xanthian@zorch.SF-Bay.ORG (Kent Paul Dolan) (08/29/90)
winans@sirius.mcs.anl.gov (John Winans) writes in news.admin: >Lets not start an a.s.p argument here again/too. We should think of a.s.p >as a challenge. I get it from one hop away over a T1 so it doesn't bother >me all that much. And I know it is read here because I occasionally look >at my nntp report(s) like a good news.admin. So I will keep it. > >HOWEVER: >What a.s.p might be good for is a place to test image compression programs >and stuff like that. The acive users of the group would probably not want >to see it go away and might be willing to work with others to try to reduce >it's load on the net. Hmmm. Good idea. Here's one suggestion for an experiment. Somewhere on the net, a month or two back, there was a suggestion that a good way to compress realistic images was to XOR scan lines after the first with their predecessor line, then LZW compress the output. The first idea is good, since scan coherence will give a big information reduction; but the result of the XOR is not the kind of data (strings of repeating bytes) that LZW does well on; it is more "speckley" or "noisy". However, since for coherent scan lines, the XOR result will have few one bits, providing a very skewed distribution of byte bit pattern frequencies, this is exactly the kind of data that arithmetic compression shines at crunching. Anyone who has access to a wrapper for doing archives and a good arithmetic compression algorithm might like to see what could be done with this approach. I'd give it a try, but I rarely remain coherent (I'm severely depressed) long enough to do serious code development, and my own improvements on arithmetic compression have sat working, but ugly and still needing speedups, untouched for months now. Anyway, a thought for the ambitious interested in making himself/herself a name on the net. Note the Followup-to: line if interested in participating in a discussion of this idea. Kent, the man from xanth. <xanthian@Zorch.SF-Bay.ORG> <xanthian@well.sf.ca.us>