home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!paladin.american.edu!howland.reston.ans.net!zaphod.mps.ohio-state.edu!uwm.edu!cs.utexas.edu!sun-barr!news2me.EBay.Sun.COM!seven-up.East.Sun.COM!sungy!news.Sweden.Sun.COM!swippy!seunet!sunic!corax.udac.uu.se!Riga.DoCS.UU.SE!andersa
- From: andersa@Riga.DoCS.UU.SE (Anders Andersson)
- Newsgroups: comp.protocols.tcp-ip.domains
- Subject: Re: Positional data within DNS
- Message-ID: <1jpiimINNogl@corax.udac.uu.se>
- Date: 22 Jan 93 19:40:06 GMT
- References: <VIXIE.93Jan14011616@cognition.pa.dec.com> <1jfj43INNr6e@csc2.anu.edu.au> <1jghnrINNl8v@daisy.ee.und.ac.za>
- Distribution: inet
- Organization: Uppsala University, Sweden
- Lines: 18
- NNTP-Posting-Host: riga.docs.uu.se
-
- In article <1jghnrINNl8v@daisy.ee.und.ac.za>, barrett@daisy.ee.und.ac.za (Alan P Barrett) writes:
- > Now that we are considering a 32 bit structure, why not just use a signed
- > 32 bit value to represent multiples of 2**-23 degrees? (Or think of it
- > as a fixed point format: sign bit, 8 integer bits, 23 fraction bits.)
-
- I agree to the single value (makes for simpler processing), but I
- think the base unit should be the second, and not the degree, to
- avoid unintuitive rounding of what is displayed after the user
- has entered integer degrees, minutes and seconds (ok, the display
- routine could probably make an educated guess as to what the
- original data was, but why make it difficult in the first place).
-
- To cram even more data into the bits, the fractional second might
- be something different from a mere power of two.
- --
- Anders Andersson, Dept. of Computer Systems, Uppsala University
- Paper Mail: Box 325, S-751 05 UPPSALA, Sweden
- Phone: +46 18 183170 EMail: andersa@DoCS.UU.SE
-
