply some of the tricks described
in section 8 below.
The SCRC's analysis server, dave2, has an archive of captured data from our
electrophysiology labs.
In addition to accessing this archive via Remote Desktop or remote X Window
sessions, as described above, it can also be accessed via a web browser.
This web-based interface to the archives is usually accessed by clicking
the View button in our
FileMaker Experiment Database,
but it can also be accessed directly by a
Login to the Data Archives Server.
Either way, you will need to provide your login user name and password.
Note that when you login to the web server, using an "http:" URL rather
than "https:", your password will be sent unencrypted over the network.
This method should only be used on trusted networks.
See the section on
password security
below for details.
As much as possible, you should use HTTPS logins to our server,
for secure, SSL-encrypted data transfers between your browser and our system.
In the future we may force non-encrypted HTTP logins to redirect to HTTPS.
You will need to tell your browser to accept the server's SSL certificate
the first time you connect using HTTPS.
Browsing the data archives is done through the showrun script on the
analysis server, the same web interface as browsing the
Sample Data to View/Download
on our main web server.
When browsing an experiment directory, you will get a directory listing as
shown in this figure:
Experiment directory listing
When browsing a run file within an experiment, you will see a summary page that
shows a thumbnail image of the data in that run, as well as a listing of
traces and waveforms in the run, and several other capture parameters.
This summary page appears much like the one shown in this figure:
Experiment data display
The web interface can also accept analysis parameters as part of a
query string at the end of the URL, or web address, of the summary page.
This query string begins with a question mark, "?", after the path to
the data file, and consists of query parameters in the form
"name=value", with two or more of these parameters
being separated by an ampersand, "&", or semicolon, ";".
The most important one of these query parameters is aparm,
which gives a series of analysis commands (keystrokes) for setting various
analysis parameters.
This aparm value can also be set by modifying the parameter fields
in the FileMaker database.
A number of characters have a special meaning when used within the
aparm value:
Character
Meaning
~
Escape, i.e. go to top level menu
!
Enter and Escape, useful for entering a parameter value
+
Space
%0A
Enter
%xx
Hexadecimal encoded ASCII character, where xx are 2 hex digits
Data display with analysis parameters
Other query parameters can also be set in the URL's query string, such as
width, height and out.
The FileMaker database will include width and height values in the query
string it sends to the web browser, to get a larger image displayed in the
summary page.
These width and height are specified in pixels.
The out query parameter is typically used internally by some of the
links that the showrun script generates, but its purpose is to specify an
alternate image format, one of gif (the default), png, or svg, or to save
the plot in plt or pdf format.
Additionally, a vnc query parameter is used internally
to launch a Java VNC session, as described below.
In addition to the image of the data and the listings of traces, waveforms,
and capture parameters, the summary page also presents a number of links to
click for additional information.
If the experiment protocol for the currently viewed experiment is online,
there will be a link to the PDF file for it.
There are also links to download the current run's data files as a ZIP file
or compressed TAR archive (tgz), or the current image as a vector graphic
in SVG, HPGL (.plt) or PDF format.
Finally, there are two links that allow further analysis of the current run.
The first of these, "Analyze this!", will attempt to open an X Window
session on the PC at your IP address.
This works if you are on the SCRC's local network and have the X-Win32
software running and configured on your system for rexec access to the
analysis server.
It also works for SSH sessions, provided you have a currently running session
on the analysis server.
The showrun script tries to detect which method is being used to access
the server from your PC's IP address.
Using the analysis software in this way is just as described above
on X-Win32 or other remote X sessions.
In cases where this doesn't work, for instance if you are off-campus with no
ssh session running, or on
a PC with no X Window software installed, the second link,
"Or analyze via Java-VNC", will open a Java-based VNC viewer applet
in your web browser, to begin a VNC session as described
below.
Either of these analysis links will open up an analysis program window with
the run you were viewing as the current run to be analyzed.
If analysis parameters were specified via aparm in showrun's query
string, these same parameters will be set for you in the analysis window.
In order to truly allow SCRC users to analyze anywhere, we've recently
added a VNC server to our analysis server, dave2.
