Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!samsung!crackers!crackers.clearpoint.com!solensky From: solensky@animal.clearpoint.com (Frank T. Solensky) Newsgroups: comp.protocols.tcp-ip Subject: Re: Message compression Message-ID: Date: 4 Apr 91 11:41:20 GMT References: <9104031437.AA14650@saturn.acc.com> Sender: news@crackers.clearpoint.com Organization: /home/solensky/.organization Lines: 33 In-reply-to: teb@SATURN.ACC.COM's message of 3 Apr 91 14:37:37 GMT In article <9104031437.AA14650@saturn.acc.com> teb@SATURN.ACC.COM (Tom Benkart) writes: In applications involving file transfers (FTP) over low speed lines, has anyone investigated any type of data compression on a message by message basis? I'm familiar with the header compression algorithms, but don't know of any data compression algorithms. The pay-off might not be very high, but any compression could be useful. One of the problems with a number of data compression algorithms (eg: Lempel-Ziv encoding, the one used by the Unix 'compress' command) is that they need to be able to look at the entire data stream before being able to compress any part of it. In this case, it would be about the same as running compress yourself and then FTPing the resulting file. Another approach would be to use something like an adaptive Huffman code (the details of which I know nothing about). The '/usr/old/compact' command on Sun's Release 4.1 takes this approach by encoding the data as it is being read in and sending it out immediately. The man page lists the following reference (and also indicates that the command will NOT be distributed or supported in future releases): Gallager, Robert G., Variations on a Theme of Huffman, IEEE Transactions on Information Theory, vol.IT-24, no. 6, Novermber 1978, pp. 668-674. I believe that the Applications Area of the Internet Engineering Task Force is planning to look into improvements to FTP such as this at some point in the near future, but I'm afraid I don't know what the time frames or immediate plans are. -- -- Frank Solensky / Clearpoint Research Corp. Red Sox magic number: 163