We can also exert control over how <fetch> behaves while it is obtaining a document, with <urlcp> :

• <urlcp nbsp;timeout nbsp;10>
  This sets the fetch timeout to 10 seconds (the default is 30): if the document is not completely returned by then, the fetch is abandoned and a partial document (if any) is returned with an error. This provides a way to avoid holding up an entire script for a slow server. Note that this is not the overall script <TIMEOUT> : the Vortex script will not exit at this timeout, only <fetch> . Thus this timeout should always be less than the script timeout.

• <urlcp nbsp;maxpgsize nbsp;2MB>
  This will allow up to 2 megabytes of raw data to be returned per page before the fetch is abandoned with an error. (Early versions of Vortex may require an explicit number.)

• <urlcp nbsp;user nbsp;bob>
  This sets the user to use for fetching protected documents. Similarly, the password can be set with pass . (This is not the Texis user/password for SQL.)

• <urlcp nbsp;proxy nbsp;"http://proxy.mysite.com/">
  This will set the proxy server to use for fetches. Instead of contacting Web servers directly, <fetch> will use this host as a proxy server to obtain the documents: useful if you're behind a firewall and need to go through a proxy server.

See the manual on <urlcp> for more options.