This allows starting a VNC session within your web browser to login to the
analysis server, without needing to install any special software on your PC.
All it requires is a current version of Sun's
Java virtual machine installed as a plug-in
for your favorite web browser.
While you could use any VNC client to access the server, or load the Java-VNC
applet by requesting it directly from our server, the most convenient way to
get the process started is by browsing the data archives
on the web, as described above, and using the link on
the summary page to analyze via Java-VNC.
This will typically open another tab in your web browser to load the Java
VNC viewer applet, which will then open a series of new windows for the VNC
session.
The first of these, shown in the figure to the right, request the connection
details, specifically the server name and port number.
The default values are filled in automatically by the showrun script, so
generally you just need to click OK.
The next window will be a login window, that request the password to login
as the user "exp".
Finally, the viewer opens a window running a KDE session on the analysis
server.
Note that when you login to the VNC server,
your password is sent unencrypted over the network.
This method should only be used on trusted networks.
See the section on
password security
below for details.
Java VNC browser window running analysis
Using analysis in this way is very much like the KDE session you'll get when
you login to the analysis server using
Remote Desktop.
There are, however, a few important differences:
The VNC sessions running on the analysis server are persistent.
The server starts up 8 of these, and the VNC client connects to one of them
(the showrun script picks one that is not currently connected to a VNC viewer).
When you close your VNC viewer, any programs you left running will still be
running, so it is a good idea to close any windows you open in a VNC session,
including the analysis window that showrun opens for you.
Otherwise, these windows will still be around when you, or another user,
starts up another VNC viewer which connects to the session you left behind.
Unlike with a Remote Desktop session, however, you do not need to logout using
the KDE start button.
The VNC viewer will usually have a very limited colour palette, so some
KDE applications, icons and window decorations will appear a little odd.
This should not affect the analysis program, though, as it uses very few
colours anyway.
An odd quirk with the Java VNC viewer is that it seems to ignore the
Tab key.
If you need to enter a Tab character in any application, you will need to use
Ctrl-I instead.
When you are done with your VNC session, and have closed all your application
windows, all you need to do is click the close button on the VNC viewer window,
then close the tab that opened in your browser, and you should be back
at the summary page for the run you were viewing.
You do not need to logout from the KDE session running on the VNC server.
A note about installing and updating Java:
After downloading and installing Sun's
Java virtual machine,
it is important that you keep it updated.
It is a complex piece of software, and patches for security vulnerabilities
are added to it fairly frequently.
Fortunately, current versions include an automatic updater, which will
periodically check for new updates, and notify you when they should be
installed.
As of this writing, the current release is Version 6 Update 13.
If you have an older version, and it is not automatically updating itself,
you should uninstall it and get the latest version from their web site.
When you install an update, or during the initial installation,
the installation software will commonly offer to install some other
application as well, such as OpenOffice.org, or the Yahoo toolbar.
The checkbox for these addtional software packages is usually checked by
default, so be sure to look for this and uncheck it unless you really want
the additional software installed.
7. Using Cygwin/X, andLinux, or other Virtual Machines under Windows
Because the analysis software is written for
a UNIX-like operating system, a strictly native Windows implementation of
the software would be quite difficult to develop.
However, it is possible to run it under the
Cygwin/X
system, which emulates a UNIX-like environment under Windows.
Cygwin/X is a port of the X Window System to Microsoft's series of Windows
operating systems, including an X Window server, libraries and many of the
standard X applications.
It is built upon the Cygwin project, which provides a UNIX-like set of
libraries and utilities which ease the porting of UNIX/Linux applications.
A Cygwin/X installation requires about 500 MB of disk space, but its
memory requirements are fairly minimal.
Cygwin/X should run well on 32-bit versions of Windows XP and Vista, as well
as the 64-bit version of Vista (though we have yet to test this ourselves).
Installing Cygwin/X is fairly straightforward.
Begin by downloading the installer: go to
http://x.cygwin.com/, click
on "Install Cygwin/X now", then save the "setup.exe" download to a file.
