Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sun-barr!olivea!mintaka!think.com!snorkelwacker.mit.edu!bloom-picayune.mit.edu!athena.mit.edu!pshuang From: pshuang@athena.mit.edu (Ping-Shun Huang) Newsgroups: comp.misc Subject: Re: Advice needed on data file design Message-ID: Date: 20 Jun 91 18:55:59 GMT References: <1991Jun20.124643.19298@kcbbs.gen.nz> Sender: news@athena.mit.edu (News system) Organization: Massachusetts Institute of Technology Lines: 25 In-Reply-To: Peter_Gutmann@kcbbs.gen.nz's message of 20 Jun 91 12:46:43 GMT In article <1991Jun20.124643.19298@kcbbs.gen.nz> Peter_Gutmann@kcbbs.gen.nz (Peter Gutmann) writes: > The idea is (I'm told by Mac people) that eg for Macs the file-transfer > programs can recognise the file from the ID in the first few bytes Whether you should bother to put in the header information depends on your application. I think the Macintosh people to whom you were talking were referring to the MacBinary header which is 512 bytes prepended to Macintosh files for telecommunication purposes which store information about the file that is special to Macintosh file systems as opposed to simpler file systems such as MS-DOS. Unless you're willing to customize the telecommunication protocols so that they can do something special with your files as they are tranferred, and unless there's some reason you would want them to do this, it really makes very little difference. I can't think of very many situations you would want to put information into the file header so the xfer protocol can give differential treatment of your data files. It is true that the leading four bytes might be used to identify the particular file being tranferred, but that's what filenames were intended for, after all, and if you're using true Ymodem (as opposed to simple extensions to Xmodem) then the xfer protocol will already know the filename anyway. -- Above text where applicable is (c) Copyleft 1991, all rights deserved by: UNIX:/etc/ping instantiated (Ping Huang) [INTERNET: pshuang@athena.mit.edu]