`file-attributes' versus large inode numbers.

Hi. As part of the work I am doing on TRAMP[1], I am trying to address the issues of large inode numbers returned from `file-attributes'. On an SGI machine, the inode of a file is `1186988379'. This overflows the standard int representation on GNU Emacs and all. Now, they deal with the issue of this by allowing the inode value to be a cons cell, similar to the various access times. This makes some sense, at least in terms of consistency. Now, looking at the source and info for `file-attributes', it looks like this isn't actually dealt with in XEmacs, partly (I suspect) because the 31-bit int here makes it a non-issue. Anyhow, would you see a problem with returning this cell as a cons under XEmacs as well as Emacs? I don't anticipate so, but I figured to check with the experts first. :) Daniel Footnotes: [1] Remote file access via an rsh/telnet shell... -- My definition of an expert in any field is a person who knows enough about what's really going on to be scared. -- P.J. Plauger