Changes for users from xpa 1.0 and 2.0
This document describes changes that will affect users who migrate from \s-1XPA\s0 1.0 to \s-1XPA\s0 2.0.
There have been a few changes that affect users who upgrade \s-1XPA\s0 from version 1.0 to version 2.0. These changes are detailed below.
\s-1XPA\s0 commands no longer have a resolver routine (this is open to negotiations, but we decided the idea was dumb). For the SAOtng program, this means that you must explicitly specify the access point, i.e.,:
cat foo.fits | xpaset SAOtng fits
instead of: cat foo.fits | xpaset SAOtng
By default, xpaset, xpaget, etc. now wait for the server callback to complete; i.e., the old \-W is implied (and the switch is ignored). This allows support for better error handling. If you want xpaset, etc. to return before the callback is complete, use \-n switch: echo "file foo.fits" | xpaset \-n SAOtng
The old \-w switch in xpaset and xpaget is no longer necessary (and is ignored), since you can have more than one process communicating with an xpa access point at one time.
The new \-p switch on xpaset means you need not read from stdout: xpaset \-p SAOtng colormap I8 will send the paramlist to the SAOtng callback without reading from stdin.
See xpa(7) for a list of \s-1XPA\s0 help pages