LaTeX

• LaTeX help is available at
• SIPB's Inessential LaTeX
• IS's Essential LaTeX
• IS's Advanced LaTeX
• The AI Lab has their own LaTeX page
• Common LaTeX questions and answers
• Separate multiple authors in the \author command with \and; to include other info for each author (email, institution, etc.), use \\ to create separate lines for each author. Thus the command \author{Alice \\ alice@mit.edu \and Bob \\ bob@mit.edu} will give you something nice.
• If lots of graphics are slowing you down, you can use the draft option to the document class statement to force latex to create files with placeholders for graphics. This should make (pdf)latex run much faster. Like this: \documentclass[11pt, draft]{article}
• To convert your latex docs into HTML, use latex2html. I personally use it with these options: latex2html -split 0 -info 0 -nonavigation
• Use pdflatex to generate PDFs directly without bothering to make DVI files. The result will use smooth vector fonts and look very pretty.
• You can fill out the Document Info fields by using the command  \pdfinfo{/Title (xxx) /Author (yyy) /Subject (zzz)}
• To make nice PDFs, you should add this after all other packages. It makes all internal document references links. \usepackage[colorlinks=true]{hyperref}
• You can use the \url{http://web.mit.edu/} command to make links that open in a web browser
• To use pdflatex, your images must be either PNG, GIF, JPEG, or PDF. You can easily turn EPS files into PDFs by using the epstopdf command. There's also a package that will do this automatically on demand, but I've never bothered with it. To convert postscript files (for example, those made by graphviz), first convert the ps file to eps using ps2epsi and then run epstopdf on the result. If you use the graphicx package, and you don't specify graphics file extensions  in your document, your tex file will work with both latex and pdflatex (provided both eps and pdf versions of the same file are available).
• If you grab the thumbpdf package from CTAN, you'll be able to generate nifty thumbnails for your PDF documents. Here's a man page.
• Source Code listings
• You have two choices: lgrind and the listings packages. The listings packages support more languages (like VHDL), but I haven't been able to get them to work.
• To use lgrind, add the newtex locker and run lgrind -s. This should give you a list of languages supported by lgrind. Note that Athena's version is old and if you can't find your favourite language, you may be able to get a newer lgrind language definition file from the net. Given a source file foo.c, you should run lgrind -lc -i foo.c > foo.tex. The -lc switch specifies the C language while -i tells lgrind to make a file for inclusion in a larger document rather than a standalone document.
• You use lgrind files by including \usepackage[nolineno]{lgrind} in your documents preamble (I don't like line numbers in my listings, so I kill them with the nolineno option). You can include individual listings made with lgrind by using the \lgrindfile{foo.tex} command.
• Bibliography Stuff
• Wherever you want your bibliography to appear include the commands \bibliography{foo} and \bibliographystyle{plain} where foo is the name of your BibTeX file without extension.
• If you don't cite something but you still want it in your bibliography, put the command \nocite{ref1, ref2} somewhere in your document.
• BibTeX likes to capitalize things on its own, sometimes doing the wrong thing. If you want a word or phrase capitalized according to how you wrote it in the .bib file, just surround the phrase with curly braces ({}). If you're really lazy, you can write things like title = "{Some Title goes Here}".

AFS

You'll need to get OpenAFS packages for your system. I got openafs, openafs-client, openafs-kernel, openafs-kpasswd, openafs-krb5, and openafs-compat. AFS requires a kernel module; if you've modified your kernel, you may need to compile the openafs module from source to match your kernel. You'll also need to steal the AFS config files from an Athena machine (the dialups should work fine). Once you get AFS working without kerberos, you should be able to cd into /afs/athena.mit.edu/user/.... and at least see your home directory. Since you don't have kerberos setup yet, you shouldn't be able to actually read or write anything, but you should at least be able to list directories. Once that's working, you can install kerberos. I grabbed packages krb5-libs, krb5-workstation, and krbafs. You should steal the appropriate kerberos config files from an Athena machine. Try getting tickets by running kinit. You can see what tickets you have by running klist. Once you have tickets, you'll need to run aklog in order to use your tickets to authenticate yourself to AFS.
 

Graphviz

E-mail/Web Browsing

• You can check your email from anywhere with a web browser using webmail.
• You can manage your list subscriptions at MIT over the web instead of using moira.
• You can login to a dialup machine from any box that has a web browser and java using the mindterm ssh client.
• For terminal mode email checking, you can use mutt. mutt lives in the sipb locker.
• For graphical email, you can use evolution, a very cool client that offers lightning fast virtual folders for organizing mail. It lives in the evolution locker and is still somewhat buggy (at least the one installed on athena is; my local copy doesn't have any major problems).
• Mozilla is on Athena as well; it's a big improvement over netscape 4. It lives in the mozilla locker; the first thing you should do is tell it to use the modern theme rather than the netscape classic one...it's much prettier. Also, you can prevent pop up ads from working by deselecting "Open unrequested windows" in the Scripts section of the Advanced preferences (this seems to make the MIT libraries site slightly nonfunctional though...). You can also block many advertisements by going into your .mozilla directory and replacing the file cookperm.txt with my cookperm.txt. Note that you may have to go a couple of layers deep to find it.
• MIT in it's infinite wisdom won't issue client side certificates to you unless you're using netscape 4.x or IE. Since both of these suck, I use neither, opting instead for Mozilla and Konquerer. To get your client side certs into alternative browsers, just get the certs using netscape and then use netscape to export the certificates to a file. Then import the certs into your nonsucky browser and enjoy.