It's a good idea to rename it to "cygwinsetup.exe" so that you'll know
what it's supposed to setup when you come back to it later.
You should review
the
installation section of the Cygwin/X User's Guide
to see what the installation involves.
Follow the steps below, which include some additional packages to install.
Run the installer and click on Next to begin installation.
Select "Install from Internet" as the installation type and click Next.
Stay with the default Root Installation Directory, C:\cygwin, and the
default installation for "All Users" (for which you need
administrative privileges, otherwise choose "Just Me").
Keep the Default Text File Type as "Unix / binary", and click Next again.
You then select the Local Package Directory, as illustrated in
step 9 of the
Cygwin/X
User's Guide installation instructions.
The choice isn't critical as the setup program will find this again
in the future, but a good choice that can be found easily from either
Cygwin or the Windows Explorer is "C:\cygwin\download".
Click Next again, and the installer asks you to select your
Internet connection type.
In most cases the default, Direct Connection, is what you should use.
Click Next again to proceed.
It asks you to choose a Download Site from a list of Cygwin mirror sites.
This is step 13 in the User's Guide installation instructions.
The best choice for North American university sites is probably
http://mirror.cpsc.ucalgary.ca, about halfway down the list, as they
have a high speed link to CA*net 4/NSFnet, as do most universities.
Click Next after choosing the site.
You will now have to select the packages to install.
There are a huge number to choose from, but only a handful are really needed.
This is shown in step 15 in the
Cygwin/X
User's Guide installation instructions.
It's a good idea to maximize the setup window at this point to see
everything more clearly. Packages are selected by clicking once on the
word "Skip" to the left of the name of the package you want. The word
"Skip" will then be replaced by the version number of the package that
will be downloaded. The setup program will also automatically select any
packages that are required by the one you selected, so by picking only
a few packages, you can select the whole system you need. Packages are
organized into categories to make them a bit easier to find, without
having to wade through a list of all packages (though you can do that
if you want, by clicking on the View button until it shows a Full list
of packages, or a list of those Not Installed). In the Category view,
click on the "+" sign next to a category to expand the list of packages
in it. As the User's Guide recommends, you should install xorg-server
and xinit from the X11 category, but for purposes of running the SCRC
Analysis Software package, you should install all of these:
byacc, gcc, libncurses-devel and make in category Devel
ed and vim, in category Editors
netpbm in category Graphics
openssh in category Net
bc, file and ncurses, in category Utils
wget in category Web
libX11-devel, xinit, xorg-server and xpr in category X11
Cygwin/X's XWin server menu
These, and the packages they require (which are automatically seleted),
should be sufficient. Then click Next.
The setup program will then download all required packages, and
after they're downloaded it will install them. When it's finished, it
will give you the option of adding icons to the Desktop and Start menu
(checked by default). The Desktop icon is optional, because you don't
use that one for running X Window applications, but you should install
the Start menu icons. Click the Finish button to complete this phase
of the installation.
In summary, all the default choices are OK, except that you must select
a download site and the packages you need, as there are no defaults for
these, and you can pick a better location than the default one
for storing the downloaded packages.
After that, Cygwin/X is ready to use, and the analysis software
can be installed from a .tgz package that we provide.
Printing support for the analysis software can be set up by installing
the GSPRINT
software under Windows.
To start up Cygwin/X, click the Windows Start button, select All Programs,
select Cygwin-X, then select "XWin Server".
The Cygwin "X" icon should appear in the task bar, and right-clicking it
will show a small menu. There should also be an "xterm" window that opens
up automatically when you launch Cygwin/X, and you can start one any time by
right-clicking on the X in the task bar, selecting Applications, then "xterm".
It is also possible to run a supported Linux system under Windows XP or
Vista using various
virtualization
techniques.
Of those we've tested so far, the most promising is
andLinux
Beta 1, which runs a minimal Ubuntu 7.10 (Gutsy) system on a custom
coLinux
kernel running as a Windows system process, and displays using the
Xming
X11 server for Windows.
