home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.protocols.time.ntp
- Path: sparky!uunet!pipex!warwick!pavo.csi.cam.ac.uk!maj
- From: maj@cl.cam.ac.uk (Martyn Johnson)
- Subject: Re: xntp3 refuses to talk from little-endian to big endian machines
- Message-ID: <1993Jan12.173201.25868@infodev.cam.ac.uk>
- Sender: news@infodev.cam.ac.uk (USENET news)
- Nntp-Posting-Host: ashby.cl.cam.ac.uk
- Organization: U of Cambridge Computer Lab, UK
- References: <1iukkuINNhmg@wintermute.phys.psu.edu>
- Distribution: inet
- Date: Tue, 12 Jan 1993 17:32:01 GMT
- Lines: 24
-
- In article <1iukkuINNhmg@wintermute.phys.psu.edu>, kenh@leps5.phys.psu.edu (Ken Hornstein) writes:
- ...
- |> It _looks_ like there is a byte-order problem, but
- |> I have not gone that far into the problem. Does anyone else run xntpd on
- |> a little endian machine? If so, please let me know what you did to get it
- |> to work!
- ...
-
- xntp2 had SEVERAL byte order problems in the xntpdc interface, though I
- did not find any in the main protocol. It was fairly obvious that whoever
- wrote the code had been fairly sloppy about hton and ntoh functions. In
- some places the wrong one of the pair was used, which doesn't cause anything
- to go wrong, but is an indication of carelessness.
-
- I sent off patches for these problems to somebody who was collecting them,
- but I have no idea whether they ever made it into the definitive sources.
-
- I haven't yet done the exercise for xntp3, so I cannot verify your assertion,
- but it would not surprise me to find byte order problems.
-
- --
- Martyn Johnson maj@cl.cam.ac.uk
- University of Cambridge Computer Lab
- Cambridge UK
-