Once you've finished a paper, you want to put it up on astro-ph. Many astronomers read the listings, so it's an important way to distribute your work. Information on how to submit is here. People have differing (strong) opinions about whether you should post on astro-ph when you submit your paper to a journal or when it's accepted. Remember that many people who look at your initial posting are unlikely to look again if you post a revision, so you want it to be right the first time around. Unless you have a good reason to post earlier, you should probably wait until the paper is accepted. Good reasons often involve the immediate fear of being scooped by competitors.

• Preprints formatted with, for example, \documentclass[12pt,preprint]{aastex} waste trees! You may write your paper using that format, but you don't want to post it to astro-ph that way. A good alternative is \documentclass[apj]{emulateapj}; you can get emulateapj.cls here. This puts your paper into a two-column ApJ-like format. If your computer chokes on this, ssh to meso and run latex again. You can also add \slugcomment{{\sc Accepted to ApJ:} August 1, 2006} under the \documentclass line for extra awesomeness. You'll have to look through your file to make sure that your equations don't run over now that the lines are shorter. If they do, you can break them up using the “eqnarray” environment or “align” environment (which requires the amsmath package, which you probably already have). For example:

\documentclass{emulateapj}
\slugcomment{{\sc Accepted to ApJ:} August 1, 2006} 
\usepackage{amsmath}
...
\begin{align}  % breaks up a single "long" equation into two lines, aligning on the left
& a = b \times c \nonumber \\  % don't number the first chunk of equation
& + d \times e
\end{align}
\begin{eqnarray} % breaks two equations onto two lines
a & = & f  \\
& \propto & g
\end{eqnarray}

You may also need to adjust the sizes of your figures. For example:

\begin{figure}      % use "figure*" instead of "figure" if you want your figure to span both columns
\epsscale{1.0}      % adjust this number to change the size of your figure
\plotone{f1.eps}
\caption{This figure rocks.}
\label{f1}
\end{figure}        % again use "figure*" instead of "figure" to span both columns

If necessary to place your figure perfectly, add commands like \hspace{-0.7in} and \vspace{0.3in} to add or subtract horizontal or vertical space (inside the figure environment). (It seems to me that these commands don't work continuously though; i.e., LaTeX won't add or subtract any arbitrary amount of space that you enter.)

• The astro-ph/new list shows papers in the order that they were submitted. The cut-off separating one day's listing from the next is 1pm PST. If you want your paper to show up at the top of the list, you should submit close to but after 1pm. But beware. If you submit too close to 1pm, the astro-ph script may still be running, and you could end up the last paper on the previous day's list, even though the timestamp for your submission is after 1pm. Even worse, if that happens, you won't have a chance to correct any errors you may have made in your title or abstract before they are posted on the “new” list. Generally, you can make any corrections on the same day that you submit with no penalty. If you wait until the next day, a new version of your paper is generated and both the old and new versions are made available. If you submit too close to 1pm, then any corrections you make will be the “next day.” So don't go overboard trying to time it closely.

• ArXiV.org (of which astro-ph is one section) has a ruthlessly enforced file size limit of 1 megabyte for your uploaded paper. It's quite easy to blow past this limit several times over if you've got a few figures, particularly high-resolution or just plain complicated observational data. Marshall Perrin wrote a Perl script which automatically converts your figures into compressed JPEGs to reduce file size. (It also does a few other things to ease paper submission, such as copying files to the ApJ-required file name conventions; see the comments in the script itself.) Mike Fitzgerald shamelessly ripped Marshall's figure compression code and incorporated it into his Python script, which performs similar functions.

• It's a bit messy to upload a bunch of files (the .tex file and figures). On a Mac, to create a single .zip file to upload, use File > Create Archive in the Finder window.

• Image files originally created in eps/ps/pdf often have many layers, which can lead to long load times (you've seen it – a plot that loads one. point. at. a. time.) A solution is to save your files initially as a png/jpg/tiff with a high resolution (600+ dpi) and then convert to eps/ps/pdf later. Alternatively, you can take the initial eps/ps/pdf, convert to a png/jpg/tiff (of high resolution), and then convert back to eps/ps/pdf. GS and OS X's convert do not handle these conversions well, but the program “sam2p” (available for OS X and Linux) does a very nice job.