An andLinux installation requires at least 2.5 GB of disk space (for the
minimal XFCE version), and will continually tie up about 200 MB (minimum)
of RAM for the Linux system, so you should have at least 1 GB of system
RAM (2 GB recommended) on your Windows machine to run this.
Note however that andLinux can only run on 32-bit versions of Windows
XP and Vista, not the 64-bit versions.
Installing andLinux is also quite easy: just go to the
andLinux.org web site,
select their Downloads link, and
get the minimal/XFCE version (200 MB for Beta 2, 131 MB for Beta 1)
from the SourceForge.net http link.
Save the installer file, currently andlinux-beta2-minimal.exe or
andlinux-beta1-minimal.exe, and then double-click on it to run it.
Most of the defaults can be used as-is without changing anything.
The only exceptions are:
1) For Beta 2, you must provide a username and password for the andLinux
account that will be created during installation (Beta 1 uses the "root"
account),
2) we recommend that you select CoFS rather than Samba for Windows file access,
to avoid the need to set up Windows File Sharing and Samba, and
3) you don't need the desktop and quick-launch icons which are installed by
default, so you may uncheck these options during the Additional Tasks part
of the installation.
After that and a quick reboot, andLinux is ready to use.
The analysis software can be installed from a .deb package that we provide.
Using andLinux is also very easy.
There is a little "yin/yang" black and white icon in the task bar on the
lower-right, on which you click to get the andLinux menu.
From there, you can launch the Terminal application,
or other Linux applications.
andLinux Menu (Xfce version)
There are two little quirks you should know about.
First of all, under andLinux, the Linux kernel - and therefore the rest of
the Linux operating system - runs as a Windows service.
This service may take a while to start up after you restart Windows,
as Linux effectively reboots after Windows does, while other Windows
tasks are starting up, and Linux will not be ready to launch applications
from the andLinux menu until it has had a chance to fully boot up.
So, if you get an error when you try to start Terminal, just wait a little
while and try again.
Secondly, after you launch Terminal or other applications from the andLinux
menu, you should click on the window that opens up, to bring it to the
foreground and activate it so it accepts keyboard input.
For some reason, applications launched by the andLinux menu don't activate
themselves on their own, so you have to click on them before you start typing.
Similarly under Cygwin/X, occasionally when you start up the XWin server,
it's slow to start up so the xterm application times out and doesn't appear.
If that happens, just start xterm from the X menu on the tray icon.
You will usually need to click on windows to activate them as well, as
for andLinux/Xming above.
The whole andLinux system runs from a small "virtual disk drive" which
is stored as a file on your Windows C: drive. Because there's not a lot
of free space on this virtual drive, you will most likely store your
analysis data elsewhere, in other folders on your C: drive. This of
course raises the question of how you can get at these other folders.
Under andLinux, your Windows files appear under the mount point
/mnt/win.
During the installation, if you chose the default of making all
of C:\ accessible, then /mnt/win will correspond
to the root folder of your Windows C: drive.
In Windows Vista, you can get the andLinux terminal to go to your
own Documents folder using this command:
cd /mnt/win/Users/yourname/Documents
where "yourname" is your Windows logon ID. In Windows XP, it would be
cd /mnt/win/Documents\ and\ Settings/yourname/My\ Documents
For Cygwin/X systems, the C: drive is accessed using the UNIX-like
path /cygdrive/c, rather than /mnt/win as for andLinux above.
Cygwin doesn't use a virtual disk, though, but rather stores its files
under the C:\cygwin directory by default, so your Cygwin home directory
will be found under Windows using the pathname
C:\cygwin\home\yourname.
The choice is yours whether you prefer to keep your files under your Cygwin
home directory, your Windows Documents directory, or elsewhere for that matter.
Note two important points.
Under Linux, it is the forward slash (/)
that is used to separate directory names in a file path name, rather
than the backslash (\) as you'd typically see under Windows.
But because you're running UNIX/Linux commands, you must follow the UNIX
convention for pathname separators.
If any file name contains spaces, these must
be quoted on the command line so the whole filename is taken as a single
command line argument. So, you can precede each space by a backslash as
shown above, or you can surround the entire file or directory pathname
in single or double quotes. E.g.:
cd "/mnt/win/Documents and Settings/yourname/My Documents"or cd "/cygdrive/c/Documents and Settings/yourname/My Documents"
Apart from these important differences, though, the use of the analysis
software under andLinux is very much the same as under any other Linux
system, using the same commands and the same techniques to navigate and
manage directories and files.
Because andLinux displays using the
Xming
X11 server for Windows,
and Cygwin/X uses a similar X11 server for Windows,
using the analysis software under either will
be very much like running it on
X-Win32 or other remote X sessions,
and the same caveats will apply regarding mouse button usage.
Indeed, you can use this setup to launch remote X sessions to other
Linux servers, by using an "ssh" command as described above.
These ssh commands may even be added to the andLinux launcher menu,
or Cygwin's XWin menu, for quick, easy and secure remote sessions.
The mouse operations in the analysis software depend on single clicks
on a 3 button mouse - no click and drag or double-click operations.
But because this software can run on so many different platforms, this
raises the question of what to do when you don't have a 3 button mouse,
or when not all mouse buttons work as expected.
First, the simple case: on most Linux systems today, you will typically have
a 3 button mouse or 2 button wheel mouse.
When the analysis software expects mouse input, it will typically show
a menu that describe the function of 3 or more buttons.
The first 3, A, B and C, correspond to the 3 buttons on the mouse, from left to
right.
On a wheel mouse, the scroll wheel replaces button B, so pressing down on it
is the same as clicking the middle button on a 3 button mouse.
Buttons A, B and C go from left to right on a 3 button mouse.
On a wheel mouse, press the wheel to click button B.
This all works well in most situations, but in some cases not all of the
mouse buttons work as expected.
This is especially true when running under
X-Win32 or other remote X sessions,
where the middle button or wheel click may not go through to the X Window
application as a middle-click.
Some mouse drivers either don't support the middle button, or reassign
the middle button to some purpose, such as a simulated double-click of the
left button.
If the mouse driver doesn't make it easy to set the middle button to a
normal, middle-button-click operation, or you prefer to keep the reassigned
operation for your other applications, then you need an alternate way of
getting the middle-click, or button B function in the analysis software.
The most reliable way of doing this is
Shift-left-click,
i.e. hold down the Shift key and click on the left mouse button.
Similarly, Ctrl-left-click will get you button C if a
right-click doesn't work, if the right button is reassigned, or if you have
a one-button mouse (e.g. an older Apple mouse) or one-button trackpad.
On some systems, simultaneously pressing the left and right buttons simulates
a middle-click, but that does not always work reliably, or sometimes not at all.
Also, pressing the A, B or C key on the keyboard
may do the trick as well, but not equally in all analysis programs (it won't
work in qm, for example).
So, your best bet is to use Shift-left-click for B if the
middle button doesn't work, and Ctrl-left-click for C if the
right button doesn't work.
Because many of the methods of connecting to the SCRC's analysis server,
as described above, send your password unencrypted over the network,
it is important to consider more secure alternatives.
Methods such as
X Window sessions launched via rexec,
Remote Desktop,
VNC
and
web browser access via http
should only be used on trusted networks, where you do not need to be
overly concerned with passwords sent as clear-text being intercepted.
On unsecured public networks, whether wired or wireless, you should not
use these methods to access the server, as your password could easily be
captured by someone operating a
Packet analyzer or
Man-in-the-middle attack.
The only secure method of connecting to our servers, when you can't be
certain that the network you're on is safe, is to launch an
X Window session via SSH as described above.
SSH will create an encrypted "tunnel" through which all your network
traffic for the X Window session is channeled, including your user name and
password, so these can't be intercepted.
It also will verify the server's encryption key to make sure there isn't
another system attempting to masquerade as one of ours.
If you're planning on travelling and think you may need to access our systems
from public networks, you should make sure that your laptop computer is
set up ahead of time for secure access via SSH.
It doesn't take very long to set up, and once it's done, you can have the
peace of mind of knowing that access to our systems is convenient and safe.
With a bit of planning, you really can
analyze anywhere.